Updated Match Maker is not balanced

[digital orchid]

Majority of my games have been camp fests with hoards of players all sitting in one corner

[proper valley]

Yeea. That sounds about right

Unfortunate. Truly

[fleet elm]

Is that new? That sounds like every match from the last 10 years

"my experience is more important than anything else, including actual data"

[paper anvil]

I agree there, i was expecting BOTS and had tons of players instead.

[paper anvil]

That or they YOLO and Die XD

[visual gorge]

I have played Westfield everyday for months and even since Komarin was reinstated. However, I am playing about 20+ battles per day. You have just have had bad luck and timing with your battles.

[digital orchid]

Na a lot of players had confidence before these changes

This is like we’re back to the dark ages

Nobody does anything

[digital orchid]

Dude I’ve had Westfield maybe 3 times since Komarin was added

Must have had that map 20x at least at this point. Sometimes it comes up 3-4 times a session

It’s absurd

Thank god it’s only a temporary map at tier 10. It’s so bad at high tiers

Why couldn’t we have gotten sacred valley back or something

Even wide park at tier 10 would be better

[visual gorge]

I am ready for Sacred Valley to return.

[digital orchid]

https://tenor.com/view/draco-malfoy-harry-potter-mewing-mew-slytherin-gif-8263298459687914233

Sacred valley enjoyer 💪

[ocean reef]

Will play even less

[ocean reef]

I recommend that nobody spend anymore money.

[thin quest]

When bad players are thrown into teams with good players to make the good player get a defeat how do the bad players get a win to get "uplifted"? When you go full tinfoil please make it make sense.

Is this something like theSchrödingers cat experiment? Both win and lose until you loot at them?

[thin quest]

Maybe get gooder?

[proper valley]

But on a serious note this match maker blows

Like. I know I'm in the mostly uneffected range.

But oh boy do the results show.

[paper anvil]

haha Yeah WTF? i didn't see that coming XD

[ocean reef]

Garbage company.

They have turned WW2 into cold war where there are minimal tactics and everyone heads in one direction to win. I would imagine they put developers in charge that don’t actually play the game.

[candid fable]

Game is not garbage but MM is for sure.
4 defeats in 4 games most under 5 mins.
So yeah, MM is obviously balanced and matches are clearly longer 👍👍👍

[vague spade]

Accuracy changes hahaha what accuracy changes

Still getting snapshotted 400m+ what a joke

[ocean reef]

Still continuing my streak of not purchasing crates. Not until matchmaker is made completely random and not based on any metric like winrate or skill.

[uneven shoal]

Ok

[lament mica]

It hasn’t been completely random in years. Probably the whole time.

[coral zephyr]

I can see where you would not want to purchase WW2 premiums when the MM is working against you whenever it can.

I agree random was better.

[ocean reef]

Adjustments were made about a month or two ago that made matchmaking “more” skill based with an emphasis on winrates. Which means players with a higher winrate will be placed in matches where they will have to sweat and carry the team each time in order to win. No fun sweating every match.

The only matches that i was winning for a while were those that i carried the team. Went from a 60% career winrate to having days where I was a 40%. My WN8 was 3-4k each day but the match making was placing a high winrate player(me) on potatoe teams that were guaranteed to loose.

All in the name of fairness and equality. The only “fair” matchmaking are those that are completely random. No “big brother” interference.

[autumn light]

Correct, it's never been entirely random. It has used battles played to sort in the same way it does now for years, win rate was added to the points calculation. That was the only change.

[ocean reef]

The “only change” of adding the win rate punishes players who have a high win rate and puts them on teams where they are required to sweat and carry the team each time. I play the game to have fun, not to sweat every battle for a win.

[ocean reef]

The “only change” of adding the winrate is what diminishes the enjoyment of playing the game. Telling people if you get too good then all the matches will become sweatfests and you will rarely win until your winrate becomes an equitable 50%. The winrate was added to help out the low winrate community.

Adding winrate to the matchmaker is like placing a tax on those that have been successful at the game. It’s like the government took over the game…..transferring wins from those who earned them to those who did not.

[daring grail]

As you can see, the game’s community has reacted extremely negatively to new matchmaker.
The only question left to ask is — why was all of this even necessary?
And when will you bring things back to how they were before?
Honestly, you (Wargaming Console) are focusing on the wrong things.

[coral zephyr]

They did what they thought the player base wanted. People said they wanted 'more even' or 'more fair' teams. Or something along those lines.

However this is not what they meant nor wanted. What did they mean then?

People didn't want all the zero damage players on their team. We got matches with 4-7 players on one team that all got zero damage. Those matches were unwinnable. That was the complaint.

Now we have "adopt a zero" if you have a positive winrate you automatically get a "zero" player on your team. A really good player? How about 2 or more extra zeros. Not much fun. The power of a "zero" player is great. Great for losing.

This has made WW2 a lot less fun. +1/-1 helped a lot. Power levels were closer. But 2 tiers down and your adopted "zero" is in a top tier on your team? Sorry you lose.

The thing is the "zeros" likely didn't care anyway. Win or lose, they're not shooting their guns either way. Maybe mid players are doing better (48-50%) win rates. Because 55% win rates get stuck with more zero players.

As you can see I am still not happy with it. Playing all tier 10 keeps you from bottom tier. But I have hundreds of tank tier 8 and lower.

Go to CW? Balance is still off there too. I am stuck between a rock and a hard place.

[fleet elm]

Has the playerbase as a whole reacted negatively? Or just the top 1%? It's hard for us to tell since the vocal community is far better than the overall playerbase

[viscid gorge]

Depends on where you look. The vocal here are the best players, vocal elsewhere don't seem to care

[ocean reef]

Still have spent zero on war chests this season! I usually buy some each week. Until matchmaker reverts to where it was a few months ago I will not purchase anything but the season pass.

[daring grail]

You could simply create a regular survey on the official website —
or even show it directly in the game hangar when players log in.

A message could pop up:
“Please take a short survey about the new matchmaker.”

Then add a few simple questions like:

-Do you like the new matchmaker?

-Has the game improved for you?

-Or did you not notice any difference?

Just basic questions that players can answer without thinking too much.

Good idea, right? What do you think?

[coral zephyr]

My guess is the vast majority of players have not even noticed.

[coral zephyr]

It is also my guess that the Discord community are the ones who brought up MM balance. They Discord community is more involved and likely averaging in the top 35% of players (1 MOE). Unfortunately the MM change has negatively affected the top player.

Be careful what you ask for...

[thin quest]

Do we have any conclusion from the sbmm test and +-1 MM? When can we expect it?

[fleet elm]

SBMM isn't a test. It's here and not going away bc WG said so

[fleet elm]

More specifically, sbmm doesn't change queue times or anything like that

It shouldn't cause any mm disruption of the sort WG cares about

[bleak oriole]

It's unlikely that we will ever go back to not looking win rate - the impact for the average player has been net positive by elimiting the really egregious outlier matches that 9.4 (and before) could produce.

We are aware that the outlier players aren't happy with it, but we did listen to you all, grabbed the data, and tuned the system somewhat. We are also paying attention to the rest of the feedback here in case someone comes up with an interesting proposal that meets allows us to achieve the goals.

I haven't really seen people present evidence that the tuned system is horrible and continously unfair - at least outside of instances where platoon distributions force "unfairness".

[fleet elm]

It's continuously unfair to me. I get worse teammates than average just because of my win rate and nothing has been done to mitigate the fact that it might take hundreds of battles for winrate to normalize

[stray star]

i get its not the largest sample ever but its still kinda crazy that this can happen

[uneven shoal]

Any consideration of a check to attempt to balance tanks between teams?

Two Type 5 on one team, one on the other vs all three on the same team

[stray star]

i love it when 3 grilles get put on one team and the other team gets 2 badger's

[fleet elm]

Subclass mm probably won’t happen. Queue times and bot games

[cunning dew]

I agree

Im tired of loosing game after game

[coral zephyr]

If you want use winrate, just balance the number of very low winrate players on each team. Give each team the same number.

However, I would still rather look at damage ratio rather than winrate. But again only balancing the number of low damage ratio players per team.

That being said it is also difficult to make completely meaningful suggestions when the 'algorithm' is shrouded in a level of mystery. It has been said 'changes were made'. However. It hasn't been said what exactly those changes were.

[coral zephyr]

From the outset. WOTC, now WOTMA, was set up to track winrate. For players aspire to improving and maximizing their winrate.

But now if one does they get saddled with the worst players at every opportunity. That doesn't seem to encourage winning anymore.

Any system that starts to discourage the highest performing individuals will see those high performers migrate to a different system. That is just how it works. That is not a recipe for success.

I say this in an effort to help. Obviously there are still players out there whom feel the MM is actively working against them. Which it is.

If outlier players (relatively really good players) are uncommon or rare, certainly in the minority, why are they being singled out? Why does the MM actively work to give them the short end of the stick?

I have seen players with near or over 100k matches yet a 44-46% winrate. Obviously losing more matches is not slowing them down. So there really is no need to cater to them.

[viscid gorge]

Damage Ratio is hard to balance around as well.

Players who play superheavies have lower damage ratios compared to that of TD players

[ocean reef]

The game was great until all these unnecessary updates. Only 3 updates were needed. +1-1 matchmaking. Max 1 arty per team. Random matchmaking not skill based or based on any other metric.

The moment you start to punish people for winning is the moment they start to play less. True for me, I play half as many games now. Probably less. I have not purchased any chests all season because I am unsatisfied with the product wargaming is making.

[autumn light]

in real world situations, what you said should happen is effectively what happens the vast majority of the time - especially since the changes mentioned to fine tune the formula. Even with outlier players. In terms of what has actually changed in results you are just finding that the low win rate players are more evenly distributed to both teams, whereas before (as you said) it'd be more random. So say there is 12 players in the game with a "lower win rate", in the past that could vary from 12 on 1 side to 12 on the other, and any combo inbetween. Now what tends to happen is there is a near 50/50 split consistently, sometimes it may be 7 on 1 and 5 on the other but the number of times it's say more extreme like 2 or 3 on 1 side are now somewhere between drastically reduced to completely gone. Which is, by all definitions and ironically often even in the feedback of people who dislike the change, more even.

The point from "outlier" players is they are most likely to get 7 or 8 on theirs, while the other team gets the 4 or 5 - and this will happen consistently whereas before they had a pretty random distribution (although that did include getting 7 and 8, and even 9, 10, 11 or 12). Which their point is that they feel it is less even for them - which brings up some kinda complicated debates on the value of a player to a team that i won't even try to get into 😅 .

to be unpopular, having seen the way the formula works, unless your win rate is over mid to high 60% then it's very unlikely there is any reality to the current MM having any "stacking" of "lower win rate" players on your team. The delta system wouldn't result/allow it now

[coral zephyr]

12 low rate per match?

Maybe we need to have low win rate and very low win rates categories.

The power of not even shooting your gun at all is strong for a loss.

But certainly you can understand mine and players grief when we are not told what is exactly going on and what was changed. Just trust us it looks better now. The mids really like it.

It seems to have been confirmed that better players will get the short end of the stick...but not TOO OFTEN. How much is too often. Many say any amount. No one wants the system working against them at any amount. That's why I say leave the average and better players completely out of the algorithm. Or at least count anyone +49% as the same. Then it would be random who gets blessed with the best set of sub winrates.

I can only go on how it feels for me. It doesn't feel great when half my team sits on spawn and don't shoot. Meanwhile the entire other team advance and tears them apart. When I get an hour of that I shut it down.

[autumn light]

"so say there is" AKA it's a hypothetical 😅 . A low win rate player is by definition is pretty subjective... for a 1% players it's the other 99% after all 😄 But just in case, for clarity there is more than 2 brackets lol

No one wants the system working against them at any amount. With a more random system then the system will potentially give you 100% of the "worst" players, which was (As you mentioned) the specific complaint often voiced here, the forums, reddit, etc for years and years - i'd wager some unhappy with the change have either said or voiced that complaint themselves - was that it was unfair that MM would often give you an "unfair" amount of the "worst" players and we should do something about it. And we did, now it is very unlikely you get an "unfair" number of "worse players" on your team, it will be pretty consistently an even split.

And again to be unpopular, i think the issue is a lot of people have focused on this change and decided it's the reason for what will/does happen in the game no matter what. And if you need some objective non wg-bias proof of that, look at how many people have complained about it's impacting their games to then have been playing only CW, or people who have had 1 bad week 6-7 weeks after the change and NOW it's the mm change fault.

In terms of the "short end of the stick" "how often" thing, i guess the complicated debate there is that in theory before they would get a 50/50 split of "the worst players" on each side... 50% of the time they had the most and 50% the other team did. And in theory that may now be that 55% of the time they have the most worst players on their team. Again, hypothetical here 😄 But, the difference is previously that 50% included anything from 1 more worse player on your team to 14, whereas now that 55% is anything from 1 more worse player, to 3 worse players, and VERY rarely 4. So i guess which is worse?

[coral zephyr]

55% of the time having the MM give the worse team is significant. Before I could say it was just bad luck. Now I never know if the MM is balancing me or it was bad luck.

That is a distinct change. Going from random to getting balanced any amount. 55%? I think 5% can separate bad from average (45-50) and average from good (50-55). So 55% if the time getting a significantly worse team is not good.

I have said before that random was better than whatever we have now. With random, I can hope my luck improves. With this a big chunk of that luck is taken away and I get shafted.

[daring grail]

The matchmaker now aims for a 50% win rate… for solo players.
That’s the whole point.
Instead of making the game better,
players with good performance are being punished by the matchmaker.

They’ll come here with negative feedback,
and you’ll just tell them everything is fine!

My advice — just play in a group if you want better results.

[coral zephyr]

I never know when the MM will be actively working against me and by how much. So if I want to win I have to play only very strong tanks, because I have to carry a couple extra players whom historically perform very poorly.

Likely tier 10 tanks are my best option. Getting a match two tiers down is very often a loss.

I was watching a CC video review on the new tier 8. He says it is a fun tank but not really a competitive tank. So I guess that is not for me. I need some very consistent to carry. But I can't pay it too much or it might get nerfed.

These are the things I have to deal with now.

I think a big part of the issue between team balance is the current state of +2/-2. The power difference between tanks two tiers apart is simply too great. It can't be balanced out anymore.

The week of +1/-1 was much better.

[daring grail]

The problem isn’t with the ±2 matchmaking.
The problem is with the matchmaker itself.

For example, tier 7 tanks have been sacrificed to tier 8s.
There are way too many battles where tier 7s face tier 8s.
Same thing with tier 9s against tier 10s.

Instead of increasing the number of 10 vs 10 battles
(on PC those were called “tier 11 battles”),
here we almost always have lower-tier tanks in the mix.

In my opinion, the issue isn’t the ±2 spread —
it’s the matchmaker itself.

[ocean reef]

Matchmaker maker is garbage. Instead of making matchmaking random we punish people with high winrates

[coral zephyr]

That is poor take on it.

Insults will get you no where.

[near pond]

There are a few. I just opened the thread and didn't see how old this message was. But yeah, Warframe managed it (albeit they use PC accounts as the mediator, something that obviously can't be done here)

[ocean reef]

Matchmaker that incorporates winrates incentivizes losing

[near pond]

It's not the matchmaking that was changed

[ocean reef]

Matchmaking changed. They incorporated winrates. That is a fact which was confirmed.

[near pond]

The players in the match are the same. Just how they are distributed was changed

[near pond]

It's not the same thing.

[ocean reef]

The matchmaker has changed, it now incorporates winrates. It was done about 2 months ago. This is an effort to equal out winrates and reallocate wins to the low winrate community.

[autumn light]

again to just point out, i do keep pointing out i'm using hypothetical numbers that are easy to use and not anything actual 😄

But even within that you also changed/misunderstood what i said.

What i said is that previously half the time you would get a random distribution, that WOULD include a significantly worse team. NOW, you could say get 55% of the time a worse team but a much closer team every time. A significantly worse team is now highly improbable, whereas before it was a certainty. But previously you'd also get a significantly BETTER team as much, whereas now you will get an either slightly better, a slightly worse or about the same.

[coral zephyr]

I understood you.

[autumn light]

then it was quite a massive distortion in what i said 😅

[autumn light]

Technically what Cheese is pointing out is the MM itself was not changed, the way the 30 in battle are chosen has not changed. It's how they are sorted into teams once picked

he is correct in this, the way the 30 are picked has not changed and does in no way look at win rate

Which is why it would be impossible for that aspect to try and "aim for 50% win rate" or work like any true SBMM

[ocean reef]

With the updated matchmaker we punish success by taxing good players and reallocating their wins.

[near pond]

it's easier to break it down into "matchmaker" and "team distributor" (even if it's not technically the case and both are part of the same system) at this point

[fleet elm]

I mean the answer is that we’d take every 3rd game with a team ranging from bad to trash rather than 80% of games having teams ranging from mid to bad

[coral zephyr]

How was it distorted?

[autumn light]

#1405193097079689458 message yeah i addressed a few back here, there is not many and i guessss a pedantic argument is it was ported to console and not made as a seperate game from the start. But the main point was more it's not common, like very very rare 😄

[near pond]

Yeah. It'd be nice. But so would 1 3 way crossplay game between PC and both consoles. Warframe managed it by using PC accounts as the mediator that you can acess from Xbox, PS, Switch, Mobile and obviously a PC

Obviously the chances of the games merging are astronomically low

[autumn light]

So 55% if the time getting a significantly worse team is not good. i said that in this example you may get 1 or 2 "worse players" on your team, compared to the old system which would allow for the worst 14 to be on your team. So the point was the literal opposite, what i said was there may be an increased chance in having a very slightly worse team, compared to a 50/50 chance where you got a FAR worse team but you got is often as a FAR better team. In no way or nowhere did i say a significantly worse team and 55% of the time

[autumn light]

well, here is 2 also completely proving nothing screenshots of past 30 day stats, but i have 2 of them so i guess i must be right?

[coral zephyr]

I consider 2 extra zeros significantly worse. Even one extra zeros I need to carry can be too much.

Before the randomness averaged out. Now we get worse teams at greater than 50%. Maybe 55%.

I am not trying to be argumentative. I am simply trying to relay the feeling of myself and other players. Don't feel attacked. I don't think you are the final say in the MM. Even that person should not get offended.

[near pond]

[ocean reef]

Instead of celebrating wins and success WOT has chosen to punish players with high winrates so we can lift up the low winrate community.

[daring grail]

Yes, you’re right… your screenshots are from Cold War, while mine are from WW2.

[ocean reef]

Wot is promoting losing. Why try to increase your winrate if it forces you to sweat every battle?

[near pond]

[ocean reef]

It’s similar to taxes. If you are successful the tax rate increases so the people that place no effort in the game can have wins for free.

The most important aspect to wargaming is the reallocation of wins. From higher winrate players to lower winrate players.

[lament mica]

Yowsers that is a terrible example to argue your point.

[near pond]

I don't think Shock understands Taxes

[ocean reef]

Wargaming should celebrate players that have high winrates. Instead wargaming punishes players with high winrates.

[autumn light]

1 is CW, 1 is a mostly ww2. But there WW2 stats don't change. How about some WW2 mostly/only ones then. And this is just looking at 1 clan so i'm being pretty low effort 😄

My point was more that a random sample screenshot proves nothing

[fleet elm]

Wins in a game = / = real life

[ocean reef]

Who does not like winning?

If you are punished for winning. Why play?

[autumn light]

not attacked in anyway, but i guess a point on subjectiveness. I think saying having 1 worse player (and how worse is not even state/given) is not a "significantly worse" team, and quoting my example and adding the words in is pretty misleading to what i said/was saying.

[ocean reef]

Uh oh you are winning to much. Let’s change matchmaker so your winrate is more equitable.

[fleet elm]

Is it “one worse player” or 3?

[cunning dew]

I have noticed it is better since you updated it. I was at 70% wins (you cant see anymore because its been averaged) but its not the worst. I dont like it but its not the end of the world for me

[daring grail]

Is this the “solo player” statistic?
It’s strange that my average damage is almost the same — or even higher — yet my win rate is 10% lower…
And all my battles were played solo.
It’s not clear here whether this is a solo player’s stat or just a general player stat (including group battles).

[coral zephyr]

You don't understand. I am saying it is significant too ME. Perhaps having 1-3 extra zeros is not significant to you. But it is to me and others. Having the odds stacked against me ANY amount is significant. That is what I am saying. I am not saying you consider it a problem.

[autumn light]

whereas now that 55% is anything from 1 more worse player, to 3 worse players, and VERY rarely 4
Most recent direct quote ^
over the several messages i was trying to say it's most likely to be even split, but that having "1 worse" player more often than another player is the hypothetical with more extremes of 2 , 3 or AT worst 4 (not sure 4 is actually possible with the changes but i imagine it mathematically is). And thanks to the way it works, a "worse" player could be you got a 55% player rather than a 56% player. The delta system doesn't care if a player is a 44 vs a 46 or a 55 vs a 60 after all.

[autumn light]

but i guess a point on subjectiveness I literally started my reply with a point that it matters what each person finds subjective?

[coral zephyr]

You added misleading.

[ocean reef]

I have played close to 25k battles. I have thoroughly enjoyed the game. Until 2 months ago I have never posted anything on the channel. This was my go to game. I disagree with the changes and do not enjoy the game anymore. When you update a game and you have longtime customers dissatisfied with the product to the point they stop purchasing product from you and reduce their playtime it will impact your Profit and Loss. If your goal is to make the game equitable then this is not the game for me.

[fleet elm]

Still would rather it be equal worse and better players per team. An improvement over the prior sbmm I guess

[native plaza]

Erm your using cold war data which is irrelevant 😂

[autumn light]

yes, because my point about it being misleading was that you were quoting back my example and then changed my wording to be your opinion, it wasn't stated as "i would find that X" it was worded as if i had said that.
That is a distinct change. Going from random to getting balanced any amount. 55%? I think 5% can separate bad from average (45-50) and average from good (50-55). So 55% if the time getting a significantly worse team is not good.
I personally do not agree that having a 55% player is a significant disadvantage over having a 56% player, but the way it was written implied that i had said

So at no point have i tried to take away from your stance, or say it is not valid, but i don't think it's fair to quote something i said but change part of it to reflect your opinion without stating that it is your opinion. Perhaps you didn't intend it that way but it certainly read that way to me.

[autumn light]

my bad 1am humour going over more heads than i thought 😅

[autumn light]

fair stance

[coral zephyr]

You don't understand what I was saying. That's okay. It is confusing talking about something happening 55% of the time and also about a 55% winrate player.

[autumn light]

I'd add a generic and not aimed at any single person comment, i think the issue is a lot of discussion around this will use assumptions and unrealistic situations to define it because it's such a high variation thing that it's almost impossible to look at it accessibly otherwise. Due to the way it works though, the system really doesn't care about an individual win rate, it compares them within the pair.

So in 1 game as a 55% player you could be an "outlier" where you get matched against a 45% player and create a larger delta, meaning you get "the worse" next player of a pair on your team to balance it out... a 62% player matched against a 65% player. In the game after you could be matched against a very similar stats player and it just leaves you in your pair order as you're delta is 0. The way/examples people are using in this chat to define expectations are frankly very unrealistic even with the knowledge of how the system fully works.

using a very crude paint thing (X = number of worse players from the pairs, Y = number of battles. Brown = old, Yellow = new), in the hypothetical i was discussing @coral zephyr what i was trying to say is that the shaded brown area use to happen, but shouldn't happen now (you get 11 to 14 of the pairs) and the shaded yellow bit would have happened in the old system too. So the change is the opposite extreme of the team tilted in your favour doesn't happen, and an increase in what people said are more balanced matches. It's not entirely accurate as the yellow should have different curves but it's late and i'm using MS paint 😅 but those should bow towards the middle a lot more.

And that is again, before factoring in that the system doesn't care about the win rate specifically, it cares about the delta. So a worse player in a pair can be any win rate. It's in relation to their opposite in the pair.

[autumn light]

it's possible there is some crossed purposes then, because i didn't mean win rate in anything i said and it would seem you did. I specifically meant the So 55% if the time getting a significantly worse team is not good. bit, where to me it reads as if you are talking about an amount of time and directly referencing this that i said But, the difference is previously that 50% included anything from 1 more worse player on your team to 14, whereas now that 55% is anything from 1 more worse player, to 3 worse players, and VERY rarely 4. So i guess which is worse?.

If that wasn't what you meant then fair enough, will put it down to end of long day trying to read a lot of stuff and reply quickly to go to bed before baby woke up (the long gap in my replies likely suggests how well that plan went lol) but i'd hope you can see why i read it as i did 🙂 It's all in all a kinda asinine issue as it doesn't change what i said - which was an open question about preferences - nor does it take away or invalidate your dislike of the change 🤷‍♂️

[ocean reef]

The game is no longer fun. WOT forces you to spend millions to re-equip previously fully equipped tanks. Reduced the quantity of skills for customization. Made skills weaker. Made equipment weaker. Implemented a “Fair and equitable” matchmaker whose goal is make everyone’s winrate 50%. None of these changes appeal to me.

I bet the next update will be to have 3 Arty’s per team, +3/-3 matchmaking, and matchmaking that always places the highest winrate player at bottom tier.

[coral zephyr]

I think you are starting to understand.

You said "And in theory that may now be that 55% of the time they have the most worst players on their team."

I replied: "55% of the time having the MM give the worse team is significant."

Why is 'only' 55% significant? I said it is 5% that separates a poor player from an average player. So a mere 5% is not a small number.

But really having a "the most worst players" 55% of the time is not a change of 5%. It is now 45% good vs 55% bad. [the total is still 100%] This makes the difference 10% of the total matches, not 5%.

[although I understand 55% is just a selected number. No one may know the actualy number. One would need to start tracking a lot of matches to see what the tread is. But that is the number you used, so I am following suit]

Before we randomly got 50% good/50% bad. The extremes may be less now but the odds are worse 45% good/55% bad. Seems like 50/50 was better.

[autumn light]

it's cheap MS paint (#1405193097079689458 message) but it probs shows the point i was asking about better, but for clarity it was a sort of open ended question saying that it's subjective if someone would find that change better or worse. While also pointing out that a lot of what is being attributed to the changes would have happened before too but is now being focused on and blamed on the new system.

[coral zephyr]

The problem arises that it COULD be the MM now. Before it was always just bad luck. Over time you luck would equal out to randomness.

The MM could also be working against a person every match. If only to a varying degree. If I take out my Fast Track, I don't know if I will get "balanced" or not, but the MM will certainly try.

The new MM only increases the "tin foil hat" theories and gives them legitimacy. The MM can indeed be working now to give a player the worst team.

[autumn light]

still cheap MS paint, slightly more accurate and slightly less 😄 but the curve is probably? and more realistic distribution with the grey filled in bits being what would have happened in both MMs... Scaling is awful lol

[coral zephyr]

That is why I say just don't try to "balance" good players. Don't introduce a mechanic that can negatively impact them. If win rate must be used, consider 50% all the same.

For blow outs and short matches, it is likely more attributed to "poor" players who get farmed. Players that don't get any damage, but die quickly.

If anything only address the worst players. That way the better players are still random and have nothing to worry or complain about.

But it is possible a completely random MM is best overall.

[autumn light]

But it was also a very common complaint that it was unfair when 1 team gets say 4 very high win rate players while the other gets none? That is FAR more likely to cause a blowout than 1 team have say 1 or 2 more lower win rate players? And again, to keep stressing the point, the way the delta system works is all relative to your pair. So i as a 66% player could be a lower win rate plater if i was matched against someone with a 69% for example.

Now that ignores that there is brackets to how it works, but the point remains. Similar for an "outlier", that could be a 51% matched against a 38%, but it could also be a 64% matched vs a 52% and so on. The assumption that keeps being applied is that it will give you the 45% player because the other team got a 65% but that isn't how it works.

[coral zephyr]

I am just relaying feedback. Currently the mm seems to be discouraging to many people. Whether getting stuck on a vastly outmatched and losing team is a result of the MM or not, the MM will get blamed. Of course, the MM is placing the players.

That person with a 60% winrate loses 40% of their matches. Now for 40% of their matches they are mad at the MM. That is simply human nature. Can it be changed? I certainly have no suggestions to change "human nature".

I am attempting to give suggestions that would avoid player complaints.

[lament mica]

I wonder if people would still be upset if the teams were perfectly balanced and they still had a lower win rate. 🤔

[viscid gorge]

Unlikely considering the main complaint is the feeling of having to carry your team to victory and feeling like whatever flank you aren't on is a losing flank

[lament mica]

They would have a lower win rate still—indicating they do need to carry more as opposed to previously.

[viscid gorge]

Not exactly. Before the changes, one could easily get 65% winrate because the teams were random, now those same people get less than 60, and if they still got less than 60 but their own teammates were better, you would not need to carry more.

[fleet elm]

If teams were perfectly balanced, people would have even lower winrates, approaching 50% so I’m guessing people would complain more

Good players would get even worse teammates

[lament mica]

That’s just not correct. But is the response I was expecting in part.

[lament mica]

Not if perfectly balanced. Teammates would be even both sides.

[viscid gorge]

Classic manuever. Don't explain yourself, just say "you're wrong"

I expected that reponse 🙂

[lament mica]

Not everything is about what you want.

[fleet elm]

Then it wouldn’t be perfectly balanced

[viscid gorge]

That has nothing to do with what I said. Great arguments here

If you're gonna ragebait and troll, the least you could do is be good at it

[lament mica]

I simple posed a question. No one forced you.

[lament mica]

Well I guess you believe no one can equal you, but if you paid attention to the original premise—perfectly balanced teams.

[fleet elm]

Ah so 5-10 min queue times for good players

Sounds even worse

[lament mica]

That’s funny because you have said before you wouldn’t mind higher queue times for better matches.

Besides the fact you have no idea what the queue times would actually be.

[fleet elm]

Not sure I’ve ever said that. Maybe in the context of tank subclass pairing or not having bots in matches but an extra minute to get the right set of tanks is different than 5-10 minutes waiting for another top 0.1% to join the queue

[lament mica]

lol again you have no idea of the queue times. To make some of you happy they should just give you full bot lobbies—since you want to cheat the mm anyway by starting and quitting the queue, then you wouldn’t have to worry about your win rate and you can check off your MoEs.

Besides the fact you have no idea how often another 1 percent player actually comes up. That’s pretty funny since you are complaining so hard about bad players on your team. I guess it gives new meaning to that phrase that there is only one thing in common with all those losing teams.

[stray star]

true! its my fault my winrate has dropped 20% since the mm update

[fleet elm]

How is that cheating the MM any more than platooning or using stronger tanks? Using the systems in place to win more.

Lol impressive troll attempt on the 2nd part. But it only works saying that if there's no other common factors between matches. Say, like, a matchmaker that is designed to give you a worse team

[fleet elm]

It is. If you didn't want it to drop so much, you shouldn't have gotten it so high in the first place, duh

[cunning dew]

Oh yea its our fault for being good

Just screw us over why dont you wargamming

[lament mica]

Laugh all you want but I highly doubt that is the intent or the design, if it is working as you say. You are essentially creaming your teams—loading in after most of the 30 are compiled and matched so the teams are balanced before considering your pair. But you are consistently manipulating it to ensure you avoid the worst of the mm and that you drop in on a relatively balance set of teams to tilt the board in your favor. It is essentially an exploit in the system that also allows you better silver and xp gains and one that should be a ban offense for players using it given the likely violation of the license agreement.

As far as the second part—that would be the new meaning.

[lament mica]

As to the proposition, yes. If your winrate drops because your team is perfectly balanced, then part of that is you or are you saying something else was giving you a better win rate before?

[fleet elm]

It didn't do much in my experience. MM is too volitile to have a big effect. I went from 55% to 60% on the prior extreme MM. My goal was to break the MM with an exploit so WG would have to revert it. Starting a queue and dropping out after a couple seconds is a common thing in general, I just did it consistently. And I told WG I was doing it and didn't get any pushback so yeah, I had permission

[fleet elm]

You’re very good at pretending to be dumb lol. I remember a meme Reddit post I made a couple years back poking fun at people who complained about the mm. The gist was that every match I was in was unfair bc my team was twice as likely to win as the other team just by my existence as a good player with random teammates

[coral zephyr]

a bannable offense? are you serious?

[lament mica]

Exploit is an exploit—if they ban people for downloading repeatedly from the store what they shouldn’t, and exploiting game mechanics for resources, it is not different—using an unintended action to gain game resources you otherwise would not have.

[lament mica]

So you won every match 100 percent right?

[fleet elm]

It’s an intended action. Entering and leaving the queue are both normal actions

[coral zephyr]

You are conflating two completely different things. Downloading extra store items and queuing. This has the effect of degrading and destroying your credibility.

[fleet elm]

Twice as likely to win means 66.7%. The point is that with random teams and no other information, me being in the lobby and being a good player drastically changes the odds. Not because of the map or team stacking, just by existing

[coral zephyr]

As we know, a bad player will have that same power. Someone who loses 55% of their matches, has a strong correlation to losing.

[lament mica]

It’s the rapid de-queueing And re-queueing with intent to manipulate the mm for gain that would be the in unintended action. WG has also said in the past that people backing out of the queue when launched causes server issues and resources. Maybe they don’t ban the account right away. Perhaps they put them on time out first.

[fleet elm]

Except I literally talked to WG about it before trying and they said it probably wouldn’t do a whole lot. They didn’t say “don’t do it” or give me any sort of warning

[coral zephyr]

If late-queueing gives an advantage [and is "a bannable offense"], that means the system is radically slanted against better players. The System needs to adjusted, not banning good players.

[lament mica]

You keep going in circles—heat already covered the degree of “advantage.”

It doesn’t really matter the degree of advantage. There is no 1/2 cheating. Either you used something to your advantage to gain a benefit that was not intended or you did not.

[coral zephyr]

I keep being consistent

[lament mica]

Yes you literally bragged about exploiting the system and WG did nothing I guess.

Good exploiting if you are a ST I guess.

[lament mica]

Ok Crazy Ivan.

[coral zephyr]

I'm not the one who wants to ban someone for how the queue. Ban them from playing.

[lament mica]

Did I say ban him? No just that it was bannable.

Funny that you think it is ok for someone else to manipulate the mm though but are in this thread crying about the tiniest of advantage the mm cannot even out. 😆

[viscid gorge]

"It's bannable"

-> several devs have read about him doing it and seem to have no issue with it.

Your arguments are kinda pathetic 🤷‍♂️

[lament mica]

Ah another Mr Consistent I suppose.

Not really arguing as I already pointed out the inconsistencies.

[fleet elm]

It was intended. It’s the way the system sorts players. The later in the queue you are, the less the system can correct a high delta pair.

[thorny bramble]

Ngl bro your kinda just saying a lot of slop

Like i schemed thru your arguments your kinda slow

[near pond]

https://tenor.com/view/youre-your-gif-22328611

[lament mica]

Sure sure. They intend for everyone to jump in queue, back out, jump in again. Clearly it was designed for this purpose. Talk about slop.

[tawdry sky]

Pretty wild that

[ocean reef]

The game is now slanted against players with a high winrate. If I do not carry the team or place in the top 3 we typically lose. No fun playing a game where you only win if you sweat. Taxing high winrate players so low winrate players can have wins. Socialism has reached WOT console.

My streak of not purchasing anything continues!

Also playing less. 380 in last 30 days. Compared to 1757 in last 90 days.

Averaged 12.6 games per day over the last 30. I was averaging 22.95 games per day the 60 days before that.

I play significantly less now and my average games played per day is decreasing. If I get punished for have a high winrate what is the purpose of playing this game?

[lament mica]

Isn’t it?

[viscid gorge]

[scarlet mantle]

I notice this for myself. I even thinking about playing on the pc version. To much over the years I disagree with

[noble siren]

🤮🤮🤮🤮🤮

[ocean reef]

Wot is still crappy today. Terrible matchmaker that punishes players with high winrates.

[cunning dew]

Yessir

[ocean reef]

Used to play 70 matches over the weekend. On pace to hit maybe 25. The game is significantly less fun now.

Trash updates.

[near pond]

This has to be satire

[fossil girder]

Agreed. The game seems more fun to me since the update even if I take an extra couple losses than usual.

[haughty terrace]

I would agree with you there. Maybe 3% of the player base is on discord. So its really hard to tell what the majority think about it. If they even care.

[noble siren]

Turn off your string theory. It's impossible to win even a single battle now. What are you doing? Where's the team balance? Turbo battles are still the same.

[ocean reef]

If your winrate was high before they implemented the equitable matchmaking update then winning will be very difficult. I only win now if I carry the team and I am in the top three. If I don’t carry the team then we lose. The game is no longer fun.

[ocean reef]

The matchmaker was implemented to level everyone winrate to 50%. This makes those with low winrates feel good. Wargaming will start issuing out Participation Medals after every match.
Those with zero damage will get an additional medal, the “I want everyone to carry me” medal.

[coral zephyr]

Yesterday I played some CW. Got a String Theory medal. I switched to WW2 mode. But I couldn't bear to press the button to enter a match. I just shut the game down instead.

Basically I never know how much I will get "balanced" any given day in WW2. Might be 55% of the time or it might be a lot more than that. All too often it seems like I get balanced more.

It is just not fun to have the odds stacked against you any amount.

[scarlet mantle]

To play solo is actually a mess

[cunning dew]

The past 2 days ive had a 50% winrate with like 45 games😭

6k WN8 btw, idk how thats possible 😭

[grand garden]

The fundamental reason we've decided to consider win rate when matching teams is probably because distributing premium vehicles means that beginners reach higher tiers before they have a chance to improve. As a result, the skill gap between players in higher tiers becomes too large.

Proposals for addressing this issue are outlined below.

1. The higher the tier, the higher the risk/reward for silver balance will be. I imagine that at around tier 10, if players aren't earning enough damage or assists, they'll lose hundreds of thousands of silver without doing anything, but players who perform well will gain over a million silver. We'll also eliminate adjustments to balance, such as experience points gained through match wins and losses.
2. To prevent beginner hunting, we'll limit equipment and crew skills at lower tiers. For example, at tiers 1 and 2, no equipment will be required and only the first crew skill will be functional, with the number of available slots gradually increasing as players move up the tier. Finally, at tiers 5 and above, we'll keep things as they are.
3. After implementing proposals 1 and 2 above, we'll abolish matchmaking that considers win rate. (However, it may be appropriate to keep it if there is a significant difference in win rate.)
4. For players who simply want to use famous vehicles, keep them as they are in co-op battles.

We expect that these proposals will encourage beginners to play at lower tiers and advanced players to play at higher tiers, resulting in a more balanced game.

[fleet elm]

That’s the move. Get to 4x or 5x in CW and switch modes

[near pond]

[autumn light]

To be somewhat awkward, it has nothing to do with your base assumption as your base assumption is largely wrong. By far the vast majority of people who buy any premium vehicle (or even the SP) are existing players with 4 or 5 digit battle counts.

The reason is pretty simple, for years people have complained that once in game sometimes a team would get an "unfair" distribution of player abilities from within the 30 players in the game. I would put money that some form of this complaint has been said by most of the people commenting in this thread tbh, whether it was "my team was full of potatoes" or "well, they got X number of people from those comp clans we've already lost" and so on. It was also blamed for "steamrolls" and for shorter game times, both of which improved with the initial change and further info has yet to be released on the results of the tweak.

Previously a system based on battles played was used, but the complaints continued and so win rate was added to the calculation (and then tweaked post release to address some feedback and issues). The change now means that that distribution is much tighter than before, so while you can still have times where either side has more "higher win rate" players, there is almost no extremes and the "gap" between teams will be much smaller on average than before.

[ocean reef]

Make the games random not skill or winrate based or any other metric I am fine with ensuring that the “types” of vehicles is matched. Otherwise you punish players with high winrates.

Anytime you get a response from the game or a moderator that takes 3+ minutes to read it means you are getting screwed.

[coral zephyr]

All that is fine unless you're the guy that is getting more potatoes on their team more often now. That is pretty much what it comes down to.

The math works out to:

🥔 + 🥔 = 😕
🥔 + 🥔 + 🥔 + 🥔 = 😩

While 😕 isn't great it is certainly better than 😩

We used to have on average:
😕 + 😃 = 🙂

Now some better players are getting:
😕 + 😩 = 🙁

I didn't want to type a bunch of words. So mixed it up with pictures.

[fleet elm]

It’s an unfun system for us. No amount of tweaks will change that. Significant exp penalties, string theory sucks now, no correction for high battle count tanks from before the patch, less fun gameplay.

Only upside I’ve seen is that it’s less daunting to see a 3 man super uni platoon on red team. I know my team probably has a lot of the rest of the good players so I can rely on them

[ocean reef]

The goal of matchmaker is to artificially give everyone a 50% winrate. If I get punished for winning then what is the point of playing this game?

[coral zephyr]

Just please don't apply this MM to CW. Let us have CW random at least.

Of course that begs the question why it isn't in both WW2 and CW ?

Just tested and implemented in WW2 first? CW coming?

Only perceived as needed in WW2 and not in CW?

[ocean reef]

It would be like saying a top sports team cannot play another team in the same league because the other teams winrate is too low. Too much focus on being equitable.

I will continue not to purchase product from the company. I used to buy multiple chests every week. Have not purchased any all season. It has now become a good habit.

[fleet elm]

Different mm, has to be applied differently. Our mm was a small change to an existing system

[fleet elm]

A different reward incentive or progression incentive would be nice

[autumn light]

didn't say it was good, bad or indifferent. Simply replied to someone saying why they thought it happened with why it did happen + whether any of those things/aims were achieved and then some anecdotal evidence that it was indeed a solution offered to a problem people brought up/complained about 🙂

[autumn light]

CW uses Era's, WW2 using +-2 tiers. There is also probably some stuff to do with the pairing system in WW2 that idk if is in CW - would need to check in with SPG for details

[coral zephyr]

Yet we have the same players in both modes. Is CW going to be converted too?

Just wanting to plan ahead.

[autumn light]

they are fundamentally different MM systems to my understanding, as Nate said the change to the WW2 was a small change to an existing system that was within it's existing system.

[coral zephyr]

I am familiar with the different parameters of the MM.

The question remains is win rate coming to the CW MM.

For some at least, the cure is worse than the disease for the WW2 MM.

[haughty terrace]

The question sir is who all is complaining? If the vast majority of players don't care or are more happy with the change then its here to stay. We will never truly know the answer. My guess is the majority are happy or simply indifferent. I have seen players post they like the change and games are more fun for them. To be honest, I don't think wg worries about unicums threatening to leave the game. The majority of the money comes from the large player base, so they probably look at it from a financial point as well. Just my opinion.

[coral zephyr]

You could very well be correct. However, my counter point to good players being few and far between is: if good players are so rare and few that they don't have to worry about losing them, why even bother balancing them with bad MM?

Either they are a large enough percentage of players to feel the need to "balance" or small enough in number to not worry about shedding them. Good players are competitive by nature. That is why they are top players. One can't expect to put your thumb on the scale and have them stick around forever.

[haughty terrace]

I believe a lot of players wanted better mm. It's a question of what you consider better. Longer games, less steam rolls or even 1/-1 games. Again, its hard to tell since wg does not share much with us. I suspect the majority want longer games, less steam rolls and better balance. They are less concerned with win rate in my opinion.

[grand garden]

Thank you for your reply. I'm actually using a translation tool to translate my English, so I apologize if I'm rude or misunderstand the dialogue.
A-. When I mentioned the distribution of my premium tanks, I wasn't referring to the sale of premium tanks for gold, but rather to the free distribution of premium tanks through events, etc.
In the first place, if the game easily distributes premium tanks later on, isn't it natural for players to hesitate to purchase them?

B-. I'm very grateful that the management is currently experimenting with improvements. However, my suggestion was that team composition should not be based on win rate, but instead, that individual performance should be emphasized when calculating silver and experience points.

C-. I've come up with a new idea, so I'd be happy to hear other players' opinions. I apologize if this idea has already been suggested.
How about evaluating the strength of both teams based on the abilities of the crew and the characteristics of the vehicle, rather than the characteristics of the player, and aiming for equality between the two teams?
For example:
C-1. The number of skills selected by the crew. The more skills a player has, the stronger they are considered to be. This refers not to the number of crew skill slots, but simply the number of skills they have selected.
C-3. Whether the player is at the top or bottom tier. The higher the tier in the match, the stronger the player is considered to be.
C-3. Map-specific adjustments. For example, on a wide map with a good view, a highly concealed vehicle with enhanced concealment will be considered to be more powerful.

However, no matter what improvements are made, players who don't make an effort to fulfill their role, such as charging into a group of enemies and dying a minute or so after the start of the match, will widen the gap between teams, so I think we need a system that makes it difficult for players who don't earn any score to stay in high tiers.

[coral zephyr]

"However, no matter what improvements are made, players who don't make an effort to fulfill their role, such as charging into a group of enemies and dying a minute or so after the start of the match, will widen the gap between teams, so I think we need a system that makes it difficult for players who don't earn any score to stay in high tiers. " @grand garden
I agree with you. Those poor players should be encouraged to play lower tiers. Not forced to play lower tiers, but encouraged more than they are now.

[lament mica]

There will always be people losing—can’t have winning otherwise. Essentially you all are asking for a funky knockout tournament and then will wonder why your win rates plummet or you have to play bots as all the “bad” players leave.

[grand garden]

Perhaps WG wants to fulfill the wish of new players to "try using famous tanks" as soon as possible.
So, why not hold regular events where standard vehicles are loaned out for co-op, with a message like, "This month, you can use the Tiger I and IS (or other famous tanks) in co-op battles only, even if you don't own them." They don't have to be premium tanks. It's also important that they can only be loaned out, not owned.
The expected benefits of this proposal are: "It will attract gamers who are interested in tanks but have never played them," "It will satisfy the desire of players who just want to use tanks, not play against others," "It will separate people who just want to operate tanks from those who want to play against others, which is expected to improve the quality of matches," and "It will also get existing players interested in vehicle trees that they don't own."

[scarlet mantle]

This is about the current matchmaker and not this lame co op mode which u try making attractive to play mate. Matchmaker should only match tank class. Any form of sbmm i don’t want.

[fleet gyro]

You Really feel the snap when the MM decides it wants to rebalance you. I think if the MM was better at sharing out the stupids in first place we would get to a better place. WR imo is a poor indecator to be taken solo. It hurts me as a person with good win rate

[lament mica]

It’s not taken solo.

[ocean reef]

The matchmaker is focused on being equitable. Wargaming cannot have anyone feeling like they won or lost. 50% winrate for everyone is the endgame.

Participation ribbons will be issued out at the end of the match. Wargaming will be redistributing wins from high winrates players to low winrate players in order to create an inclusive environment.

Eventually winrates will be removed and the scoreboard will not have any stats but say “good job comrade, now go to the store and buy something”. This way everyone feels good about themselves and wargaming gets some money.

Wargaming said your winrate is too high and it makes people feel uncomfortable. We’ll tweak matchmaker and “fix” that high winrate. Does 50% sound “fair”?

[ocean reef]

I am still on my streak of not purchasing anything this season!!!! Used to buy multiple chests every week. It’s now a habit. When you implement terrible changes to the matchmaker and the game I will refuse to purchase the product.

[thorny bramble]

Aye send me gold so i can get some chests

[scarlet mantle]

Hopefully you doing good mate. Probably make an request on go fund me

[thorny bramble]

Aye help me out mate

Send me a quick 3000 gold twin

[cunning dew]

Get outa here homeless bum

😅

[ocean reef]

Your next free daily chest is on me.

[upper rain]

Lmao this is so funny

[ocean reef]

More terrible matchmaking. 3687 WN8 today. Winrate of 38%. Top 3 majority of games. I will continue not to purchase anything until the matchmaker is reverted to the way before all the changes. Used to play almost triple the matches on the weekend. Not anymore. The developers chose to punish players with high winrates.

[carmine nebula]

[ocean reef]

I am sure there is some wargaming developer implemented this matchmaking because his winrate was below 50%. Good job ruining a product that required no changes. Wonder what will happen to sales after the. Nights talker event is over?

[near pond]

Deserved for playing the nightstalker

[ocean reef]

You’rewelcome. It’s why the game is free to play.

[near pond]

Yeah…….no

[tawdry sky]

Dude is back to ranting about DEI again lmao

[near pond]

There’s games that are ENTIRELY ftp with no DLC that have been around just as long as WoTc and are performing better.

The whole “wot needs op stuffs to keep it ftp” argument is bogus. Plus that thing was in an earn op

[ocean reef]

Already started playing other games. I play about 50-60% less than before the updates.

[near pond]

[ocean reef]

Done for the night. I think I will have a better time plucking my nose hairs than playing WOTC.

[near pond]

Disgusting but have fun with that ig

[ocean reef]

Thank you

[thorny bramble]

@ocean reef One way to help at least in my experience is platoon with other good players

[rotund veldt]

The interests of wargaming and the players are not aligned. War gaming make money by selling tanks, boosters and commanders to players who think spending money will make them better. Looking at screenshots of players who have played a few years and want +1 -1 they have enough gold, silver and xp to never buy anything else from wargaming.

[cunning dew]

Its the only way to fix it. If you do that actually can have a higher winrate. However you have to be platooned

[scarlet mantle]

What a sweat fest this kind of mm is. Not fun at all if you play solo.

[coral zephyr]

Since I have been mostly solo lately, I am just not playing very much.

[near lichen]

I know there's alot involved with the match making algorithm but explain the light and heavy match ups for this game seems like one team was stacked with better tanks and was most likely going to win

[lucid swift]

Imo dont play cw 😂😅

[vale maple]

Why do I always get the shittest heavy players on my team. Is this a MM thing I think so I have stream proof of my heavys failing in 90% of my games since mm change

[lament mica]

Why confuse players and put this in a WW2 MM thread?

Pretty sure they have said platoons cause issues across the board and there is a lively platoon thread if you want to raise it there.

[scarlet mantle]

3 games all massive losses.

I just switch the game and keep my coins. No black Friday deal for me with this kind of mm.

[coral zephyr]

Who is this new MM helping?

and

How is it helping them?

then

If it is helping someone, who is it hurting?

[fleet elm]

Literally half the playerbase is gonna benefit by being carried more often. 40% of the playerbase above 50% wr but below 53% won’t really notice much change. 10% get shafted and have noticeably worse mm. But most importantly (at least WG hopes) in the matches with no outliers, the match will be more competitive and less blowouts

[coral zephyr]

How are they going to get carried, if the MM stackes enough bad players with one really good player?

The complaints by really good players seems to be that they are getting too bad players. So many bad players that their teams are impossible to carry. (There is a limit to how many players one can carry. Especially when they don't even shoot, for example)

[unreal hound]

People are assuming they are the only good player in world of tanks most of the time players have another great player on the enemy side which makes that even and both teams get equally bad players to fill that up.
But of course you cant predict what they will do

[coral zephyr]

The effect of the new mm does not seem to allow good players to carry mid players. It is to saddle good players with poor players in order to suppress the good players.

If mid players are getting more wins, it is at the expense of good AND bad players.

[unreal hound]

i still get a solid 60% W/R with 90 % solo so it cant be that bad but people always will remember the 1 bad 0:15 round in 10 Games.

[bleak oriole]

But most importantly (at least WG hopes) in the matches with no outliers, the match will be more competitive and less blowouts

The amount of outlier matches is much much diminished compared to 9.4 - the only reason that there are still some is Platoons (tm)

[coral zephyr]

Before we could say it was bad luck. Now we never know if we got balanced by the MM. That is significant.

But worse MM is still worse. It is likely worse for bad players too. As they get grouped enough to drag down good players. But maybe bad players don't notice or care.

[coral zephyr]

So I could mitigate the effects by platooning with another good player? Unfortunately, that has not been available lately.

Really I don't epect anyone at WG to tell me that I am correct. If there was not 100% confidence in the new MM, it would not have been implemented.

Although it has been said that the negative effects on higher win rate players has been softened. But it has not been said exactly how. Yet the MM still negatively impacts higher win rate players. That doesn't seem conducive to them sticking around.

I can only guess that the "quality of life" initiative was to improve player engagement. Then WOTMA would be better off with more player engagement. But the MM currently picks winners and losers. or at least sorts them.

I can only tell you in is not good being on the short end of the stick. My engagement is going down. If I didn't want it to go down, I would not be here giving feedback.

There must be a way to get outlier matches to an acceptable level without discouraging good players. Outlier matches are likely due to bad players getting farmed quickly. Seems like it would be better to work on that problem. --They get farmed by mid and good players alike when they just sit there in the open.

Again if the good players are so few to not worry about losing them from the game, they are also too few to bother with slanting the MM against them. If the good players are numerous enough to feel a need to "balance them" and give them worse MM, they are too numerous to risk losing their support of the game.

[fleet elm]

By outliers I meant players of significant skill not balance of teams. Has there been a significant change in match length (prior to the equipment/crew changes) and blowouts?

[near pond]

Tbf short matches cant realistically be eliminated without some sort of objective only gamemode with respawns and a lower team size. Once a disparity in remaining team members develops there’s very little that can be done to stop a blowout

[fleet elm]

Blowouts are different than competitive matches are different than down to the wire matches. A competitive match might end 9-0 because one team got just enough of an edge to get map control and 6 tanks are one shot

[lament mica]

I’ve noticed this a number of times. I certainly try to direct my team to take out good players.

Also, I think people forget that some players are up and coming and others are diminishing. Some of the people in this thread had some terrible stats before the lightbulb clicked and now play well. Some of those “bad” players being sorted to their team will likely over perform old stats as well.

And truth is we don’t really know what percent of games actually disfavor a high win player (even if WG published an overall percent from data) simply because of the various factors such as queue order sorting with various queue sizes and error rate in considering win rate from actual capability.

However, people lose a match, feel like the team performed poorly and now the mm is the excuse.

[fleet elm]

As a good player, statistically, you're gonna have worse teammates than the other team something like 90-95% of the time with the new matchmaker whereas before it was 50%

[lament mica]

Sure, show your work.

[bleak oriole]

As a reminder, the system looks at your vehicle win rate is the primary balancing stat. If you don't have a statisically meaningful number of battles in the vehicle, then your account win rate comes into play.

[fleet elm]

I mean the burden of proof is sorta on you here. The stated function of the system is to make the teams have closer to equal win chances. The way to do that when a >60% win rate player is in a lobby is to saddle them with worse teammates who are more likely to lose.

[errant saddle]

Kinda ruins it for me when I run the Sturm now since the reload is basically 30 seconds and having loads of field maps compared to city...

[fleet elm]

Reload change is literally completely irrelevant

Everything got the exact same change. The sturm arguably benefitted from the reload changes if anything

[lament mica]

What are you talking about? Perhaps you should reread. I was noting issues that would likely undermine making the teams equal—AND we don’t know—WG doesn’t even likely know. You proffered it doesn’t matter because stats. Show your work.

[fleet elm]

I mean I could send my matlab sim but it’s a simple logic path to follow. The 14 other people on the team of a very good player are going to be worse players than those on the red team

[lament mica]

That’s like saying that using camo on a tank is irrelevant to the tank or speed equipment is irrelevant to the tank. The affect on reload, or affect on camo, or affect on speed is very relevant to a tank. And very relevant to the player and their enjoyment in said tank.

[fleet elm]

Everything got the same dpm change (except tanks with intraclip reloads)

[lament mica]

Sure send it.

Because that last statement just isn’t true—particularly as the absolute you offer it as.

It’s one thing to complain about it but at least be somewhat accurate instead of just pushing an agenda.

[fleet elm]

Explain why I’m wrong

By what other mechanism are winrates of very good players dropping?

What else causes that?

[lament mica]

Then instead of dropping dpm, they could have just increased hitpoints by 10 percent across the board then, right? Would affect everyone the same right? Just increase/decrease the accuracy of everyone by the same percent right—doesn’t really affect anything between tanks right?

[fleet elm]

Yes exactly, nothing really changes tank to tank with dpm changes, hp increases would be identical. Accuracy is a bit different with the max render range. Poor accuracy tanks suffer a bit more

[lament mica]

Because it’s likely good players are on both teams so you don’t always have worse players—particularly as win rate is not absolute determinative.

[fleet elm]

You’re saying it’s likely that an even number of top 1% skill players get in any given lobby of 30?

There’s a ~25% of getting even 1 top 1% player in any given lobby

[lament mica]

Dropping can be caused by age, diminishing skill, other players being under estimated, or even if they were overinflated to begin with. Some of the things I pointed out previously actually.

[fleet elm]

lol it happened to every good player exactly when WG made the mm changes

[lament mica]

But it wouldn’t—you could do the same with alpha—it affects the psychology of the game. It’s not just 1-1.

[fleet elm]

The game yes. Individual tanks, no

[lament mica]

He was talking about the tank in the ecosphere of the game/-not some theoretical comparison of dpm.

[lament mica]

Every? Or just the ones you heard from? Not saying the initial mm change did not affect players, but that is a different discussion—far away from my post. You are bootstrapping your complaints here. Wintates did not drop because they always have worst teams. You said 90-95 percent of the time then went absolute. How about you provide proof for your statement that 90 percent of the time your team is worse just because of mm. That is what I asked about.

[fleet elm]

It can’t not happen unless you platoon every battle. If you had a 60% winrate and played a significant amount of battles in your normal tanks, your winrate went down overnight

[fleet elm]

There’s 2 scenarios where you don’t get a worse teammates. The 1/15 where you’re last in the queue so teams are equal already or there’s another 0.1%er in the lobby. Case A is easy to figure %, case B is harder but sufficiently rare enough that 90-95% is a good enough estimate. In either case, it’s a bit of hyperbole mixed in too, the statement mainly being made to express that you get shafted on teammates really hard as a good player

[thick cosmos]

Another issue i havent seen brought up is that if team A is all 50% win rate players and team B is a mix of high-low,
Low win players die first leaving highs to 2v1 resulting in a loss for team B.
This drops ALL win rates not just the good players.
Meaning that your now gonna begin to see people likely drop into sub 40%
When that happens this will make the MM issues even worse as it will be dealing with extreme lows

Already struggling players are now likely to be locked in a cycle of loosing teams

Dragging high win rate players down with them.

[fleet elm]

Not quite accurate. That type of distribution is rare. Good players win less but still well over 50%

Bad players will still see an increase in winrate. Bad players tend to camp and not do much rather than die early for nothing

[thick cosmos]

Hmm fair enough.
My win rate is low from when I was a trash can years ago. I was beginning to make it up slowly but my win rate over the last three days has been 25%

Im struggling to find how thats possible

Even intentionally trying to sabotage my team my wins shouldn't be that low

[thick cosmos]

Could you provide some more info on a typical distribution?

[fleet elm]

Pairs of tanks get split between teams. Pairs are as likely to be avg/avg as avg/bad or bad/bad. So the good and bad vs average scenario will end up being rare

[fleet elm]

That’s just bad luck possibly combined with bad play. But mostly just a poor streak. Under 100-200 battles won’t tell you much of anything statistically

[lament mica]

Ah so you don’t have any real math just your guesstimate.

[fleet elm]

There’s no real math that we can do. We don’t have enough specific information. We can assume stuff and sim but we have no way of knowing accuracy. But I also don’t think I’m that far off either

[grand garden]

Before adjusting the matchmaker, was there any discussion on any of the Discord forums about how to improve it?
In the first place, I don't think the issue of large disparities in matches should have been solved by adjusting the matchmaker, so was there no opinion like that beforehand?

[grand garden]

I apologize if I misinterpreted, but based on this, wouldn't players with accounts that have a high win rate be at a significant disadvantage if they start with the initial package of a new tank?

[near pond]

It wasnt the matchmaker that was adjusted

[fleet gyro]

An older Dev Said once. "A game should never cheat the player or make them feel cheated." The GOAT that is Jeff Minter. This every 15 matches just decides, you losing that one. Just so the 80% can feel better more of the time?

This mechcanic is dreadful, it is cheating the 20% to make the 80% not feel like their bad.

All it does is reinforce the Dunning Krieger effect. Something WoT has a problem with already

[fleet gyro]

and secondly, these wins are stupid 1 sided. So nothing is worked for, nothing is learnt

All it does is treat the symtom, not the actual problem

[lament mica]

They have been tweaking the mm since 2016 or so iirc. Reducing the effect of various player complaints to give an appearance of better balance. Things such as having the same amount of tier on each side so one side is not loaded with tier 10s and the other is half tier 8s. Matching tank type, reducing scout spread, adding games played as a measure, etc etc

[grand garden]

Thank you for your reply.
I suspect that the cause of various problems, including this one, is that WG makes corrections or additions before players give their opinions. In fact, WG makes frequent mistakes and players post a lot of complaints.
I think that if there was a situation or thread to collect player opinions before WG makes a correction, rather than after, complaints would be less likely to arise.
For example, WG would start a thread to address the problem they want to solve, such as "we want to reduce games with large wins," and then collect solutions from players to use as reference. This is not the same as starting a thread after making a correction.
Alternatively, if such a thread already exists, I would appreciate it if you could let me know.

[lament mica]

Part of the issue is the playerbase has various opinions and there is no “right” answer. It would be nice if WG engages more before, during, and after implementation.

The equipment/skill and tank adjustments had some engagement but WG drops the ball of continuing through and after final implementation to continue to get feedback and adjust. What is worse is they make adjustments now based solely on perception (sometimes ancient perceptions in contradiction to current reality), not use any data, and don’t seem to care about actual outcome.

[bleak oriole]

I would hardly consider this thread as having a "lack of engagement from WG" in it - both BFG and myself have been fairly active in it.

[lament mica]

You are correct. The engagement in this thread has been above and beyond standard. Sorry I was not clear in referring to tank and game mechanic changes. And I’m not saying every tank change needs so much either, but nerfing premium—paid for content—should be extraordinary, exceptional occasions where there is repeated, ongoing engagement—not appearance tokenism. Major continuous engagement on major mechanic/equipment/ skill changes as well, which were slightly better than the premium tank nerfs. The equipment / skill changes were at least over a substantially longer period of time whereas the premium paid for tank content nerfs were mere flashes where it appears the decisions were already made and in regards to the micro Maus, the feedback was not even listened to. 5/4 showed similar disregard to the community.

And not to beat you up in any way because I know you don’t cover all areas and are here because of your specialty in this facet but over indulgence in response here and scarcity elsewhere gives the appearance of bias to portions of the playerbase. Not saying to engage less, but other areas engagement needs to be more like this.

[bleak oriole]

For engagement - I typically try to say away from these feedback threads as they meant for you all to provide feedback. I do read quite a few of them - but not all, because there aren't enough hours in the day to do that - to look for things that I might be able to chime in.

A lot of these threads are "why don't you do X", "why do you do X", and "It would be cool if you did X" type of threads. There is no real input that I can provide for those because they entirely fall within the "game design" areas of responsibility.

Where I can provide context is "how does X work?", "X doesn't work the way that you all think it does", and anything that generally falls under the unbrella of "how". I can't do that for every thread - as above, not enough hours - but also there's a lot of repeated questions.

For some of the other stuff covered above - I'd like to point out that for the tank nerfs in particular, players have provided feedback for years of feedback for some of those tanks. The 5/4 has been in the game - and been a problem for a long time. We aren't going to disreguard the volume of feedback we've had simply to focus only on what was said in a particular thread.

[lament mica]

Why have a thread if just going to disregard it? That’s why I complained of tolkenism.

As far as years of feedback—that’s not really terribly valid because there have also been years of changes as well. Essentially you are saying that those opinions from year one are more valid than opinions in year 11. —Things not actually based in reality. I understand addressing perceptions, but there is a limit imho. I could go on but have digressed enough from mm thread. I appreciate the discussion and time offered though.

[fleet elm]

They haven't disregarded it. They've seen it and concluded that our experience playing the game matters less than that of average skilled players since they make up the bulk of the profit

[autumn light]

As nate slightly said, not engaging doesn't mean not taking it into account. Dev's and staff can read something and make changes and you'll never know because they took it into account in their own workflow/process.

In terms of the years of feedback thing, i think you may slightly misunderstood, i believe SPGs point was that a lot of recent changes have been focuses on feedback that has been a constant for years. These MM changes are an example of that. So it isn't about some feedback being from X years ago and we're "just getting to it" it's that it was complained about X years ago, and is still complained about in the same or similar way now. Many of those things we have made changes too here and there, and that will always be the case in the future too.

[native plaza]

Honestly at this point changes to mm won't make that much more of a difference, the best/biggest changes would be to limit tds and lights and even make arty 1 per team

[bleak oriole]

My $0.05 is that the above belong in a separate suggestion - although, tbh, the various versions I've seen of limit class X to Y - have never gotten much traction from the playerbase.

Edit: Pennies are no longer being made, so min charge is now a nickle

[lament mica]

Ok but…You can actually see the threads that say otherwise—particularly the Micro Maus there. So as not to clutter this thread more, I would be happy to continue in that thread or if prefer, I can make a different thread.

[paper prism]

I do agree with OP proposal but with small yet important addition - aside WR, system should also take tank tier, tank class and “tank class role” into an account. Makes zero sense to balance equal wr players using tanks with different roles. Say you dont match a brawler ht vs support td. Not even speaking about different tiers… Id say even support ht vs brawler ht of same tier is NOT right. Each tank should have its specific role set (as PC version) prior. The MM should match players by mentioned credentials. Matching time may increase though.

[unreal hound]

the matchmaking would be 1 game per hour with all those restrictions we dont have the playerbase to include tank roles into the MM, even without the addition of Winrate.

[fleet elm]

It does help with the current issues of getting scheduled loses based on map. With the new mm, if I get a type 5 on Death Valley or Mali, it’s basically an auto loss for my team.

[coral zephyr]

That loss makes no difference to the next match. It isn't giving you bad teams a percentage of the time. Every match it trying to give you a bad team, if you are really good. The mm is only successful at giving you a bad a certain percentage of the time. Varying from session to sesion.

[fleet elm]

You've got a ~90% chance of having a worse team than the red team (minus you). That 90% will vary from sorta close to much worse. On average though you're gonna get the worse end of 2-3 extra pairs. In those cases on bad maps for tanks, you literally can't impact the match before the odds become overwhelming

[lament mica]

So you are jumping back on the 90 percent again eh?

[fleet elm]

Again? I've always been there...

[grand garden]

I have a question for WG: have you already recognized that taking win rate into consideration when matching is a problem? Have you removed the consideration of win rate, or is it now your policy to remove it?
Also, I have another suggestion: while the usual procedure is to "select one tank and start waiting for matching," wouldn't it be good to allow players to "select multiple tanks and start waiting for matching" if they wish? Of course, as before, players will participate in the match with one of the tanks. MM will flexibly arrange matches by matching those multiple tanks with other players waiting to join.

[lament mica]

Really? Then please show your work.

[lament mica]

Part of the problem is they take anything into consideration. It’s been that way for years due to population. Once you start down that path it is inevitable win rate comes in. Either you let players kill off newer players or you try not to.

[fleet elm]

Why bother? Youre arguing for the sake of it, not in good faith. We both know the lack of info from WG means you can choose to refute any number. Give me your number and how you arrived there and maybe this conversation will be worth continuing

[coral zephyr]

I see the 90% mentioned as the odds. When one considers the amount of 65% win rate player vs everyone else. 90% is likely a good estimate of the odds.

It wasn't like he said 91.02%. It was a round number estimate.

[proud sluice]

Am I supposed to enjoying losing? youve taken away my agency and bake in losses. before this change I was 63% wins and climbing and now im down 2% and dropping. 2 platoon mates used to play 3-6 hours a day now we barely play 3 hours a week. before this change we saw maybe 1 of these players. now we have an entire team of them vs the other team that only has 5-6. I DONT WANT to play these players that dont know how to use a controller.

[cunning dew]

Mine went from 70% to 60%😭

[proud sluice]

its really tragic to hurt the players that play the game to get wins instead of just zombie out or zombie back to get marks. 🙁

[coral zephyr]

This new MM has completely destroyed WW2 for me. I just have zero motivation to play it.

Why would I when the mm will actively try to put the worst players on my team?

Was that loss because of just bad luck or the MM? Could be both.

This is quite unfortunate. Previously I was one who really liked the game.

[lament mica]

You are the one who proffered and continues to push a very specific narrative. If you don’t have any math or sufficient reasoning for it, that isn’t for me to create to refute.

[lament mica]

Good estimate based on what exactly? Because it sounds good to you?

[fleet elm]

Bad estimate based on what? Because it sounds bad to you?

[lament mica]

I didn’t say bad—I asked for your evidence.

[fleet elm]

I answered you. You didn’t like the answer

[fleet elm]

I also asked for your evidence/number

[lament mica]

You must be a dupe of max miner—the way you talk in circles contradicting yourself. Maybe not a dupe—maybe a cousin.

[fleet elm]

It’s an estimate, not sure what more you want. It’s in the definition

[coral zephyr]

based on looking at player distribution graphs. How much area is under the curve below 60%? What are the chances a player with 60% will be paired against a player with a worse win rate?

This example is the tier 8 SU-130PM which players average a 49% win rate. Will a 60% paired against a higher win rate person or even equal win rate person? not very likely. Odds are it will be someone worse and closer to that 49% average.

This creates the WR delta that the next pair(s) will try to balance. The delta between 60 and 49 is +11. That is a lot to balance. It will likely take more than one pair. [but they made a change. reported as a change but not what exactly was changed. I think they put a cap on the delta. So the delta won't be +11. It will be the cap.]

Yeah. so the odds a 60%+ player is going to be saddled with more bad players is pretty good. 65% player, even more so. Some would even say 90% chance.

[lament mica]

I didn’t proffer a number. I said it was hard to come for a number which was initially took umbrage with then sent you in circles stating you cannot come with up means of determining a number and at the same time pushing a number.

Entertaining the testing acrobatics you are trying none the less. But please continue pushing your number and just expecting everyone to accept it.

[fleet elm]

Yeah you don’t like what I used to provide my estimate. Experience plus 100k simulations with whatever info WG gave us

[lament mica]

You seem to completely ignore queue order or the number of other items I noted that would cause errors in team assignment as far a good vs bad players.

I’ll stop there for now.

[lament mica]

You didn’t provide anything but your own good word, sooo

No

[fleet elm]

Nah dude. It’s nda stuff from WG. I can’t just release it

[fleet elm]

That’s why WG isolates new players for a bit with mostly bots. It’s why tier 1-4 don’t have MOE. It’s a big flaw in CW that era 1 has MOEs

[lament mica]

Sure sure. So you’ve been talking about NDA stuff. I’m sure WG would appreciate that.

[lament mica]

Seems like most people that play CW don’t care about MoEs. Actually that’s true of 95 percent of the playerbase in WW2 as well. One tail that certainly should never wag the dog.

[fleet elm]

The “most players” aren’t the issue, the sweats are

[fleet elm]

Not against the nda in this case

[lament mica]

Talking about the NDA and bragging about it even.

[fleet elm]

What does that matter?

[lament mica]

It’s NDA. You keeping talking about it. Makes about as much sense as your 90 percent in your mind I suppose. But sure keep talking about it.

[fleet elm]

🤷‍♂️

[proper valley]

Can you blame them?

Look at a score board and tell me you aren't surprised by just how low some of these lads score-

[lament mica]

Well there is RNG. We’ve all had the fun of shooting someone with 99 percent chance of pen, or low rolling what should be kill shot and then getting lag shot in the most obscure place that 99 percent of the time would not pen, sets you on fire, destroys ammo rack. You patch, and next shot blows your ammo rack anyway. Stack those plus bad play and see some terrible games.

[proper valley]

Yea. Thats true. Thank you for the reminded brother.

[autumn light]

The delta between 60 and 49 is +11. That is a lot to balance.

I have alluded to this a few times now, but there is a repeated assumption that there is a huge delta that must be overcome in these "extreme variance" pair situation, but this is not the case. Part of the tweaks was to address exactly this concern. Without knowing the exact brackets and system obviously you can't know for sure, but what i can say is that the delta is not as extreme as seems to be being repeatedly assumed. It's not quite the system Nate at one point was suggesting (only takes into account the previous pair) but that would be a more accurate assumption than some of the ones being made here.

And for the back and forth on the odds of meeting certain levels, can i suggest you stop using the win rate and start using a percentile? AKA if you are in the top 5% percentile what are the odds you come across someone in the bottom 5/10/15/etc percentile. Doing it this way you don't need our numbers/data, you can make very accurate calculations that will give an accurate approximation

[coral zephyr]

As you see in in my post I reference that:

"[but they made a change. reported as a change but not what exactly was changed. I think they put a cap on the delta. So the delta won't be +11. It will be the cap.]"

But we don't know how the sausage is made. So every time we get heartburn, we blame the recipe.

I was looking at the player distribution graphs. Of course the ones I have access to are limited. But they give a basic picture of how players a distributed according to WR. Looking at them one can get a feel for the odds of being paired with a lower, same or higher WR player.

[autumn light]

i should probs have not tagged you TBF 😄 it was more a generic statement after reading through a bunch of messages, but your message was just the most recent and relevant with something i could quote 😅.

You did acknowledge the change but then it somewhat does take away from the point you were making IMO, there isn't a cap but things like brackets, values, etc were all changed to specifically address the complaints raised here while also aiming to further improve the reductions in steamrolls and increased match times we had seen. Because although as you say you don't know "how the sausage is made", to keep within that analogy, basically we have said "these do not contain lamb" but there seems to be an ongoing discussion about how much/if they contain lamb, and how much the lamb may be influencing the taste. So as much as we haven't told you the exact ingredients, if we have told you it doesn't contain lamb then a discussion about how much the lamb is influencing the taste seems a tad superfluous surely? If you get what i mean. Not a perfect analogy but hopefully not awful 😄

[autumn light]

the latter part was 100% not aimed at what you said FWIW, it was more reading the back and forth between nate and t prime about the odds of seeing a certain win rate delta and there not being a good way to get an accurate guess. Technically true but my point is that if you switch to percentile you can actually do pretty exact (if not 100% exact) math about those odds rather than making estimates. And then from there you can get something more concrete to point at.

[coral zephyr]

It doesn't matter too much. I have pretty given up on WW2 for now. It isn't fun to have the MM purposely placing bad players on my team. There is no limit to how much the MM will do it. It is all a matter of queue distribution and how team forming shakes out. There are also an unlimited number of potatoes to place on my team.

It wasn't fun before when I would get randomly saddled with more of the worst players that drag the team down so much that wins become all but impossible. Now the MM was changed to make that happen more often to me and my favorite tanks.

I like light tanks because I could use strategy to get significantly more wins than the global average of ~49%. So, the MM will hate me for it.

[fleet elm]

The odds of having the worse team are relatively straightforward for someone above the max skill cap. Last in queue or finding someone else above the max cap. One of those is simple, the other is guesswork

[uneven shoal]

I do feel like the overall experience playing solo has gotten worse. I’ve taken a few screenshots of outlier matches where significant portions of one team do little to nothing each round. I haven’t had time to run them through my program, but I’ll try to get the screenshots sent here. It’s been over several weeks, but I’m obviously focusing on the worse experiences as they are what stand out

Here’s a few over the past two months. Would need to dig further for others. Obviously only a small portion of the total number of matches I’ve played in the past two months, but some of them are pretty egregious outliers

[uneven shoal]

Accidentally included two screenshots twice

When I’m looking at team/personal performance, what looks worst to me is when you do less than your HP in damage. Some of these games that was me, but you can see in many of these cases that 2-3 players carried each time, and the match was determined by which team had more players carrying or less players performing poorly

The games that stand out to me the most are those when a small portion of the team performs decently, but the rest struggles to do the basics.

In my experience, this occurs more often on maps like Cliff, Erlinburg, or Vineyards where teams are given a spot that looks good to camp on. When you don’t move and give the opposing team control of 60%+ of the map, you are most likely going to lose.
Map design is obviously a completely different conversation, but it’s a trend I’ve noticed.

I know I’ve mentioned before that it would be nice if the MM attempted to swap which side of the map you get, but I have no idea how difficult that would be to do

[fleet elm]

That’s a deep change. You need twice as many maps to guarantee no b2b maps or some fail safe but it’s unnecessary dev work imo

[uneven shoal]

I don’t mind b2b as much if I’m not getting the same side of the map. Most of the time they play differently enough

[mint brook]

Since the October update and new season, My WN8 is trash. Teams die so fast, or I find myself abandoned when the whole team lemming trains, or I get rushed and team members dont even attempt to shoot the rushers… it’s frustrating. Only thing I can do is only play the best tanks like an E5 or something that has good armor, a fast reload and accuracy and try to make up for a garbage team. (Or like some of the high WN8 players… I could always pad my scores and make it look good, but alas Im stuck grinding that Merc Contract and must play high tiers). Looking back at poor games I see some players never fire their gun, or maybe fire it twice and hit nothing. Why are those people playing tiers 8-10? Oh.. it’s because you keep giving away high tier tanks. Maybe not do that? The problem I see with the match maker is that it’s ultra complicated due to an off balance player skill level, but could be successful if the overall player base skill level was better. But high skilled players are leaving the game and have been for years, so your stuck with trying to keep this business going with a majority of players who have never figured out how to play the game nor seem to care to.
You could work on ways to influence player skill, then the Match maker wont blow a gasket trying to put a balanced team together. Feels like the Computer on the Hart Of Gold trying to make a cup of tea that turns out to be almost but not quite entirely nothing like tea.

[near pond]

WG isn’t going to balance around a player made statistic

[fleet elm]

It's accurate enough to give us a decent glimpse of skill

[near pond]

Eh……kinda…..

[mint brook]

Most likely not, however, Im asking for effort to be spent on helping their customers learn how to be successful with their product at least to a level of learning basic fundamentals. Pushing people into the high tiers before they are ready is a big part of the problem here. Those players who have been around a while and are reasonably effective get punished with being stuck into teams of low skill level and therefore start loosing games more often, or getting wiped out early due to lack of effective team work while they are trying to have an influence on the game. I dont sit back and use my team as a meat shield like some, but actually attempt to lead out and influence the game. But lately finding my self quickly abandoned, or my side of the map is already dead.

[near pond]

It is until it isnt

[fleet elm]

I mean most of us in this thread are just using winrate

[near pond]

Win rate too is a terrible metric

[mint brook]

The picture I put up is irrefutable evidence that something changed when the new season dropped

People keep asking for evidence.. here it is

[fleet elm]

It's the best metric for solo play

[mint brook]

Or.. I just forgot how to play?

[fleet elm]

Why do we need evidence. WG literally said it happened

[near pond]

Thats all well and good. Sooooooo, how do we tell if matches had platoons?

[fleet elm]

Did winrate go up? Platoons. Did it go down? Solo

[near pond]

Wow. That is genuine brainrot.

[fleet elm]

After the mm update, this is literally how it happened

In actuality, it's mostly been for personal tracking. WG can find out what's happening on their back end platoon or not

[proud sluice]

Its "OK" though because these are changes made at the request of the community for YEARS 🤡

[fleet elm]

Matchmaking might be the number 1 complaint in the game. WG just addressed it in a way that benefits 95% of players but shafts unis

[proud sluice]

thats great. glad they hate me. now my platoon runs all nightstalkers to get our wins back.

they really shouldnt cater to the player that literally are just enjoying driving their tanks around. I have sent probably 1k messages in the last year to players when they do things that result in obvious losses. the #1 response is "I dont care, its just a game. wins dont matter" why cater to these bozos? its like 80% of their responses " I dont care"

[fleet elm]

They spend the most money. WG wants them to enjoy the game. Really easy to understand despite the frustrating result

[proud sluice]

you got the data to back that up? or you just flinging crap as usual?

[lament mica]

I think you missed my point as well though at least understood some.

Whether win rate or percentile—it is still at best approximations. Not even sure where you would get percentile but for other approximations—more errors.

It’s not definitive. The game still needs to be played.

#1405193097079689458 message

[fleet elm]

You need data to be convinced that 99% of players spend more than 1% of players? Oof

[acoustic mist]

I'm really not good enough to carry a team. So being in the top 3 of the losing team regularly doesn't make me feel any better about being steamrolled. It's truly not any fun babysitting players who are even worse than I am because the reality is that if I'm on your team & I'm 3rd you did something wrong, if I'm 2nd you made some poor decisions & if I'm 1st you seriously need to rethink how you play the game. I'm old, disabled & really bad at the game I should never ever be in the top 3 ever...

[errant saddle]

Reminds me of the light ebr tier 9 I had on prok...sitting next to tds this previous week...

[uneven shoal]

Here’s an example from Death Valley.

The top 4 players from team 1 completely outperformed the rest of their team. And 2 of those 4 performed notably above the others. This group of 2-4 clearly carried their team.

Then we look at team 2. 4-ish players clearly matched the 3rd and 4th players from team 1, and overall the rest of team 2 (excluding top 5 in xp) performed better than than their counterparts on team 1. Unfortunately for team 2, there were a pair of outliers on team 1 who were able to dominate the game.

It’s also worth noting what tanks under/over performed. On team 1 the top 4 were 3 tens and 1 nine, team 2 was similar. However, over the rest of the board team 1’s tens seemed to do a little better than team 2, which may have contributed. Tier ten tanks typically have the strongest presence on the field and if they don’t shine, it really hurts the team.

Yes this match was overall balanced, but it clearly came down to the top 4-5 players on each team, which I’d argue is not the best way to balance things

Do I have a solution for this? No, but I think it’s worth looking at as I think it clearly points to the outlier issue and just how much those players can influence the match

[coral zephyr]

I think the solution is just go back to random. It was better. You then randomly get bad teams, average and good. It all equalled out over time.

[uneven shoal]

Next match I played.

Once again, looking at outliers in damage, team 1 had 2-3 players that stand out, but looking at team 2, there’s only 2 players that performed well in dealing damage. However, team 1’s players are two tier 10s and a tier 9, while team 2’s is a tier 10 and a tier 8.

Tier for tier isn’t as clear in this match, but team 1’s nines and tier 8 TDs performed better than team 2. At the lower end of the board, team 1 actually had a stronger concentration of top tier vehicles, but the tier 10 and 9s from team 2 played noticeably worse (1.5 shots to 0 damage vs 5 to 1).

It’s clear here that team 1 was overall better. Could just be luck of the draw, but I’d also point out that Gambit was on team 2, and since he represents one of the most dramatic outliers and he was in a bottom tier vehicle, his team was likely at a disadvantage.

My point here is that the MM likely does not consider tier when attempting to balance the outliers. Without checking, I’d assume I’m the closest to Gambit in WN8 in this game, yet he was in a tier 8, I was in a tier 10.

Could I be pulling all of this out of thin air? Sure, but it’s my thoughts on what’s happening here.

apologies if this does not make the most sense

[lament mica]

Can I just point out that the losing team had more PS players? 😁

Like significantly more. 😂

Perhaps they should balance by console as well. 😅

I also watch the number of late loaders. I would guess that has a significant affect.

Also ping. It is very clear when you see a red with poor ping and what happens in cqc.

[errant saddle]

I love how ruthless did only 810 that match.

[uneven shoal]

Killed the Conq in the final 60s or so for most of that damage

[bleak oriole]

I'm pretty sure I've said it before - we don't try to balance win rates within the tiers. That way lies madness, because after we do that, then it will be "why aren't you balancing win rate within the classes?".

In the above example, there was one light per team, and there is a noticeable skill difference between the two players.

It's hard for me to trivially see how many players on team 1 survived - but looking at just the score line, the above looked like a fairly balanced / even match.

[coral zephyr]

MisterSPG, that match you are referencing is exactly the type of thing that gets me. [gets me losses, that is]

I play one of my light tanks and create the delta. Since win rate isn't balanced within tiers, the created delta very often gets balanced by pairs 1 or 2 tiers higher than me.

Which results in the opposing team getting the better of few top-tier pairs. But me, who created the delta, can't carry the team as my bottom-tier light tank simply can't output enough direct damage.

To me this is a recipe for disaster. Not a recipe for fun. The new MM is about as fun as it was to have 3 ARTY per team when you are in a bottom tier heavy tank. I COULD play, but it isn't fun. So I don't want to play and end up not.

[fleet elm]

Performance in lower tier tanks have the -2 games baked in to begin with. Most -2 games are an uphill battle without worse teammates. It's kinda like I was saying eariler, there are way more battles that are predetermined as a good player now

[coral zephyr]

It is as you say. Before it wasn't great. Now it is unbearable.

[mint brook]

I keep seeing games with far too many Zero damage players who never fire a shot. Either they are conscientious observers, or forgot to add shells before launching, or just out for a joy ride, or maybe they all died instantly on start up? But the match maker is really effective at putting them on my team.

[grand garden]

I'm assuming you're aiming for a close match. However, that's assuming all players are participating seriously.
However, as others have said, the reality is that there are still many irresponsible players who AFK, die for no reason, abandon their role, or give up on improving. This is rampant even at high tiers.
These players, whose win rates should actually be well below 50%, are now approaching 50%, so they'll probably continue to be irresponsible and not reflect on their actions.
Also, on some maps, if some tanks abandon their role, the game becomes painfully uninteresting. This is a characteristic of the game. For this reason, the irresponsible players I'm talking about are a major problem. A good example is when a light tank on a map with a good view and where scouting is important rushes into the enemy as fast as possible, abandoning its scouting role.

[fleet elm]

WG has data on the trend of match length and competitiveness. It seems like whatever they're doing is working (for their goals at least) since they're not changing anything

[modest ivy]

I’m not a good player by any means, but I have had more games like this lately than I can remember. It’s really frustrating, and honestly, makes me not even want to bother playing.

[bleak oriole]

Not sure that I see the issue in the above - if anything Team 2 is the one most likely to have complaints as they had 2 of the "0 0 0" players on them.

The matched ended with 3 survivors on Team 2 - so it looked rather close, but no idea how many HP they had.

Both teams had the majority of their damage concentrated in just a few players hands - but that doesn't seem unreasonable considering that they are TDs, SPGs, and Heavies. The rest of both teams did decent assist damage.

So, you'll have to specify what the issue is, because the above looked like a reasonable match.

[bleak oriole]

The amount of threads about "the MM is bad because every match is a blow out" seems greatly reduced.

[coral zephyr]

I would caution against just using the amounts of threads complaining as a metric.

[cunning dew]

No we just got tired of writing them cause nothing ever changed.

Quality of games has only gotten worse aswell, when the goal was to improve this😅

[bleak oriole]

Since we did tweak the system based upon feedback, and we implemented this system as a result of the "too many blowouts" feedback, I'd say that the statement of "nothing ever changed" is highly sus.

And the question was what about the match in the screenshot above was "worse" than the 9.4 experience?

[cunning dew]

Ok sorry somthin did change my bad. It got worse😅

Getting matched up with terrible players is not fun

[cunning dew]

Why should I try to be good if I will be punished? Tell me that, cause everytime I ask it gets ignored

[coral zephyr]

I suspect the 'quality' of players has dropped. There simply are more players who do little to no damage. Is there a solution for that? maybe, maybe not.

While I can appreciate what WOTMA was trying to do, I don't think this was the solution. -- there also might not be a GOOD solution.

The new system plainly penalizes better players by saddling them with more terrible players on their team. This doesn't encourage the better players.

More of a challenge? More like why bother. Better players are competitive by nature. Having the MM force more losses on them is not an acceptable solution to the better players.

IF the MM needs to be adjusted to account for 0 0 0 players, do so in a way that only considers those players. DON'T consider better win rate players any different than average 49% players in the equation.

this is it in meme form.

[near pond]

Player skill can’t be fixed with a flick of a switch. Trust me.

[cunning dew]

Also guys what happened to -1+1?

Its was loved and you never said anything else about it

[near pond]

Thats a different thing entirely

[cunning dew]

There's many things

It seems everytime they are about to do something good, something ruins it

Or they just dont implement for some reason

[near pond]

I know this is a thread about the matchmaker…….anyway who wants to hear bout the circulation issues in my arms

[cunning dew]

Me!

😅

[coral zephyr]

I think that match highlights what I have previously mentioned. Was it bad luck or the MM? We never know now. But the MM WILL try to put bad players on the teams of better players.

In the past we could say the MM was to blame for a loss. That the MM stacked the teams. But it was random. So just bad luck.

But NOW better payers know the MM will actively stack teams against them. Any particular match could be more bad luck than MM. But the MM may get blamed every time.

Do you see the difference?

Of course, now better players are reporting a lower win rate. The new MM is having a negative effect on them. Before people would complain about the MM, but those were considered "tin foil hat". The response and solution was to get better at the game. But now if one "gets good" they are penalized.

[cunning dew]

I love winning 10% less games. That just makes me wanna play so much more😁

[near pond]

Can we please stop pretending like winrate matters more than personal performance?

Wow. Your winrate went down. If your personal stats within those matches are still high then it does. Not. Matter.

[cunning dew]

I love doing 10k and loosing. Wow thats so fun

Ita not

Why try

I play the game for stats, if I dont have that I dont enjoy

[near pond]

What difference does winning make outside of the score on your profile and the end screen of the match?

[cunning dew]

I barely mark anymore as that was my favorite thing to do

[cunning dew]

The actual enjoyment of saying I did that

Cause when I pull off a 1 vs 5 or whatever im proud

Like how about doing a 1 vs 5 and you die to the last guy

Your gonna be upset

Im not alone on this either

Because the overall objective is to win the game

I mean thats what the game is about

Being better than others to come out on top and win

Competitiveness is about winning not loosing

[coral zephyr]

Some people play to win. They like winning.

[cunning dew]

Exactly

Thats what almost every video game is about

Atleast team based games

[cursive tinsel]

Id personally like to know if visible. How much platooning has increased myself. Because eit just feels like you trying to sort one area that some people have had issue with. But then if platooning has increased by a fair amount to avoid the issue, the issue has just been pushed elsewhere. At least to me it just doesn't make sense.

[fleet elm]

Quality for you. WG sacrificed the 1% to appease the masses. Their games got better

[bleak oriole]

Looking at some of the players in that match, I'm thinking that Team 2 - the winning team - had the higher skilled individual player.

As for the rest of this, it's a fundamental question of, if we are supposed to be a team game, then should the MM be balancing around that or not

[cunning dew]

Bro I have lots of 50% friends and they didn't even feel a difference. If thats who thwy are appealing to it didn't rly change

Just ruining the good players and basically didn't change the average and below average

How about wargamming teaches thw bad players? All they do is reward them atm

[fleet elm]

Eh that’s not feasible

[cunning dew]

Like why are our players so bad?

I cant understand

[near pond]

This is going to be a REALLY spicy opinion. But i feel like high winrate has been idolised as the defining factor of a good player for too long

[cunning dew]

Some of them have like 50k games aswell. Idk how that is

[cunning dew]

Well personally I like WN8 metrics but wargamming cant use that

Winrate and wn8 together I like

2 factors instead of 1

[near pond]

Wn8 is just as bad. If not worse because of how easy it is to manipulate and how many things it cant take into account

[cunning dew]

But wn8 can be farmed etc

[cunning dew]

Thats why I like both together

Because if they have low winrate and high wn8 it means they just farm

[near pond]

Still does not define the metrics of a good player

Remember WoT is genuinely meant to be a team based game

But right now players idolise the numbers that come from throwing your teammates under the bus

[coral zephyr]

One of my main points is that "better players" now will blame the MM for losses. Whether it is the case or not in any particular match. That is a significant change. It is also demoralizing for better players [whether it is true of not for a particular match].

[cunning dew]

Well whats a good player to you?

If someone wins more than another that means they are good and help the team

Because you cant win without the whole team winning

So a high winrate player would not throw others under the bus

Thats just a true statement

[near pond]

A win can be hollow

[bleak oriole]

I disagree - high winrate != skilled

Wins can be farmed via platooning

[cunning dew]

Well then why are you basing it off that if it doesn't mean skill?

[bleak oriole]

because there isn't anything else that is better

damage can be farmed

kills can be farmed

[cunning dew]

Then dont do anything at all?

It was fine before

[near pond]

Player A captures the enemy base before Player B had the opportunity to deal damage

Did player A help player B have a good match?

[cunning dew]

Now your quite literally helping bad players when they rly should just learn the game

[cunning dew]

Nobody likes cappers. Thats not how most wins are

Always the sh**er that caps

[bleak oriole]

it wasn't fine before - see previous comments about the blow outs.

[near pond]

But player B won! Thats good right!?

[bleak oriole]

the goal object is to make two teams where both sides have a roughly even chance of winning. it's never going to be 50/50 - but having close to that is judged to be better than having matches where it's 75/25

[cunning dew]

We still have the same blowouts! You guys dont even barely play the game. You just make a statement every time and say we are right you are wrong. Im trying to help this game as many others are and you guys dont listen. I love it to death and im rly trying to help

[near pond]

World of tanks is full of so many moving parts that telling someone to just get good is genuinely dumb

[cunning dew]

Alof of players i know have quite from the last matchmaking update

Its sad

[bleak oriole]

no, I play the game, I just do it under an account that isn't under the WGW tag

[cunning dew]

Ok well most of the wargamming dont

Theu literally have like 20 games in last 30 days😅

[coral zephyr]

before with random, it would balance out. But now if you are good, it never balances out.

[near pond]

You are aware that most WG staff will have personal accounts to prevent stalking right?

[cunning dew]

But pls can you answer this question ive been asking forever. Why should I try to be good in the game if im punished

I rly want to know

[bleak oriole]

hmm.. if there are two 60% win rate players, and 28 50%, the match will most likely two teams of 1 60% and 14 50% players?

not sure what can more even than that

I have to run in about 5 minutes, so I'll keep this shortish - what's more important: the journey or the destination?

If we go back to full random, then the 3 minute 14-0 blows out will increase, and your win rate % will go back up, but is that kind of match considered "fun"?

[coral zephyr]

Damage ratio would be a better stat. Yes damage can be farmed. But that is contributing to wins. With win rate you can get carried.

The skill difference between a 44% and a 49% win rate is likely much, much higher than the difference in percentages might look.

If someone has a very low damage ratio, they are likely rather useless to there team (yes some exceptions for spotting). but in general they are a net drag.

Damage ration would be a much better stat than win rate.

[near pond]

Yenaaaaahhhh

[bleak oriole]

the problem I see with that is that immediately there will be howls about the Heavy Players who sit back and Farm Damage...

[near pond]

Damage ratio is so easy to farm with tanks with good guns

[coral zephyr]

Yet the MM creates pairs and compares them. It can't even win rate balance between tiers. So again there might NOT be a good solution.

[cunning dew]

So theres no way to balance it sounds

Thags what im getting

[near pond]

T249 literally vomits damage ratio

[cunning dew]

How about you take a look at how PC does it

[bleak oriole]

well, the current system is "better" than the previous system - at least, at the 30,000 ft view

side note: I saw the other thread here about how the CW MM should use WN8 to make more balanced teams...

[cunning dew]

Question. Why is this thread open if there will be no change. Sounds like its set in stone

If anything it just wastes people's time if nothing will actually change

[bleak oriole]

PC has much large player base, so they can do slightly more complicated stuff - they have also spent a lot of time building up the subclass system

[cunning dew]

I see

[bleak oriole]

it actually isn't set it in stone - we keep it open in case there are reasonable suggestions

[cunning dew]

Have you came across any? Just curious

[coral zephyr]

I my "day job" I receive complaints and concerns all the time. I have to carefully consider each situation to determine if it is a legitimate complaint and if it something that can be solved. Some times action needs to be taken other time no action is required and or possible.

[bleak oriole]

but they seem to be "revert" - which doesn't fulfill our goals of having more balanced matches

[cunning dew]

Reasonable ones

[bleak oriole]

well, the feedback from the first pass of the system that we got from here was used to build up the second pass

and from looking at the results, most of the players seem ... happy is not the right word ... but not as unhappy

the whole snake system that I think Vishaw proposed original (or maybe someone else) - we looked at it and decided that it wouldn't do what we wanted

[cunning dew]

Also a question. Has this brought more players in or more players away?

Because all I here is leaving players😭

Its sad

[near pond]

This sounds kinda insane…….would adding a proficiency rank for a tank work?

[cunning dew]

Whats that mean?

[bleak oriole]

well, it's hard to quantify players coming back and players leaving - what thing changed that caused them to either leave or come back - skills? equipment? premium tanks balance changes? mm?

anyways, running time

cheers

[cunning dew]

Aight

[coral zephyr]

I am saying exactly why I am not playing WW2.

Everyone is okay with someone else paying the price, as long as it isn't them.

The new MM will result in good players leaving. It is basic human nature.

[autumn light]

changes like this won't reflect overly quickly in something like an active player count unless it's catastrophically bad, which it is not.

So to answer a slightly different question of "is there any sign that people are leaving as a result" the answer is no. If the question is are people staying around longer/more as a result, that is where it gets trickier and isn't something that will instantly show.

[coral zephyr]

Does WOTMA want to trade good players for bad? Is that a good trade?

[coral zephyr]

If one wants to persist with the new MM here is my suggestion:

Consider average winrates and above all the same. If you are 49% it is considered 49%. If you are 55% it is 49%. If you are 65% it is considered 49%.

This way you are only balancing out the low performers, which seem to be the biggest problem. There certainly are a lot of them.

But you are also NOT negatively effecting higher win rate individuals. They would be considered in the MM the same as average players. This would be fair to them and remove any reason for them (me) to complain as the MM would not be stacked against them.

The biggest complaint has always been people getting stuck with 0 0 0 players. This would attempt to address that rather than simply shifting the burden of more 0 0 0 players to good players.

The average and better players would randomly fill out the team. They need to "fight it out".

All that being said, the inability for the MM to balance out win rate within tiers is a fatal flaw in trying to balance out win rate at all. The complexity of creating matches with varying classes and the tier system may prevent a good solution at all. (will it be -2 or +2? or -1, +1? etc.)

Once you lose players it is more difficult to get them back. Certainly with a MM issue where the details are somewhat secret. How is one to know when/if it changes back?

Will that player even be paying attention if it eventual did? not likely

[proud sluice]

this new system is trash. go use a press accoutn and set your wins to 60%+ and go play. tell me how much fun you have. now maybe yall at WG are so bad at the game you wouldnt notice a difference. mayeb thats why you have to constantly ask for how to fix things stop listening to the community for solutions. you are PAID to make the game we just play it. the previous MM the sausage was occasionally too spicy and the current solution is to add EVEN MORE spice to the spicy sausage and triple the amount of spicy sausages in the pack so to speak. its an absolutely awful solution and its shameful to hide behind WELL the COMMUNITY told us to. Go get a spoon and dig it in the concrete and harden the heck up. YOU ALL are responsible for these changes not us. youre so shameful you wont even put out what the solution was because it must be so heinous it has to be kept hidden and esoteric.

[scarlet mantle]

For me it is easy… just do what the pc guys do and iam happy. All your console exclusives was just bad for the game 🤷🏻‍♂️

I can remember times with 20k + players on the servers

[lament mica]

I wonder if xp earned/minute is part of the formula.

[olive bolt]

Dang this thread still going? You all missing the most important tracked metric which really matters

Revenue

[proud sluice]

Yeah I for one cant wait for the real revenue crunch when a LLM comes in to replace these devs. learn a trade coders.

[modest ivy]

I agree with your analysis of the screenshot I posted, and the other comments that you have made in this thread. I’m not a great player, my overall WR and 30 day WR is around 54%.

In response to SPG, I’ll be honest, I don’t know what the solution to the MM is. I’m not the knowledgeable on the system and/or anything else regarding the background of the game. Doing damage is fun to an extent, but I like winning games more. I like running boosters. I like the feeling of accomplishment when you win.

My overall WN8 the last 5 days is 3797. My WR percentage… 42.7%. Not a typo… 42.7% win rate. As an average player. How is that fun? That is extremely frustrating and defeating to me. I logged in today, scrolled the garage for a few minutes, and then exited the game and went to WoT PC. I don’t want to play the game if this is the experience I will have.

[haughty terrace]

Key word being Better players. Since the majority of players are probably average or below. Does wg focus on them or the small group of good players. I don't have the solution, but from a business model. They will probably focus on the majority of players. The issue is that most players are not on the discord. Therefore they may actually be happy with the changes. We simply don't know. As wg stated in the past. When they run a test, players simply assume what the results were, without having all the info. Yes, I agree wg could probably do better at letting us in on it. However, that is their decision. I did enjoy +1/-1 when they ran the test. However, I also know players that prefer +2/-2 for the challenge and more xp. It's easy to say you want something that benefits you, but everyone would like something that benefits them. ( human nature) As I stated earlier. I don't have a solution, but instead of telling wg to go back. Which probably won't happen. Then think of new ways to improve mm and help wg. Something that could benefit both groups. Of course, it has to be realistic.

[coral zephyr]

What I am proposing wouldn't benefit me. It just wouldn't negatively affect me, as it does now.

If better players continued to get shafted, and they figure it out, they will leave. It is that simple. Unless they don't care about winning. But the whole change in the MM is about win rate. So win rate seems to matter.

It has been implied by some that the mm change isn't a lot of matches-certainly not enough to worry about. That is the MM is only really stacking a small number of matches. If that is the case, why is it even doing it? It is not worth the bad taste with good players for a small number of matches. If it is NOT a small number and having a great effect, good players are really getting screwed.

Again, there doesn't seem to be tons and tons of "good players". They are a definite minority. So why bother screwing them over? If I am mistaken and the better players are a significant number, one better not push them away and out of the game.

I am told that I am only getting screwed on a small number of my matches and don't worry about it. (That I have plenty of wins to spare). Let's ay that is true. I am in the 10% and only 10% of my matches are being made losses now. If that is really the case, the other 90% of players can't being seeing any real change at least nothing significant.

The 10% of my matches spread out over the other 90% of players is a very small number to them. Is it even 1% of matches for someone in the 90%? (bad and average players) Probably much less.

Bad players are getting stacked on teams due to good players. So those bad players end up losing those matches. That doesn't help them. Average players have not been saying their win rates are jumping up. How could they?

Again. I see no good reason to set up the MM like this.

It has been said the many players have stated they like the change and see "less blow outs". But how can that actually be true?

When WOTMA staff are asked if it is true, some have said "we need more long term data" Of course they do. And they have access to all the data. That is likely a very reasonably response.

YET, an average player with a poor to mid win rate can make a 100% call on how it is going with their limited matches? So why fall back on saying those players know what they are talking about.

[fleet elm]

How can less blowouts not be true?

I think the issue here is the WG is happy with the results of the change despite screwing over the top players. Hurting us is worth long term match quality. I think some rework of endgame high skill reward space is in order to make this change less awful for us @autumn light

[grand garden]

People are pointing out that if this matchmaking approach is repeated large enough, everyone will approach 50%.
To take an extreme example, if this is repeated and everyone reaches 50%, the matchmaker will no longer be able to distinguish between strong players. As a result, matches with large score differences will once again occur.
Furthermore, if victory or defeat is decided by 50% no matter what you do, beginners will not put in the effort to win.
What's more, incompatible with this game, wins and losses will also affect experience points and silver earned, leading to frustration for many players.

Adjusting player win rates to resolve dissatisfaction with large differences in match performance abandons any analysis of the causes of the large differences in match performance in the first place.
I apologize for saying this many times, but when matches result in large differences in performance, the most serious problems are "irresponsible and lethargic players" and "tanks with performance disproportionate to their tier."

[coral zephyr]

It is certainly possible there are not a significant less about of lopsided matches. I have yet to see an official stat. Saying less could mean 0.01%. Also what someone considers a significant amount could be different. Some might be happy with 1% drop in the frequency of lopsided matches. Other will say that isn't worth the squeeze.

But with players who don't fire their guns and/or don't leave spawn, all bets are off. How can one balance that? If you get a couple extra of those on one team, it will often be a blow out.

[fleet elm]

How do you ever balance that though? This mm makes sure that the people who do that get better teammates to counteract their awfulness

[coral zephyr]

That's the thing. You can't balance that. If 2-3 really terrible players are stacked on a team, there isn't much stopping that loss.

[proud sluice]

I could tell they changed the MM before ever reading the patch notes. I knew the MM was changed by the players I was seeing. I then went and read the MM was changed. In the old MM I very rarely saw these players that LITERALLY dont know how to use a controller. these people can barely navigate their turrets. and suddenly I was seeing 5-6 of these guys per team. its not fun to play with OR against. Some consideration needs to be taking regarding that now the wins feel less high and the losses feel WAY LOWER. The game just isnt as fun. whatever the MM was doing before it kept these players away from me and I was happy for that.

[coral zephyr]

It doesn’t seem to help to group 0 0 0 players on one team.

[fleet elm]

That’s the point of the mm though. 2-3 of those players won’t be on the same team in the new mm, they’ll be evenly distributed between teams. Unless there’s a good player to group them on the same team

[fleet elm]

You’d still see those players before the mm, just less of them.

[fleet elm]

It helps since everybody’s win rate is now closer to 50%

It’s all sorta relative to what your definition of “helps” is

[coral zephyr]

That is what happens. A good player is on one team.

[lament mica]

Yowsers

Not the “whole change” —it has not been and never been only winrate.

And you guys make it sound like in every game you and every good player are getting extremely and significantly shafted, which is just not the case.

[scarlet mantle]

For me solo it is noticeable 🤷🏻‍♂️

I just stop playing because of this uncomfortable feeling.

[fleet elm]

It's obviously not unmanageable, >55% win rates aren't hard to achieve. The feedback is literally that it isn't as fun for us anymore.

[grand garden]

I haven't seen it myself, but I think it's highly likely that the person who starts a thread saying "the MM is bad because every match is a blowout" is the person who is causing their team to lose.
I've certainly experienced losses by large margins before. But it wasn't to the point where I'd say it was "every match" before.
Isn't it highly likely that the opinions they were using were biased?

[uneven shoal]

I'm sitting at 60% right now, but thats still 2-3% lower than I was running before the change and it def feels like I have to carry more

[uneven shoal]

Gambit started this thread... I dont think he's one to cause his team to lose

[grand garden]

Of course, I'm not talking about this thread. I don't think Gambit said "We'll lose every game by a large margin" in the first place.

Did you see the MisterSPG chat I was replying to?

[proud sluice]

these baked in losses sure are fun wg

why are you balancing the matches AGAINSt the only players that care about winning 90% of players dont care about wins

[mint brook]

I have deep respect for the OP of this thread. The guy is one of the best players in the game, earned his status by learning how to win even when the odds are stacked. Gives me confidence when I see him on my team and concern when he is on the Red team. This thread is an honest validation for me that something changed and has made playing the game Much Less Enjoyable. Looking at my playing history. On THE day the changes came my WN8 dropped a lot. Its been super frustrating. Winning means playing ONLY powerful tanks. And even then its really hard to get a win. More Players than ever are only shooting a few rounds per game if they shoot at all. I have video one 000 player in a game yesterday who drove in circles shooting random objects. I looked them up and confirmed they are a teal player with a healthy 46% win rate and not a bot.
The broken MM and bad games for those who enjoy this game is a Symptom of a greater problem. My training as a fire investigator has taught me to search for the Origin and Cause through analysis, Evidence, and scientific method. The steady erosion of good players since 6.0 has upset the balance and now were at a point where there simply isnt enough good players to compensate for those who dont know how the game works. WG has consistently made business decisions that continue alienating those who care enough to try to win. And those people keep leaving.
So… they cant fix a match making problem when the problem is the ratio of good players vs new or bad players is so far off balance that only one to Three decent players per team are possible. ( look at how many people deal decent damage or kills).
WG needs to figure out why people leave and make those changes to win back their business. The MM will naturally improve.
Hint… its not Anime colab or cartoon turtle tanks or Premium tanks for a $100.

[summer plover]

Can confirm. That claim is basically insult, but I’ll also assume the person in question doesn’t know who gambit is

[lucid swift]

Can confirm, this match maker is sh_t 😅😅 do something with it..... No one ask who Gambit is 😅😅

[summer plover]

I’m a discord mod…. Not an employee…

[lucid swift]

I know, i just said...

[near pond]

[errant saddle]

Even worse now because I'm getting more messages for cheating when I'm using what's in the game for examples being in a sfac and the poor tier 6 not being able to spot me and not understanding the ring system that I said would be an issue and was told pretty much too bad for them if they wanna learn join discord or watch a streamer.

Or there's the occasional imma boot up a match and walk away and never move so they are afk, or players on older Gen who now load in after what seems like a minute

[near pond]

Tbf the sfac definitely should not see tier 6

[grand garden]

Perhaps players who normally play carelessly in PvE may feel that the quality of players has declined when they come to PvP because they think, "Even with careless play, you have a 50% chance of winning."

[lament mica]

Sure

Not so much directed toward you but there are others in here with higher winrates than life time rate and improved wn8 whining that it’s not as fun anymore. Surely has nothing to do with equipment and skill changes reducing the effectiveness of individuals right? Reducing the individual effectiveness to react in a shorter time span makes the game “more strategic” right? Surely “slowing” the game down has nothing to do with being able to cover over the inabilities of teammates right? More strategic surely does not mean having to think and plan more instead of just having fun. Surely not being able to move as fast to dispose of threats has no affect on the mental tax of feeling helpless when the other team moves in a more organized way and your individual skill lacks the time and reload speed to overcome. Just sit back more and plan strategic. All is more strategic now. Just need to plan better because more strategic is more team dependent and it is more fun.

[proud sluice]

what are you even saying? if you have a good winrate you are now forced to play very aggressive to stand any chance at all. this basically eliminates 90% of your garage. you have less than usually 3 minutes to impact the game before you are down 3-7 players. again this further reduces your garage choices.

[lament mica]

Need to strategize more on using that 90 percent. You have less reload and less accuracy.

[fleet elm]

And worse teammates exactly 88.7% of the time

[proud sluice]

the worst part is when you do get a winnable game you dont want to play anymore because your next 5-10 are going to be just awful experiences 🙁

[mint brook]

I imagine thats how people feel sitting in a casino…. “What the heck put another quarter in! “

[fleet elm]

Good thing that’s not how it works

[mint brook]

Really?

Loss #5

Say it ain’t so

[stray star]

[native plaza]

Also platoons break the mm, so technically you're not affected by the changes

[fleet elm]

No, the mm can still sorta account for platoons, but not if there are 3 extremely good players

[fleet elm]

Cool, the quality of your team isn't dependent on the prior game

[lament mica]

Considering the lower pop, seeing the same people from game to game is not out of the ordinary though.

[fleet elm]

More accurately, the quality isn’t dependent on the result of any previous games.

[bleak oriole]

As I've said before - the process for selecting the 30 people in a match remains exactly the same as it has been since the CBMM launched.

The only difference is the logic for the team assignment. Specifically, it used to be only by battle count, and now it is by vehicle win rate / account win rate with very minor contributions by battle count.

[scarlet mantle]

This little adjustment is enough to make me not playing Sir. It is just not enjoyable for me solo. 🤷🏻‍♂️

[cursive tinsel]

It seems to very from tank to ta k for me, but personally im just giving up playing solo, which was the main why Ive been playing for the most part. If I do play im playing as a platoon of 3, which to me, we are just shifting the issue to being what will be, now we have too many people platooning and MM struggling in that regard.

Its just not enjoyable to attempt to play after work. Some days were rough but it just seem near ever day and battle is rough. All because I went out my way to learn the game and improve.

[fleet elm]

They’ve been very clear that they’re willing to sacrifice the top 1% of players entirely for everybody else to have a slightly better time

[fleet elm]

The worst is when the map is unfavorable. Death Valley in a type 5 might be a 70-80% loss chance from the get go

[cursive tinsel]

The map pool and variation is a nother that just add to my general unwillingness to even bother, but thats really more a sperate subject. It just annoying as, the other day I played the type 61 god knows why as its awful to start, to then get shafted game after game over the course of a few days. Played the M46 patton, similar in WR, way easier time in comparison to start with.

I just dont get why punish those that were willing to learn to start with. Tank balance is where half my areas of concerns are. Not MM

[fleet elm]

For us maybe. For most players, having relatively fair matches that last longer is more fun

[cursive tinsel]

Don't get me wrong the game are last longer for the most part, but tahts more the QoL changes personally. All this is going to do is force the higher players to platoon far more frequently. Which just causes more issues.

[coral zephyr]

We are getting even more improvements for this weeks update. Outlines, UI, etc. Previously it was equipment, skills, etc. Lots of improvements (in my opinion).

However, ALL those improvements are completely cancelled out for me by this new MM. The game SHOULD be the best it has been for years. Yet with the MM it is the worst it has been.

I just don't want to play WW2 anymore. CW still has its issues too.

[paper prism]

Then why not to match each player in a given tank with another player who plays on a tank of same tier(!) and class (!) with similar WR(!). Summed up WRs of teams may be not quite similar in result due to WR different tolerance but player to player matching will be much more fair. There will no situations when you match 65% player and bunch of 40-45% on one team and mostly 50+-% on opposite. Odds aren’t fair in this scenario.

[bleak oriole]

because there might not be such a player in the queue? we might have a 65% light tank driver and a 65% td driver?

[coral zephyr]

Decrease the maximum delta to 1. Basically alternate better winrate player in each pair between teams.

[cursive tinsel]

Id personally say turn it off so a set period and see if even notice it back to how it was. As we know the same scenarios can and do still happen. Most the MM complaints get dealt with by a tank rebalance. Or, by the QoL changes that shredded half the busted skills out for good.

[paper prism]

That may be happening. Matching time will increase for sure but fair balance worth waiting imo.
Also if waiting time happens to exceed some tolerated threshold just fill gaps with random tanks of required tier & class controlled by AI that mimics a player skill required to make a match.
Lets run a test event and see how come.

[fleet elm]

Which is why it shouldn’t be “target equal teams”

So much of the effectiveness of a good player in certain tanks is map dependent

Manticore on Ensk. 70% loss. Type 5 on Death Valley. Same

I’ve literally lost that Death Valley scenario before I had a tank to shoot

[bleak oriole]

but ... wasn't that also true in 9.4? a Manticore on Ensk is generally never going to have a good time

[fleet elm]

But the rest of my team isn’t at a disadvantage

So the match is still no worse than a 50/50 on average

With a worse team from the current mm, it’s much worse than a 50/50

[fleet elm]

@bleak oriole can you check in any way stats on “bad” maps for good players? I’m curious how bad the disadvantage is

[coral zephyr]

That was always the case. NOW we can have a bad time on ANY map, if the MM stacks our team with terrible players. We can get the best scouting map (Westfield?) and have MM chosen team that doesn't leave spawn and/or even shoot.

Before we could have bad luck with a bad team or bad map. Now the MM is creating "Bad Luck" for us.

[fleet elm]

It’s specifically bad map bad team for me. Scout tank on Westfield can carry some of the worst teams but a type 5 or maus needs some reliable teammates to do the same

[scarlet mantle]

Not even there… nobody will shot at your spots. At least this is my experience in light tanks lately.

Or they shot but without doing damage.

Or the positioning of your mates is so bad that they can’t do anything.

This kind of battles happens way too often now.

[errant saddle]

Not enough city maps even in cw

[autumn light]

This is very close to how it works now somewhat ironically. Fortunately jokerz said it so i don't have to be controversial lol, but i'd imagine that the majority "bad games" being attributed to this change are actually games that'd have happened more often previously. It's just now with something to blame it creates a perception that it must be because of this, when the reality is what changed is the extremes, all the "random" like stuff in the middle will happen basically the same just more often, with very small shifts in what "outlier" players may receive. But they are more likely to be impacted by an extreme caused by platoons than by the changes... but with the other extremes removed by the change it is the change that is blamed for those games caused by something else (say platoons).

Given that the complaint is still the idea that by being an outlier player you are getting teams "stacked" against you, but that just literally isn't happening as a result of this change... you were more likely to have a "stacked" team against you with the old system than the new, and the heavily stacked is now basically impossible compared to older system.

But as often the case the issue exists in that perception = reality, and to some extent what is actually happening has almost become irrelevant. We could reverse the change tomorrow and i'd imagine that for the rest of the games life we'll see people saying we never actually reversed it, and i'd argue something similar is happening now. We made changes and have said repeatedly that now it will not "stack" a team with the lower point of a pair - as part of a direct attempt to address the feedback. Yet "Stacks our team with terrible players" is still said even after acknowledgements that we have said the changes will not cause this. The old system would allow that, and things like platoons could cause that now. But the changes literally do the opposite and reduce "stacking" of teams - even for outlier players.

[fleet elm]

Lack of transparency does nothing to stop speculation and complaints. And winrates (and fun) solo queueing are still bad

At some point, stop the waffling and tell us to deal with it so we can properly quit.

[mint brook]

I still argue the real issue is imbalance of good players vs new or bad players. Too many good players have left. Now one to three decent to good players possible in a match. If one of them is taken out. Its Basicly game over for that team. Now more than ever the good players must 1. Choose a good tank, which limits what they can play, 2. Play their very best. 3. Take out the other teams good players as early as possible.

[cursive tinsel]

To be fair my, bad games if more toward what the rest of the player base probably felt prior. Bad MM get more the blame of poorly balanced tanks combined with not the best maps in rotation and variants. because everything has an assault version, because reasosns.

The second point yeah, I could garentee those that raise the MM issues before the changes even started, if you removed it tomorrow, they wouldn't even notice.

For me, this is just going to push the issue elsewhere. For certain players, solo play just isnt worth it. Myself included. However flip side to the coin is that if platooning on the higher end increases, you still going to have the same issues as prior. Just now behind a platoon which is harder to account for as whats been stated numerous times.

In my eyes, MM gets the blame when its really other factors that make it feel worse. I get doing it, but at this time of life, I dont get why focus so much on it. Pubs is meant to be random in my eyes, trying to make it 'fairer' just add more issues elsewhere or shift the issue from one area to the next.

[proud sluice]

did yall vibe code this change then? theres no way in heck we arent being stacked way harder. maybe you are right and its what @fleet elm said earlier. most of the good players have left from the first implementation or perhaps in combination of having to redo every tank and commander. Now theres just not enough decent players to make teams, I guess my platoon is waiting for WoT:HEAT then.

Yall are absolutelty certain theres no bug going on here? it feels like a substantial change

[uneven shoal]

It might be that I'm now paying closer attention to how my team ends up performing, but I feel like I've seen a lot of this structure of teams:
2-4 good players
10ish ok to bad players
3-5 detriment players

I'm not sure if this is the structure the MM is trying to make, but when I look at the scoreboard, this is often what I see.

Maybe @bleak oriole can answer this. What is the expected % of HP remaining for a winning team? Typically I see in the 30-40% range for really bad games (a steamroll in either direction) and 10-20% for a more "average" match.

Close matches are usually under that 10ish% line

[proud sluice]

Its really nice to hear that we the players are mistaken and that the system is so much better. meanwhile EVERYONES wins in this thread have dropped dramatically. what a clown take. "Pity the emperor who sits alone on his throne" -Izaro

[bleak oriole]

tbh, I'm not sure what design ranks them at - my personal scale is:

• < 5% - awesome match
• 5% <-> 10% - great match
• 10% <-> 20% - good match
• 20% <-> 30% - okay match
• 30% <-> 40% - meh match

• 40% - bad match

The thing to remember is that RNG has an impact - an unlucky ammo rack at the start of the match could turn what should have a been a 5% match into a 40% result because a team lost their only scout tank on a map like Prok.

[native plaza]

I find it massively depends on the tank I'm playing, I would say games on AVG feel better but I still have more wash outs with 8+ doing only 1 pen of damage

[native plaza]

Personally I think this method has some oversight with the possibility players hoarding their own hp while the majority are low hp and the enemy has a more even spread hp

@bleak oriole my point of view as a player. Games which are active (IE, people not camping spawn or one position) are the best followed by long battles then close hp difference

[autumn light]

There is no real target structure, just that there is a delta of 0 the delta is as close to 0 as possible. But the tweaks done ensured that even for an outlier the number of players required to "undo" the delta created by a worst case outlier vs outlier situation (lowest possible vs highest) was drastically reduced, and effectively capped... of course with the caveat i mentioned of why mister SPG loves platoons 🙂 😄

[bleak oriole]

there is an attempt to get a delta of 0 - not always possible

for platoon reasons mostly, but other situations as well

[autumn light]

no one said your win rate didn't drop, and it did for some and didn't for others. What i said was that the blame being attributed to the delta system for "stacking the team" with bad players is not true and goes against how it works. Basically the system can not put 5-6 "worse" of the pair on your team, and therefore blaming it for that is clearly not accurate. So you are trying to counter a point i never made 🤷‍♂️

[native plaza]

Does this data included capping

[autumn light]

thanks for the better wording/clearer point on that, changed my wording

[native plaza]

Or is that something which is not considered a outlier

[autumn light]

Outlier in this situation is in terms of the points assigned at the start of the game, that the "sorting" system then uses to assign which team players get on. So outlier in this case is someone at an extreme of these points, and the situation that was originally brought up here was that increased chance that a high win rate player will consistently get a large delta to be "overcome" due to the likely chance of being matched against someone with a much lower point amount.

What i have been trying to repeatedly point out is that this maximum delta has been massively changed with the tweak that happened to ensure a more consistent and tighter spread of the "better" or "worse" players from the pairs.

[native plaza]

I know sorry if I'm not making this clear 😢, I was more on about what a healthy game is/ what is considered to be a healthy game internally.

[fleet elm]

Well the expected is 7 so having an additional 5-6 would be absurd. 3-4 is already enough to feel like my teammates are useless

[autumn light]

oh, no clue. I'm not sure if there is an official "this is the definition of a perfect game". Obviously in terms of that the only stated aims were to increase match time + reduce "steam rolls". Which the initial changes did achieve this but not to the extent we'd have liked to see, in terms of the since the tweaks i haven't got anything i can share on the results since.

[fleet elm]

well when I get a type 5 on death valley, there's gonna be a steamroll. My side likely has a 1/3 chance of winning at best

[native plaza]

Tank limits would have a bigger impact than any more mm changes

[fleet elm]

Not really. Tank distribution isn't really an issue unless something just got given away or something broken got released

[native plaza]

Personally I have always had 3-4 bad players on my team the issue when these changes first came out I was getting 8-10 consistently now its back to around 4-5

[native plaza]

Disagree having 8+ tds or 5+ lights ruins the game before it even starts

Same as double arty on most maps

[fleet elm]

That's rare enough that I don't consider it an issue. Double arty isn't awful either

[native plaza]

It's not rare from my experience, pretty much every game is one or the other, whether it be 6+tds or 4 lights or double arty, worse case it's all three combined

@fleet elm also considering you have only played 250 battles in the last 90 days and 150 being in the techno, I don't think that's a good sample to base experiences on

[fleet elm]

6+ TDs is different than 8 and double arty was a different message entirely

[fleet elm]

What does that matter though? Most of my battles since the change have been solo Techno and I still hate it

[native plaza]

Playing a TD against tds is different experience

[fleet elm]

It's more like a light/TD/med hybrid though. Unfair to call it a traditional TD

[native plaza]

? So like 60% of the other tds then

[fleet elm]

Literally just the grille and taran

Ig the T114 lol

[native plaza]

Bruh there are other tiers 😂

[fleet elm]

Nah bro

I got lost in tier 8 comp for a while bc I didn't know what anything was. And that was with the most egregious prems banned

What you got BFG lol. You've been typing a hot minute

[native plaza]

His kid probably has the phone or he has sat on it

[autumn light]

well 7.5 would be the dead middle of 15 pairs, but if we're assume that you are always going to be the best in your pair then 7 would be the midpoint of the remaining (0 to 14).

And the point i was making is that for a team to be "stacked" then having a maximum variation of only 1 or 2 "more or less" is what the sorting/delta system will achieve pretty much every time in a perfect world. The thing being said still here is that the changes are resulting in say 4-5 more "worse of the pair players" being put on the higher win rate persons team as a result of the delta system. My point is that is as good as not possible in a perfect situation. And whereas the old system would result in anywhere from 0 to 14 in terms of the stacking/spreading of "better" pairs, the new system will not, it will keep a much tighter spread that is far closer to a variation of 0.

So the odds of coming across a team with say "4" stacked against you is drastically more likely before than it is now. I'd even say that having 3 is significantly less likely now. So those games may still happen, but it isn't due to this system. If that makes sense/you get my point. That's why i have said at some point in this thread about the real change/factor is more what you have brought up i believe, and it's the exposure/rate of having a smaller disparity. AKA if before the average was say 7, but the max variance was also 7, now the average may be 7.5 but the max variance is 2 and before the tweek maybe it was 8.5 and 4. But how often you see the games within that 2 variance has gone from say 40% to 99%, and with that slight shift in average that means the games where your team is say 1 to 2 players worse, has gone from being 20% to say 51%. The number of games where you have 1 to 2 players better has also gone from 20% to 48%, but basically people won't notice that part anywhere near to the same extent.

[autumn light]

fault is somewhat correct, trying to get daughter to sleep lol

[fleet elm]

Ah pacing while holding a baby and typing a message on my phone is all to common of an occurance lol

[fleet elm]

Well I mean that's good to know but it's still noticeably worse than "random" mm to me and others obviously. And I'm pretty sure queue cancelling is very effective since I leveled out at 60% with the worse version of the sbmm and have gone down since while queuing normally

[autumn light]

There is a chance people are platooning more due to a perceived issue, and platoons are actually creating more "stacked" teams as a result and that ironically creates a feedback loop of confirmation. Entirely guess work with no data backing, but that is probably more likely than the delta system creating a lot of "stacked" sides.

[native plaza]

Is there any way to filter winrate on your end, into platoons and non platoons?

[autumn light]

for someone, probably. For me specifically, not at hand 😅

[native plaza]

Yh I kinda didn't me you, but yh you got the point 😂

Maybe instead of you using tanks winrate/overall winrate, it's better to be spilt into only solo tank winrate/solo overall winrate

As that would be a better representation

[fleet elm]

Wouldn’t change anything. Most affected players are so far over the winrate cap that it wouldn’t matter

[mint brook]

Blah Blah Blah!!! Talk talk talk.

Are you guys going do something to fix this game or not?

Just say.
Hey. Were sick of the game trying to let you guys down soft. But. The ride is over…
or
Were trying to fix it, but its really complicated and we Accidentally laid off the people who knew what to do…
Just tell us. Because it sucks loosing game after game and isn’t much fun anymore

[proud sluice]

My winrate has dropped from 63% to 60% thats substantial

[autumn light]

no one said your win rate didn't drop Literal first line of what i said, and as the final line said So you are trying to counter a point i never made

[proud sluice]

its the only data I have to back up my claim what do you want me to say? my playtime has dropped more than 90% and my wins have dropped 3% in that time thats SUBSTANTIAL

[fleet elm]

They know it dropped. That was the point. They don't care about good players and made our experience worse to improve everybody elses

[cunning dew]

Ypu mines even worse now

Am currently at 57% and was 70% before the update

I dont even care anymore

[lament mica]

They don’t care so much they spend all their time in discord in one thread. 🤔

Repeating the same thing over and over though people are too busy not having fun to pay attention 90 percent of the time. 🫠

[proud sluice]

thats the saddest news in gaming right there 🙁

[cunning dew]

I dont even mark tanks anymore and stuff. Killed my love for the game

[fleet elm]

They really need to update the reward space for good players if they want to keep us around

[fleet elm]

easiest way to mark is to throw 100 battles in a tank then start trying

[proud sluice]

I think it was nate who said the system itself has muddied where the players belong. its brought players up higher in wins than theywould otherwise be and brought down the wins of good players. i suppose eventually it will be back to random matchmaking XD

[fleet elm]

That's not true. 55% win rate is still easy solo

Just a narrower spread of winrates

[proud sluice]

tell thats to @cunning dew

[cunning dew]

Yea and im platooning aswell

57

Its sad

I feel like i suck at the game now

[proud sluice]

dont sweat it. if you want to try hard you cant play any slow tanks/ slowest a tank can be is nightstalker speed. also need to have good armor and DPM so very few tanks available

[proud sluice]

hey dont mention the "P" word they say the matchmaking wasnt changed for platoons LOL me and my platoon doesnt play much at all anymore

[autumn light]

I think you are somewhat missing the point i made. My point was blaming the change for "stacking" a team with a large number of "worse of the pair" players is blaming the wrong thing. As a result of feedback the changes to the topic of this thread would, in a perfect situation, give 1 side no more than say 2 better/worse pairs to compensate for the hypothetical worst case possible points mismatch. This means that the number of times your team gets a large number of "worse of the pair" players as a result of this system is almost certain far lower now than before - the same thing also applies to "better of the pair" situations though. So i'm not saying things aren't impacting your win rate, i'm not saying you aren't getting times where you get a team "stacked" with worse of the pair players, i'm saying that given the mechanics of how this system works, it is not the cause.

Perhaps due to how much less common those kind of games (in both directions) are, the ones caused by platoons messing with the system get a larger focus and they feel more jarring and obvious. Perhaps people's perceived issues with the MM have resulted in more platoons, and more platoons have actually increased some of these kind of games... maybe more so if you have platoons increasing largely above higher win rate players... i have no clue, pure speculation. But at a mechanic level, this system will not cause your team to be "stacked" with worse players.

Does that make a bit more sense? I was writing the original stuff at like 2am 😄 i promise no perfect explanations at the best of times, nevermind at 2 to 4am lol

[cunning dew]

Problem is all my favorite tanks were sh** tanks😂. Nemesis mbt, ebr 105, kranvagn,thunderchief. Now im forced to play competitive tanks

And I dont enjoy them

Because of matchmaking

[proud sluice]

thunderchief might be ok but t10 is tricky to carry in

[coral zephyr]

I think this is where is disconnect and/or misunderstanding lies.

I and other are saying that the MM will "stack a team against me".

Devs think we are meaning that we think the MM will stack the teams with "a large number of 'the worse pair' players". That is not really the case.

What I and others actual mean is the MM will stack our teams with 1-2 extra of 'the worst pair' players. That 1-2 extra is enough to complain about for a good player that cares about win rate.

Devs seem to think having 1-2 extra of "the worst pair" players is nothing to complain about. Some players feel like we are being told that the MM isn't changing back, it's no big deal and get over it. Not in those specific words of course.

Much of the misunderstanding arise from the difficulty of trying to have a meaningful and complex conversation over Discord. That is why even with discord, emails, etc. face to face meetings are still needed for many things. If not face to face in person, at lease over the phone/voice or online video meeting.

It is simply difficult to clarify text without seeing a person face to face. There also inherent difficulties in expression something over text.

Long story short:
We don't think the MM is giving us several extra bad players. We know it is only 1-2 extra terrible players. Also 1-2 extra terrible players is too much.

1-2 extra by chance (random), fine.
1-2 extra by design, not acceptable.

[errant saddle]

Noticed a lot of players never loading in or after a minute has passed resulting in losses most of the time because there still at base providing nothing to the team

[coral zephyr]

That is a big issue.

[fleet elm]

That’s affects all mm equally stacked or not

[fleet elm]

My issue is that we’re obviously few enough that WG doesn’t mind forcing us to platoon or leave. But in that case, why not just ignore us and do a softer mm where there are equal “worse” halves on each team?

[coral zephyr]

That type of player is likely to have a lower win rate. If they do nothing for a minute, it effects their win rate.

The new MM is more likely to put an AFK player on a good players team rather than on an average or worse player's team.

So the assignment of the chronicly AFK player is more often with the good player.

[coral zephyr]

I have said this as well. Just balance poor players and leave everyone else out. OR count good players the same as average. Don't give good players worse MM than average players.

Count everyone 50% and above as 50%. Although the real average is likely lower than 50%.

One can't expect to keep giving someone the worse deal, purposely and expect them to stick around.

I don't the that you're the problem with lopsided matches, Nate. The big problem is AFK players or those that are effectively AFK.

So they shouldn't treat good players any different than average ones.

[lament mica]

Yeah your alter ego cousin doesn’t think it’s 1-2.

And it’s not necessarily the worst players of the match.

You also still seemed to be stuck on the every match mentality.

[fleet elm]

It is >=2 per match most matches. If you aren’t getting the worst players then it’s 4-5 since the game has to balance the fact that they got put on the other team

[mint brook]

Thought I would try a few games. Seems improved to me…
Im not a huge fan of statistics… but this isn’t possible without some sort of rigging. Teams were blow out so fast I hardly had time to get going before being surrounded and killed

[fleet elm]

You know nothing about statistics if you think 4 games means anything

[viscid gorge]

Need atleast 30 to add on to that

[mint brook]

Ended up with 6 straight loses before a win.. You guys keep talking back and forth about all kinds of things.. at the end of the day, the game is broken.. stop defending it

[viscid gorge]

Nobody is defending the MM. However, we will refute baseless claims like how the MM is "rigged"

And also, your screenshot means nothing until you have more battles

[scarlet mantle]

Tier 10.25?
Cw mm doesn’t changed right 🤷🏻‍♂️

[viscid gorge]

Also that

[proud sluice]

you have to get the losses out of the way for the one winnable game. they say it dont be like that but it do.

[proud sluice]

game is just totally ruined

[mint brook]

The only reason I got some wins is i platooned with my clans best player. We both rolled out FV4211 configured with max armor equipment. Played ultra conservative and stayed together all match.
Then we won games

Before that… solo… bull crap teams. Blow out games

[fleet elm]

Say it with me, CW matchmaking hasn’t changed

[mint brook]

As far as the mm being rigged… it is, and for those who say we havnt shown enough data to prove that…. Not talking to you, because you cant fix it. Were talking to WG. And they have all the data…
So. Like Iron Man said…. Thought we wouldn’t notice? But we did. Fix it !

[scarlet mantle]

But with your feelings you destroying this tread. Cw is not involved in the changes they did.

[coral zephyr]

Previously I asked if the CW MM was going to be changed and I never saw a response.

[viscid gorge]

https://tenor.com/view/guy-arguing-guy-talking-to-wall-talking-brick-wall-gif-18667615

[mint brook]

Its pretty much the most accurate thing in this thread

And helpful…. Really.

Going away now. So long! And thanks for the games!

[viscid gorge]

(You are the wall)

[coral zephyr]

Did any staff confirm the CW hasn't been changed recently? The WW2 change was some time ago. We have have a new season now.

Why would they say the CW was changed, IF it was. To get another 1000 post thread?

I have no idea if the CW was changed recently. But they should say it wasn't if it wasn't. Last they said anything was a bunch of updates ago. No need for conspiracy theories. WG can easily clear it up.

[cursive tinsel]

CW wont be a priority yet until the whining eventually lead to it being looked at. Though I imagine trying to sort any MM changes across the eras is more hassle then worth. I was stated in the past numerous time that CW hasnt and wont be changed any time soon if at all.

Instead us poor sod playing WWII its either suffer solo, 3 man platoon constantly or give up playing 90% of the tanks in the rooster because its not 'meta'. Because working all day and having to play like your in a comp game continously is so fun... oh wait

[bleak oriole]

The CW MM remains different from the WW2 one. When we are ready to deploy a similar style MM to CW, we will let you all know in advance.

[viscid gorge]

Does the current pairs maker take into account tier?

Like my 60% winrate in a tier 8 won't match a 60% winrate in a tier 10?

Idk if that's been asked and I am not gonna be able to read back through 1000 messages 😭

[autumn light]

it doesn't match win rates, but yes pairs are class + tier (and other stuff, but it makes sure same class + tier... doesn't care about stats)

[viscid gorge]

Oh okay, thanks. I just wanted to be sure

[fleet elm]

It doesn’t look at tier, just delta between the pair

[scarlet mantle]

„And other stuff“
So basically skill based matchmaking 🤷🏻‍♂️

[autumn light]

Still no. But to quickly repeat, stuff means things like time in queue. As in simple english as possible so no one can misunderstand.

THE MATCH MAKER DOES NOT LOOK AT WIN RATE OR ANY PLAYER STATS.

There, that should help 🙂

[paper prism]

Simple english is always welcome🙂 It relates only to CW MM, correct?

[coral zephyr]

The Win-rate Delta MM might work better in CW.

The tiers are balanced closer together in CW. So when a delta created in a low tier pair gets balanced by a higher tier pair delta, it won't create such a team imbalance. Still an imbalance, but not as great since the power gap between tiers is less.

[cursive tinsel]

Well, not for the pulling of the 30 players. But SPG said in a comment further back that it that rather then sorting with battle count it now sorts with vechile WR/ account WR so...

[near pond]

The matchmaker is the pulling of the 30 players

The sorting of those players is a separate but connected system

Im fairly certain this was said like 12 times

[cursive tinsel]

Yeah, this is the message I was referring too since this is really the latest of it in my last reply.

[unique lance]

So your saying the consideration of player Win Rate has been pulled out? It doesn't consider the winrate of the player in that tank or if not enough games consider the general winrate of the player? You need to get BAM1500 to communicate that if this is true. Some how people aren't going to believe you, here.

[unique lance]

Notice I said nothing about matching winrates...I want to see you say it doesn't make use of or consider any portion of winrate, even in scoring and evaluating player points. If this is true, then there is truly no player skill being consider as part of the match matching or anything leading up to the actual match maker. If this is true this needs to be communicated as part of the stream for WoTC and come from BAM1500 since he was the one that told players this was the case.

[unique lance]

I'm not sure that's the entire story DeathPigeons. RudoplhBobrBFG may not be actually answering the questions that you meant to ask.

[unique lance]

Is Win Rate used in calculating anything like pairing as you mentioned above? You guys make this so confusing and really doesn't have to be, but instead of taking my advice from a month ago and really laying out for everyone how the Pairing and Match Maker actually work, you constantly chase your tail in trying to explain. As long as you can please, confirm Win Rate isn't used in any way to determine who is on a team in a match then that needs to be communicated to the player base.

[unique lance]

Just so I'm clear this would be for both CW and WWII.

[unique lance]

This is why I said a month ago they should lay out in detail how this works and then provide a link to the webpage explaining exactly how it worked before the change and then how it would works after the change. This way they wouldn't need to repeat themselves over and over again. I believe at this point they want misunderstanding and are hoping people just give up.

People now have 2 months + of data and we know this is occurring and know players with high winrates are being affected by the "Creation of matches" no matter how it is accomplished.

[autumn light]

My answer was to someone talking about the pair based MM system, the way the 30 players are selected... this has not now, nor ever cared about stats of any type. As has been said probably 50 to 100 times at this point. The sorting system AFTER the 15 pairs have been selected by the MM is the part that uses a points based delta system to assign which of each pair goes to each time, this system also having existed for a long time, was adjusted to add win rate to the points calculation and then adjusted again to address feedback and results we were aiming for but to a lessed success than we had hoped.

[unique lance]

You can keep saying it, because people believe it to be the same "Creation of a battle". You maybe technically right and I do understand what your trying to say, but that's why this all should be explained on a webpage that you can refer to...thanks for replying by the way.

[autumn light]

no i exactly answered the question he asked, but as if a pretty common thing in this thread people are mixing up the different parts of the process of getting into a game.

[autumn light]

i mean i could provide it down to the last detail and ironically this thread and a few others have probably proved that it won't matter 😅 many people will have decided what is happening before they have read anything and will not believe anything we say... BUT overall i think it's pretty fair to say that the attention showed by red names to this thread show we have been trying to improve the communication around this issue. And we'll keep looking to do more to clear up any confusion 🙂

[unique lance]

Yes, I agree with people are mixing them up because they don't care they are separate. They think of it as one thing not two. I understand that technically they are different, but if you keep referring to them as separate people are going to continue not understanding. That's why "again" I would encourage you or some at WG to document the process on a webpage and then you can refer to the link instead of answering over and over again.

I'm sure people would very much appreciate and do appreciate you trying to clear up the understanding

So then to be clear to get an answer to my question, yes the winrate is still being used in the points calc, to determine player "pairing"?

[unique lance]

Sorry the "Sorting"?

[autumn light]

them being different is a massive thing though. To create a kinda odd but hopefully relatable to many analogy:

If you think about being at school and then playing a sport in gym/sport class/break/etc.. Your class was decided by one set of factors... this is what decided you are at that school and in that class (encatchment, age, siblings, etc). Later when a teacher selects the team by lining you all up and giving you a number from 1 to 30, and then saying "odds to team 1, evens to team 2", it had no impact and was not related in any way to why you were in that class/school.

That is what we keep trying to explain basically

[unique lance]

I agree and understand, document it and put is on the website and point people to it. You will get people using the same terminology and it will be easier on you to point them there.

Document in full context like your anology.

[autumn light]

in the analogy, the pairing is picking the 30 players is being in the school (No player stats involved), and being picked in gym class is the "sorting" (player stats are involved)

[unique lance]

Yes, but there is more detail to both and they both need to documented and displayed, so that the community can understand and see they are separate. I truly believe it would help with understanding and getting people to align on the same verbage when talking about it.

By the way I would even use the anology on the webpage too!

[autumn light]

Did the analogy/extra info help explain this btw? I'm about to try go back to bed and don't want to just vanish if you have something a bit unclear still 😅

[unique lance]

Yes, though I already understood this, but YES

[unique lance]

Let's be honest here...the whole point of playing your game is to win your match, it even use to say this on your website and it still might. It's not a good idea (in my opinion) to undermind this by putting the weight on the scales so to speak. The real problem is experienced (knowledable) players have an advantaged over inexperienced (less knowledable) players. This isn't going to address the real problem. Their are in my opinion much better ways to address the real problem. However, it's not to punish the more knowledable players, because they will just leave if they determine the game to be unfair to them, but maybe that's the goal. Either way I really do appreciate what you do Rudolph, as I know it's not easy to monitor threads like this. I've been in IT for 35 years. I have a CS degree and been an IT PM for 25+, so just know I truly mean that.

[unique lance]

Have a good night sir

[coral zephyr]

I don't think the problem is that good players are better. How can it be a bad thing for someone to learn, put in the effort and get better?

The issue is the vast number of players doing poorly. This game has been around well over 10 years, yet a lot of players are doing worse.

They play tanks that are too high of a tier. They should play lower tiers. At least for the ones that try. Then they can practice and get a little better.

But there seem to be a lot that don't try and/or don't care. Those ones are really ruining the game for everyone.

We need more good players not less. But I agree with you this mm will lead to less good players sticking around. You can't hardly play anything but tier 10 now. If you get -2, you are often really out of luck. A higher winrate opposing player has 2/3 chance of being higher tier than you. With the substantial power differences between tiers, you are at a disadvantage.

But I am sure you already know all this. That is why you are posting here.

[near pond]

At this point just give the team distribution system an official name so confusion around it can die.

[cursive tinsel]

Or just remove it and see how it handles for a set period given the QoL implementation and see if that alone effects it rather then trying to take into account players WR and Tank WR post the 30 pick. Because you know, just because I have a decent WR in a horrendous tanks pre changes doesnt mean im going to be pulling the same when facing a similar player in a vastly better tank that also got pulled in.

[near pond]

Not a good idea. Remember the whole reason it was changed was to improve the blowout match situation.

Besides. The other player with a similar WR isn’t being pulled into the match because of the winrate similarities. They’re being pulled into the match based on the tank class they are playing. That same player would have still been in your match. They were just put on the enemy team because of the WR similarities

[cursive tinsel]

I get that, but as it been stated, blowout can and will still happen so I'd gernally be interested to know the levels of complete blowout prior versus now.

[cursive tinsel]

Yes that I get, but say me in a WZ-120 just for example is not going to really be the same as a similar player with a M46 patton for example the vast majority of the time. Tanks that haven't been buffed or a power creeped because everything else around it has been buffed to oblivion isnt going to impact the game.

Hence why if something like this were to happen, all tanks across the board need fully assessing because there's some may out standing tanks across the board and classes to others.

[near pond]

You say you get that but you just said something completely opposite

[unique lance]

I really said more knowledgable players are better in my post , not sure what your point is here, other then I do agree the game needs the more knowledgable players. I did express there are better ways to solve the problem, though I didn't share them here, because this is about the "general" subject the Match Maker.

[near pond]

Oh dear god

[unique lance]

Yes we need common terminology and documentation(explanation) so that everyone is on the same page, before the discussion starts. Then the actual problem trying to be solved should be stated as well. This would provide more a less confusing discussion and a more productive feedback. I think Rudolph understood that.

[near pond]

This thread is only a thing because the OP doesn’t know that the MM is separate from the system that distributes players between teams

The MM finds the players

The secondary system then distributes those players according to set rules. Those rules are what was changed

[unique lance]

The reason for the change is also confusing, because though I've heard WG staff give the reason you state, other staff have given other reasons. If it's about blowouts, then I agree that's the reason and results people should look at and then also look at the consequences of the result of the change. Either way it should be officially stated and documented. I hope you would agree on that. How WG goes about fixing the problem could be multi-facilitated and I would say would need to be closely tested and then validated to the community (in my opinion).

[near pond]

Blowout matches are caused by a few things. There is no one size fits all solution, but balancing the teams by winrate is a start

[unique lance]

I agree people don't understand or have limited understanding of how it actually works, but instead of repeating over and over again, why it so bad to document it and place it on a webapge and refer to the link for all to see. Both are important as both are involved in how a "Battle" is setup. Though, they are separate both are important to the players and certain concerns can be discussed based on either the specific separate part or looked at in totality.

[near pond]

They’ve told us how it works. If need be i’ll put together a watered down explanation and have BFG double check it or something……..

[unique lance]

Assuming what you say is true "Blowout matches are caused by a few things". You don't state what these things are, so it's hard to agree with that statement. I agree if we are solving the problem of "Blowouts" this is not easy to solve. As I think there are many reasons for a blowout, I would be happy to list them, but trying to not type a novel. However, the point I want to make is even if you think it's a start is WG okay, is the Community okay, with changing the primary goal of the game, by saying to players go win the battle and then (as a result of said fix) punish some players for doing that. In my opinion this is a terrible result and has profound effects longer term on the game. I wouldn't simply say give up on trying to solve the stated problem, but just this is an option for the solution, as I'm undermining the game. I would consider and offer other solutions for solving the "Blowout" problem.

[unique lance]

No watered down version. Are you WG staff?

[unique lance]

I disagree this thread is a thing because some people are experiencing negative effects from the change that was made.

[near pond]

1. lack of recovery options
2. team sizes causing an acceleration in losses if one team gets an early lead
3. the game’s older mechanics rearing their ugly head

How exactly are better players being punished? What is the punishment?

No. The game is not attempting lower your winrate. What it is doing is creating an average winrate with the enemy team using the players of both teams

[unique lance]

More experienced players win more games, because they have knowledge about the game and are able to perform better in battles, this results in higher win rates. There are also other factors that result in high win rates, like platooning with experienced players, etc. However, those that have achieved higher win rates are being effected by this change. Many have actual date, as do I to show this over 100s and approaching 1000s of games now.

[unique lance]

I don't disagree that the game isn't directly attempting to lower your win rate, it's forcing a higher win rate player to play along side more lower win rate players to overcome the opposing team that has more high win rate players, which is "stacking" the game against said first player, so to speak. AGAIN I ask you directly are you a WG staffer?

[near pond]

No i’m not WG staff. But i am apparently the only one in this thread who actually understands the change outside of WG…….

[unique lance]

That's simply not true

[near pond]

Uh huh

#1405193097079689458 message

[unique lance]

Do you not believe that players are being negatively impacted by this change?

That would be awesome if that was a link to the documented way this all works! ROFL

[near pond]

Winrate maybe

But the game is not stacking the odds against better players

[unique lance]

Correct there are no odds involved, you are correct it's not chance it's a forced "balancing". ROFL

I would agrue that win rate is truly the only thing that really matters when it comes to stats (in my opinion).

[stray star]

loud incorrect buzzer

[near pond]

Oh no…….wow yeah you were never going to understand this

[unique lance]

Because this is the primary purpose of the game. Though I appreciate your opinion I do disagree this is the correct solution to pursue to address the "Blowout" question.

[near pond]

There’s a difference between best and least worst

[unique lance]

I didn't say there wasn't and I'm not going to get into a discussion about the "other" stats people use for this game, as I understand all of them very well and this isn't the thread for that. I simply made the point I did, because this solution under roads the primary goals/incentive of the game. Have a good day Cheese.

[unique lance]

I probably could have stated that better, it's the only stat that really aligns with the primary purpose of the game.

[modest ivy]

maybe not a good idea to release something similar with this much backlash coming from ww2 playerbase

i for one have not played the game since the last session i had. wotstars is down right now, but it was something like almost 4k wn8 with a 40% WR

if i have to play cw then fine i guess... boring game mode. id rather play ww2 and finish the 3 tech trees i am close to finishing

needless to say, i have put most of my time and money towards wot pc as i just dont see why i would spend any money on this game right now until something is fixed with the mm in ww2

[near pond]

That my dude was for you

2s imma put something together. Clearly someone has to.

[haughty terrace]

True and some are experiencing a positive effect from it. So, which way does wg go? That's the real question. You will not be able to please everyone, but so far I have not seen a compromise from the person who has created the thread. Only go back to the old way.... Give wg some ideas on how to create a better solution.

[fleet elm]

There is no better solution. WG wants to handicap good players (and boost bad players) in the name of “better gameplay.” Meanwhile, we don’t want to be handicapped. There’s no middle ground. The best we can hope for is that the new matchmaking doesn’t do what WG wants it to do and they revert it

[lament mica]

Revert to what? Because it hasn’t exactly been doing what they want for awhile which is why they made the change. Random was looong ago and they probably could not nor given everything built on go there. Not sure players would want to tolerate the lack of other development.

.
WG would have saved themselves some headaches if they would have called it team division changes or something different.

[fleet elm]

No. They didn't even specifiy the matchmaking changes and we all noticed how much less we were winning. We just don't like getting constantly handicapped

[fleet elm]

The system isn't particularly important as long as it doesn't handicap anybody, especially for improving at the game. It's not a system issue though, it's a "WG is ok sacrificing some of their playerbase to improve gameplay" issue

Which sounds fine unless you're being sacrificed

Which is why the only real course we have here is to hope it doesn't help and gets changed back. Honestly a word one way or another on several issues would really help decide if I uninstall for good

[coral zephyr]

I have not seen anything to think they are going to reverse course. Or even considering it at this point in time.

[bleak oriole]

This ... is not true. It was discussed on-stream some time before the season launch - I want to say 2 week, but it's been a while.

[coral zephyr]

Not everyone watches Streams or if they do they don't watch every Stream.

Since the Friday Stream changed times, I can't hardly watch them. They don't line up with my lunches any more. I don't think I saw the Stream where they talked about the MM.

[fleet elm]

Mm changes were mentioned. Vague “technically true” but somewhat misleading statements were made. Nothing was described until we started complaining

[cursive tinsel]

Was going to say, all I remember hearing was, the first iteration was being tweaked. I what regards was not informed in any articles mentioning it and only found after it being pushed som much to get an answer.

[bleak oriole]

"We are going to use the Win Rate in order to assign people between teams" is what I remember being said

[cursive tinsel]

All im going to say, just saying WR, doesnt mean much consider the spread of which it could be taken from. Also, what other said, if you miss the stream and its not published else where, that its good for anyone, because it lead to situations like this where it turns into a he said, she said, they said. Hence why backing it uo is a necessity to point to off the rip or actual put line the exact areas.

[fleet elm]

Yeah there’s a big range for what that means. On one end theres making sure no more than 9 of the top 15 wr players in the lobby end up on the same team. On the other (what was implemented originally) you have attempting to force even teams by pairing the best with the worst.

And the whole “not sbmm” nonsense was very misleading when we had no other info

[bleak oriole]

we didn't implement pairing "best with worst" - there is zero guarantee that the actual worst player in the match is actually on your team

[fleet elm]

I know how it works. It’s a more general statement on the best players getting the worst teammates compared to everyone else

It’s precisely the system that we didn’t want. A proper sbmm (playerbase issues ik) would be preferable

[fleet elm]

It’s slaved to the old system bc it’s been stress tested and was super unlikely to break. Pairs mm has limits when you start sorting

[unique lance]

It was not discussed, it was told to the players what WG was going to do. Let's keep it real , there was no feedback asked before this change was implemented.

[fleet elm]

The feedback was realistically a decade of mm and game length/closeness complaints

[fleet elm]

It’s all moot though. If y’all at WG think this is positive for the game, that screwing us over helps everyone else enough, it’s gonna stay regardless of our feedback

[unique lance]

This is not the solution to address "Blowouts", they still occur and will continue to occur until we truly address the reasons for the "Blowouts". However, if WG is trying to solve some other problem then I can't speak to that. It's hard to get them to be constant as to why this change was made.

[unique lance]

It was put in without our feedback and just might stay regardless of our feedback.

[unique lance]

Sorry, I should have been more clear. It was announced during a stream that they WG was going to do this. However, it was not asked or feedback wasn't collected prior to this change being made.

[unique lance]

I did comment during that stream that is was announced that this was a bad idea and that it underminds the primary goal of WoTC, to win the match, because WG is directly stacking (for a lack of better way of saying it) the game against certain (not specific) players.

[autumn light]

the reasons stated were:

• long standing player complaints about "stacked" teams and "unfair balance" of teams
• To reduce "blowout games"
• To increase average match times

[unique lance]

Thank you Rudolph for being consistant, that is exactly want has been stated here before.

[autumn light]

Basically players have for a very long time pointed out it was very possible to get into a game and say 6 players on 1 side were all very high win rate and the other had 0. Or similarly the other way around of 6 on 1 side were all very low win rate while the other side didn't have that. Under the modified system of sorting the pairs those sort of things don't happen in terms of the "stacking" extremes. And there was reductions in blowout games and increases in average match time, BUT they were not as large/impactful as we'd like. The tweaks were then made, and no more info has been released on the impact/success of that.

In theory, with how it's meant to work (aka without platoons messing with it or some other outside factor), it should be highly improbable that you end up with such disparity across the teams.

[unique lance]

Wait let's take the first part of this "Basically ... other side didn't have that". Do you have stats on how often this occurred?

[fleet elm]

And was at least partially addressable without actively shafting good players

[autumn light]

me personally, not to hand, but AFAIK yes. But it is also just a mathematical certainty with standard distribution. Basically what the change aimed to do was take a standard bell curve that meant any combo of "all 15" to "none" of the "best" of each pair being on 1 side, and shrink that to be that there is as close to an even split on each side

[unique lance]

In other words can WG quantify the problem? Yes or no? Even if it happened all the time how often did it happen to a specific player.

[fleet elm]

Why not just do 7/8 down the middle and be done with it though?

[autumn light]

^i did a very low effort paint visual of what i mean back here

[autumn light]

In a perfect situation, it will effectively do that. But basically the current system can't do exactly that really as it would need to basically track the info of all the pairs before sorting and SPG can correct me if i'm wrong, but it doesn't/can't do that atm.

[unique lance]

Stop with the solution and talk about the problem first....can you answer my question above.

[fleet elm]

Delta rounds to exactly 1 or -1 for every pair. Solved

[autumn light]

I can answer in 2 ways,

First one, enough that it has been a constant complaint by players basically since day 1.
Second one, i don't have exact numbers and idk how much i can share of what i do know. But i think it's safe to say that it would follow a relatively standard distribution curve

[unique lance]

1. Not what I asked you...

[autumn light]

again SPG would have to correct me if i'm wrong, but to my understanding that would require a new system/more coding. Basically what you're on about is having an almost secondary number system AFTER the points value is worked out. And i think it was said that changes made were all tweaks within existing code and no new code.

[unique lance]

You can make all the point changes you want you aren't going to eliminate the problem your trying to address, because your not addressing all the caused for the problem.

[fleet elm]

It’s no new code. It’s within the current system. It does the delta comparison for each pair. If that number is positive, set to 1. If negative, set to -1. Then it builds teams exactly the same as it does now

[unique lance]

There's multiple reasons you could have a high win rate...

[autumn light]

i think you are misunderstanding, it has nothing to do with a single players win rate

[autumn light]

To reword i guess, the issue was the existing "sorting" system relied purely on battles played to work out the delta. This didn't factor things like win rate and so people would complain that the win rate spread across the teams would often be very uneven. This is mathematically true and certain, but in theory would also balance out. AKA if you have 6 "very high" win rate players on your team, statistically you'll eventually get 6 "very high" win rate players on the other team. Effectively this is still what happens, it's just the extremes are now far narrower. The complaint brought by the OP and people like nate is that statistically by being having a very high win rate they will create a larger delta more often, meaning their "average" will be skewed to having "worse" overall delta balance more than before. Correct me if i'm wrong @fleet elm.

The issue of the system sorting people and resulting in a quite "random" distribution of win rates across the teams was indeed addressed by the changes

[unique lance]

1. You actively work with others to win the battle (such as mic usage and platooning) 2. You have more knowledge of the tanks capabilities (pros and cons) then others 3. You have more map awareness about the map the battle is occurring on. 4...I could go on, but I've given enough to make by point.

[autumn light]

If that number is positive, set to 1 This bit i would imagine is new code, and as much as i'm not saying it's 10s of weeks of work... i have 0 clue how much work, it is new code. That's why i said that what happens now is effectively similar, but not exactly that because that would have been new code. If that makes sense

[unique lance]

Yes, I also was provided feed back the old system should have been pulled out (number of battles doesn't mean any thing, because it does speak to knowledge about the game).

[coral zephyr]

That is not an entirely true statement. A single player can create the largest delta possible. Subsequent pairs would be assigned as needed to counter that delta. It could take multiple pairs to counter a single maxed out delta.

As such the lower performer of those pairs gets assigned to the person who originally created the max delta. Every pair needed is another chance for a really bad player to be put on the good player's team.

[autumn light]

for clarity, i mean the aim of the change had nothing to do with addressing a single users win rate. The context was about the aim of the change

[unique lance]

First we reall need to understand what needs to be fixed, if there is a real need to fix the "Blowouts" because that's 2 of the 3 points you said WG is trying to address.

Longer battle times are easy

[coral zephyr]

Yet that is what it has done.

[fleet elm]

I more mean that no new functions would be written. It’s practically the same as capping delta for a pair if it’s too large. It should be a trivial code change

[unique lance]

Just multiple all the tanks hp by 10 or any factor you choose and this will result is longer match times...Again I don't this that needs to be solved, but there you go.

[autumn light]

but the question/discussion was about the aims, and i am stating the aims. So to say it's not an entirely true statement isn't accurate 😅

[fleet elm]

Yes outliers will wind up with worse teammates than before pretty consistently

[autumn light]

i can't comment to the work required, all i can say is it was stated that no new code was to be done and any changes were simply within the existing system.

So that's why i say the "tweaks" resulted in a very similar outcome to what you are saying would be ideal, just done a different way within the existing system

again, in a perfect situation... Platoons can mess with it

[fleet elm]

Any change requires code change. Within the existing framework, if there’s a function to cap delta, capping it arbitrarily low would be the same as +1/-1. It’s a variable change, not a code change

[unique lance]

If you don't really acknowledge what the root problem is you won't solve this, again you can do all the point calc. you want it's not going to solve the "Blowouts"

[autumn light]

we didn't say solve, we said reduce. And it did do that. The changes achieved exactly what we laid out to some extent, some to more of an extent than others.

[unique lance]

That's the illusion that you have created!

Eventually when everyones at 50% win rate, you'll be right back to the increased "Blowouts"

[autumn light]

• Teams will no longer get wide disparity in win rates across the teams under ideal conditions
• Blowout games were reduces
• Average match times were increased

3 aims were set out, all measurable, all were achieved to some extent. Although we openly stated for the latter 2 we would have like to see more change, and for the first we made tweaks to improve that while addressing feedback in the process

[coral zephyr]

That's not am entirely true statement either.

But eitherway, I am a results oriented person. The intentions are often irrelevant if the result ends up bad.

[autumn light]

due to the pair based MM not caring about win rates, it won't be very successful at trying to force win rates down to 50% (and it isn't aimed to do that either). Even the people who hate it the most here can probably admit they don't suddenly have a 50% win rate or anything that close - over any longer/larger number of battles.

[unique lance]

Wait bullet one is different then above, but I would agree your driving everyone to 50% win rate which reduced disparity in win rate, but NOW you won't even be able to determine better vs worse

[fleet elm]

Debatable that it addressed feedback. The main feedback here was “I don’t want a worse team bc I have a high winrate”

[unique lance]

I didn't say that, it's the result of what your doing!

[fleet elm]

Winrate distribution is narrower but it’s still easy to win >55% of matches solo

[autumn light]

i mean i think it's fair to say that most in hear said it got "better", even if not "fixed". So that's all i meant by addressed, perhaps "attempted to address" is more fair/accurate?

[fleet elm]

I honestly can’t tell if it’s better. Queue cancelling in the more extreme version of the mm has given me better results than playing normally so far

[unique lance]

It's definetly not better for players with 60+ winrate

[coral zephyr]

How long do you wait to re-queue?

I need to move to that or move on.

[autumn light]

fair enough, i'd say anecdotally the feedback in here seems to suggest that it is certainly better post tweaks than pre. So that's why i'd say attempted to address at the very least, as the outlier player feedback was directly considered into the tweaks

[unique lance]

I also believe players with <50, don't even notice a difference

[autumn light]

better in context, better than before the tweak

[fleet elm]

No, the original version of this mm was more extreme. It’s all worse than pre change

[fleet elm]

3s in queue at most. It’s a pain

Cold war is irrelevant. No changes there

[autumn light]

^

The changes were WW2 only

[fleet elm]

We’ve listened to feedback and will only be torturing you twice a day instead of 3 times. Technically better but not really addressing the issue

No. Not applicable at all to this discussion

[errant saddle]

Fine I'll just say the mm is fine then since you wanna go that route.

[fleet elm]

It’s likely gonna be a “feel” thing. Only really noticeable in large scale data like player retention

[unique lance]

I will happily provide feedback when WG is wanting to really understand the root of the problem, until then this conversation with keep going in circles and more experienced players will continue to leave, resulting in the continued slow death of the game (sad to see for sure).

[autumn light]

Although i understand that with the limited information available it's hard to make much judgement beyond "feeling", the change was significantly more drastic than that. The tweak has resulted in everyone's delta "imbalance" being MUCH closer to 0, even outlier win rates players are likely very close to 0 now.

[coral zephyr]

Shorter than i was trying. Still might be better than wasting an entire match.

[fleet elm]

The issue is quantity of worse teammates due to how good a player is. You’ve made the number smaller, it should be 0 on average

[coral zephyr]

Do you not see the inconsistency? Our anecdotal feedback is dismissed. Yet the anecdotal feedback that agrees with staff and the change is used to promote the change.

[near pond]

All I’m getting from this thread is players don’t like consistently being against others of a similar skill because it lowers their winrate

[fleet elm]

Total pain but I was doing like 60% in the technodrome vs mid 50s not doing it over 200 battles each sample

[fleet elm]

I don’t like being forced to play with worse teammates. It’s not about who’s on the other team

[unique lance]

I have hard specific data that my win rate is dropping, but I guess I was just always on the other side getting the benefit of the "Blowout". The good news is when your able to platoon this has less impact as you get atleast a few others taht have some knowledge of the game.