Increase Grandmaster rank to 30,000 rank points

[spark perch]

In many games Grandmaster is a title only given to the most skilled few. In chess for example only .5% of tournament players are Grandmasters.

I've noticed many players who have grandmaster rank who frankly seam to be intermediates at the game.

I've heard many other top players share the same sentiment that the ranked system is too watered down with players given high ranked for a group who's skill is highly deviated. In other words the difference between a novice and a intermediate is very low while the difference in skill between grandmasters is often very high.

To make ranks more true to the skill level I believe that we should up the skill points for GMs to 30,000.

That might also involve a shift of the rest of the ranks up slightly too to make everything fit.

As of right now 360 people on the leaderboard are above 30,000 sp and it isnt even halfway through the season.

because the grandmaster rank is so low players are forced to regain so much sp over a new season that they already deserve to start at.

There is no point in forcing grandmasters to regain all their hard fought rank every season to show their true rank.

By raising the rank points to 30,000 for grandmaster this problem would also be fixed.

[dense kraken]

I definitely agree that the term Grandmaster is used far too loosely in this game. Comparing it to Chess it's nothing close.

It's also extremely arbitrary since someone can be amazing at Classic Fixed and have way less SP than a very average EU Caps enjoyer.

So until SMG revisits continuing to improve the ranking system I don't know how legitimate any threshold is. Or incorporating the Sabr system which we have found to be a far more representative data set for actually measuring ability in our game.

That said relating the term Grandmaster to a certain percentage subset of active players does seem like a wise move. Perhaps rather than a different arbitrary floor for GM the threshold fluctuates based on that percentage?

I would be interested to see where it goes long term. I do agree with your fundamental point Truth. I would question how many problems this change actually solves vs continues to perpetuate the illusion that we have a functional ranking system for Risk currently.

[exotic bloom]

After thinking about this, what about a system where the season resets “GMs” back to the lowest RP that a GM has. This would obviously need a system where there’s a fixed amount of GMs (maybe top 500 players?). So if the 500th ranked player in the world has 33,000 points everyone above that would be reset back to 33,000.

You could either reset everyone below down to the current base GM or reset them down to master. This could encourage more GMs to be active as well. There’s a whole lot GMs like myself who don’t play a single ranked game in a season.

I don’t particularly love the system of making GM a certain percentage of active players. There’s a lot of players that only play casual, single player, pwf, etc so I’m not even sure how you would define that. I’m also not sure that just making GM a few thousand more points than it is now would accomplish the change you want. If you can get to 26,000 you can get to 30,000 at about the same difficulty in the current system.

[brisk mulch]

Because of how elo works, gaining elo gets twice as hard for each 10,000 skillpoints you go up.

If they pushed FFA GM up to 36k and 1v1 GM up to 46k, a lot of people on the edge would feel the burn.

Of course it would be best to accompany these changes with other improvements to the ranking system.

[river fiber]

Arguably 26,000 may be much more difficult to reach if you have fixed settings you can only play similarly ranked opponents with and there's only 3 or so automatch modes for joining

though the "same rank" may actually make it easier, if it's preferential settings, but it'll mean players will get better at the settings because they're the only ranked ones

[cold hill]

Well, if you increased the 1 v 1 GM point total to 46,000. That would currently be a total of 7 Grandmasters, that might be a bit of an “overcorrection.” As I am sure (and agree with Truth’s original point), but I also don’t think he wants the Grandmaster total to be less than 10. That being said, why do rankings reset at all? IMO, it is a method that SMG uses to encourage more games to be played as it forces games to be played to see your name on “the list.” The point of an ELO system is to actual know what your max point attained will be. Well, to the casual player who isn’t grinding games, they will never know their ceiling. The top players grind for a season or two…. But then they drop off once they are done with “the grind.” Why reset at all? If you have a good product, people will play it. Why have a drop at all? If you want to ensure games are played, why not program to be “viewable” on the top list, you just have to play a game once every two weeks (or whatever standard you want to set.) if you truly want a season, why not make it annual?

[brisk mulch]

Plenty of people go over 46k, but there’s currently little reason to do it more than once, let alone every season, when it keeps reseting to 26k.

In Chess it’s far harder and rarer to hit GM, but they go by the mantra “once a GM always a GM”, and it’s a far more prestigious title within their community

[cold hill]

Yes, it is good to see you agree with what I wrote.

It sounds like you question the resets just like I do.

[brisk mulch]

Yeah, there’s too many season resets, making top ranks current too much about sinking in a ton of time almost every month, even if you were one of the most skilled players in the world

What I’d really like rank to reflect, especially for 1v1, is “if I played this person now, on their settings, how likely are they to win”

[plucky mural]

I agree that given the SP requirement for the GM title is raised to somewhere in the 30k area. Grandmaster should be a title that is upheld, not taken away every 50 days. Although it doesnt take away the GM designation, getting downtiered from 46k skill points to 26k might as well be.

[cold hill]

If the community wants the Grandmaster title to be revered (as it should be). If there were no season resets ever, and instead a mandatory playing requirement to keep your name viewable on the top 500 list. Couldn’t the GM title actually be set at the top 5% of actively playing players? Once you had that determination, let’s say the top 5% was at 30,000. Then the divisions of the other ranks be divided up equally. Just a thought.

[exotic bloom]

I don’t think SMG is going to go back to a system where there’s no resets. As much as the resets are annoying, it never being reset was worse. Now I’d have no problem with longer seasons but almost every multiplayer game has ranked leaderboard seasons resets and it works well.

Also what does top 5% of active players mean? There’s a huge community of players that plays nothing but single player or casual. Then there’s huge sections of the community that basically only play either 1v1s or FFA. How do you draw that 5% line? Also 5% of players being a grandmaster is quite a lot. I’d put it closer to 1% if you want it to be “revered”.

[river fiber]

The additional benefit of resets is that it chunks games together (so when we're archiving games we're limiting the total number of games being grouped together)
It makes it easier to revisit / pull games from archive to check stats or whatever else - basically keeps the game data *bite sized"

[plucky mural]

I feel like its not THAT difficult to reach grandmaster, its only when you get to 30k and more then it feels like a challenge but by that point there are no ranks up there

I think increasing all of the rank requirements by 4k skill points would make it better.

Its not a permanent solution but definitely satifys a lot of things for now at least

Increasing the rank requirement will also disincentivize skilled players grinding on new players for skill points as they have a deeper cut if they lose.

[cold hill]

I have to explain what 5% of active players means to you? Seriously? Well, okay… here goes… active, to me it would be defined as players who are playing games. So, for the current season, you evaluate how many players are playing 1 v 1 games and determine 5% of it. You want it to be 1%? That might actually be reasonable. It would depend on the numbers. Does anyone actually know how many 1 v 1 players are actively playing games in this season right now? I bet SMG knows that number. Same thing for FFA, determine how many active players who are playing and see what 5% or 1% actually is before making a determination. You may be right that 1% is a better reflection of having Grandmaster being “revered,” then again… 5% might be better. Whatever it would be changed to, it would be an arbitrary number so it would be subjective in the end, just like someone made the subjective decision to put it at 26,000 currently.

[exotic bloom]

Thank you for being condescending for no reason. I asked for clarification on active players because how you define it is very different than how SMG currently defines it. Currently you don’t even have to play ranked game, you can play one single player game, and that counts as being active for both FFA and 1v1 so you don’t lose your rank.

Additionally if you are just talking about 1v1s I would say like 90% of “active” players are novices. And most that I play have played less than 10-20 games. That means most players play a game or two of 1v1s quit and don’t play again. Is that really an “active” 1v1 player? I wouldn’t say so. If you ask me I would say you would need to cut out novice accounts for the percentages regardless of if they are “active” enough.and I think that goes for FFA and 1v1. There’s too many of them that play a few games and never play again.

Lastly if you are going to say determining whether it should be 5% or 1% depends on what the number is means you care more about the number than the percent. So why not just have it be a number? Maybe the current number of GMs is already at 1% of players. Maybe it’s 10% but does that really matter? We all agree there’s too many so why not just make it a number

[cold hill]

Well, we all believe there is a problem. Maybe there is, maybe there isn’t. Truth wrote that there are currently 360 Grandmasters above 30,000 (although he didn’t specify if that was 1 v 1, FFA, or both combined.) Well, 360 compared to what? In order to actually know if it is an accomplishment, it would be nice to know how many grandmasters are there compared to active players (actively playing ranked games or however it is defined). Sterling thought it should be over 46,000 for 1 v 1, and that would have been 7 players total. Well (imo) 7 is a little lite, but it’s a subjective standard so everyone is entitled to their opinion on what the right number would be. As for defining active players, I am not Merriam Webster, so my “definition” isn’t the one that matters. How SMG defines it is what it is. If they want people who attained a rank to maintain it by playing casual games, that is their choice.

[scarlet stag]

Why are season resets necessary at all? (Hint: they aren’t)

If the goal is to have opponents start from an even starting line for a season long race for the top then why not tie leader board position to performance since the start of the season. Comparing my resent results to your resent results has no requirement to involve our positions on the rank ladder at all.

A players rank and their performance in what is essentially a many weeks long competition (the seasonal leaderboard) are separate things and should never have been mixed together in the first place.

[exotic bloom]

I'm confused. How would a player receive a rank without it being tied to a leaderboard? I think I get what you are going for (A GM in skill is a GM regardless of whether they play on the leaderboard that season) but how do you determine skill level without the leaderboard?

[scarlet stag]

The seasonal leaderboard position would be based on wins loses since the season start and movement up or down the leaderboard could take into account rank points gained or lost and opponents rank in games played during the season. Rank position itself would be a separate numerical score denoting your rank level only. You could only lose rank by losing games (or in some systems by long periods of inactivity).

So in some sense you could have a worldwide permanent rank leaderboard that would show you who has achieved the highest rank, maybe it would be a risk GM who stopped playing 3 years ago. But the seasonal leaderboard would be based on performance since seasons start. And relative gain or loss based on each individuals starting point. No reset necessary.

There can be some argument for an algorithm to process the data to correct for differences in difficulty of climbing from different rank positions. But the best system would involve peers playing peers (so for best results SMG would have to institute camouflage mode as well allowing repeated play among high ranking players) and wins and loses tallied up with relative ranks between opponents in any particular game taken into account.

Then the number of games you won and the difficulty of each of those games based on opponent strength would be used to compare you to every other player participating in that seasons leaderboard. If a GM sat the season out they would not show up on the seasonal leaderboard, but they would remain a GM and they would still show up on the “universal leaderboard” which I’d prefer to simply call “the list of ranks”.

[brisk mulch]

I’ve seen 100+ 1v1 players reach over 46k over the years, but there’s no reason to sustain it currently, when they’re being pushed back to 26k so frequently, it takes so long to rank up, and there’s no reward for ranking up again.

There are currently around 650k players on the monthly battle point leaderboard, which is a good proxy for active players (those who’ve earnt battle points this month)

Of the 800k ranked players this season, 1800 are FFA GM (1 in 444), compared to Chess where there’s 20 million active ranked players on chess.com and 2000 GMs (1 in 10,000).

This would be like having only 80 FFA GMs for Risk.

Ideally I’d want to see something somewhere in-between Risk’s outlook of “GM should be easy, because it’s having fun that counts”, and Chess’s “GM should be hard, because it’s winning that counts”

https://www.chess.com/forum/view/general/how-many-active-members-on-chess-com

[river fiber]

The comparison between chess & RISK is still very loose at best though, primarily the massive luck factors that apply in RISK. If we set similar parameters / point scales to Chess you'd just never get anyone reach GM too 😄

It might be better to look for inspiration from leaderboards that contain a larger luck factor

[plucky mural]

Wait if we set the ranking system like as hard as chess, i dont believe it would be impossible to reach GM

and though the luck factor is there and very obvious in some matches, in the grand scheme, good players still manage to go positive on skill points

if we completely mirrored the chess system, GM at 30k, master at 22k, expert at 16k, and below

I think its overall a big help

Since the trend as risk gets more popular and gets a larger playerbase means, that the average skill level will increase.

So the only obvious solution is to in turn raise the ranks

[brisk mulch]

The elo requirement for GM doesn’t need to be as high as chess (2500 elo, something rediculous like 150k skillpoints?).

But it could be higher than it is now

[plucky mural]

Yeah for example, i had to go digging aorund, but this is the leaderboard from a year and a half ago, before i told them it was broken, and you can calculate the average rank using this data. Point being is that SMG can look at their leaderboard now and compare it to this one and see that it has increased. Obviously you will have to interpolate and convert it to the modern system if its possible.

but i assume the result will be that the average rank is higher than a year and a half ago

[plucky mural]

As of 2022/08 ish

im allowed to leak cause you cant do it anymore

and its old

[brisk mulch]

26k reflected the top level of play back in around 2015 when it was introduced. At the time Risk wasn’t on PC yet, it was impossible to turn off camera animations, surprisingly little was known about how dice worked in balanced blitz or caps, tournament scene was tiny, and Risk youtube was a wasteland

These days the game, the players and the community have come a long way, so the standard required to be a top level player has increased

[plucky mural]

It has every risk player ever played too, so you can also measure other things such as engagement.

[plucky mural]

Thats a very good point

[brisk mulch]

Wow, that’s a lot of data. How did you make this?

[plucky mural]

Scripts

Web scrapers

If your curious enough you can do anything

[plucky mural]

Also when i shared it, SMG instantly set the leaderboard viewable to only 10k people

I made the leader board in an effort to convince them to bring back the ability to see rank in game lol

I thought that they would simply have no choice but to implement it since everyone would begin to use my method

But i was a doofus and realised they could just simply take away the leaderboard

So yeah you could say that i ruined it for everyone, or you could also think of how many people about it did it before me and didnt tell anyone and got an advantage for a while.

[plucky mural]

But then again back to the ranking system, i think the spreadsheet will help a lot into deciding how much to increase the rank level down to the specific amount, as you mentioned that it was introduced in 2015, which just goes to show how it definitely needs a change

[cold hill]

@brisk mulch thanks for posting the data, it makes for a convincing argument, and I agree with your conclusion. GM should be harder to attain, but probably not has hard as it is in chess. @scarlet stag you have presented an interesting concept to mull over. @plucky mural I’m not sure if I’m doing something incorrectly or not… but I’m on mobile, and I am unable to open the data you provided. Based on the description given, it would be interesting to see.

[exotic bloom]

The data that Centauri provided is the old leaderboard put into a massive spreadsheet. So you have every player put in rank order with their position, name, link to their profile on hasbrorisk.com, and the number of SP they have.

It’s essentially an easy way to quickly look up anyone’s profile in the game (at the time) so you know the skill level of your opponents. This obviously defeats the entire purpose of SMG hiding profiles in ranked (but it’s 10x worse IMO because only PC players that have very specific access would be able to use it, and on top of that your opponents would have no idea you know their ranks)

[cold hill]

Oh, I think people with PC’s and multiple monitors check the ranks of their opponents already (if they are ranked high enough to be viewable)… well, at least I suspect it for FFA games or 1 v 1 ranked.

[brisk mulch]

Seems weird that I could be double GM, 26k 26k, and it still be possible for people to have more 1v1 skillpoints, than my total skillpoints

[exotic bloom]

That's me currently lol . And 2 others. And people do it in FFA every season as well

[silver lily]

I play from a phone and have only reached master level then down a level or two. I consider myself good and probably would've won 25% more of my loses if I was on a PC and didn't run out of clock.

[brisk mulch]

“Ask not for lighter burdons but for stronger shoulders” - Ancient Proverb

I think the solution to mobile being slow, has always been to speed mobile up, rather than to slow the rest of the game down

[silver lily]

It would be great to speed up mobile. I would probably be the best in the world then.

[cold hill]

Sadly, mobile players typically need 90 second turn timers to some on PC needed 60. Not much can be done about it until other changes are made to make it faster for mobile players (like us).

[scarlet stag]

There may be a solution here not in speeding up mobile play directly through internal corrections, but rather through the user interface itself.

If SMG would optimize click box sizes, correct the mobile slider interface and make several other small tweaks including simplifying the attack move entry method (in some way similar to the click enter enter method for PC) then with effort mobile players would be faster on their own without the need to remove whatever delay is built in between clicks (though I wouldn’t object to that being looked at as well).

I find that the primary problem in mobile speed is the additive error of many many small delays caused by click failures requiring multiple attempts for nearly every small action.
While none of this would match the raw potential for PC speed it would create a fair playing field where the average fast mobile player could compete with the average PC player on speed.

If these changes where made then the limiting factor on mobile speed would become technique and that is something that serious players couldn’t complain about. Optimizing your technique on mobile would be just as important as getting your click rhythm right for ultra fast PC movement.

While the speed gap would never fully close. Mobile would stop being unnecessarily handicapped by such an avalanche of small unnecessary “glitches” that break any attempt at actual speed.

(My mobile experiences are primarily on an iPad Air 3rd generation but I have also played on an iPhone6 in the past. Curious if other mobile users find the same problems with click entry failure on their devices.)

[cold hill]

iPhone 12 here… the problems are still the same.

[placid wraith]

In my opinion, I think it'd be cool if players in the top 500 had a new rank.

Maybe like this

16,000 - Expert
21,000 - New Rank
26,000 - Master
Top 500 - Grandmaster

or

16,000 - Expert
21,000 - Master
26,000 - Grandmaster
Top 500 - New Rank

[brisk mulch]

Yeah, it fits to have grandmaster at the top though. Would just sound quite slapstick and inflated to put another rank above it

[river fiber]

If taking the "titles" for sub sections of ranks, could go with:
Expert I, II, III, Master I, II, III, Grandmaster I, II, III
(And unofficially we can call grandmaster 1, basic grandmaster 😄)

Segregating a rank based on number of players is a good route overall as it is only really changed by the number of players that can play at that level

[scarlet stag]

Unless that rank was titled “Top 500 Gramdmaster”

[valid sierra]

how about a god role for the top player (:<

[dim path]

How about we rename the top 500 to Grandnoobgrinder?

[brisk mulch]

If there needs to be more than one type of Grandmaster, the threshold for Grandmaster isn’t high enough

[cold hill]

If there were sub sections within each class, perhaps on the season resets, they would only be “reset” to the lowest of their sub-section instead of dropping the entire distance.

[river fiber]

and having it percentage based / proportional to the number of players - makes it more considerate than simply any fixed value to reset to

I'd still like to see all classes being percent based on points - so you don't have fixed brackets, instead they change as players abilities change

[plucky mural]

The 1st option would only work if they changed the way season reset works

[cold hill]

Would it be better to do it the way you listed, or the reverse? So the title “grandmaster” would be reserved for the highest level of achievement, and Grandmaster III would be the initial entry level GM? Just a thought… no big difference, just a personal preference.

[fading tulip]

What about adding a new rank below gm, chess has international masters so what if we made it master 16k, international master 26k, and then a further to percentage or number cut off for grand master?

[river fiber]

Doesn't really matter whether it's reversed or not, as long as it's clearly explained when it's added (the benefit of 1 < 2 < 3 is that it could have levels added without having to completely overwrite other stuff, so more from a "keep it simple, future proofing" to make future changes have minimal disruption/confusion too

[fading tulip]

Also, I think that the resets are important overall, that makes it so players keep having to earn their rank and understand how the game is played in the modern day. If someone from 2+ years ago played a modern tournament they wouldnt have as good of a knowledge of today's common settings and how they're likely to go down, just watch some of the older YouTube videos. The reset means players have to be playing at a gm level in the recent game.

[fading tulip]

I actually kind of like that idea, I play another game with 5 different categories and each category has 5 rankings (I - V) within it, I being the top within the category. That could be usable for risk and could allow us to expand upwards a few thousand points. In game it could be represented with the Roman numeral after the title as I think that would look better than just typing out the number or something.

[cold hill]

KISS method. Isn’t that the patron concept behind all programming. Avoid all future issues when planning!

[river fiber]

Technically not my area but I try to do the same with design of @formal spruce - more important to ensure all of the original info (timestamp of use, id of user/channel/role, etc) is used, regardless of how it's presented

still has the same basic topic of creation of threads that it was initially designed for 😄

[plucky mural]

I like the idea but a different name is needed as the designation “international” doesnt make sense

[fading tulip]

Yeah fair

[cold hill]

Well, I’m open to a different term, but, there are many of us who play risk who also play chess. In the chess world, the IM designation makes sense. It’s a step below grandmaster.

[plucky mural]

I know how it works. There are no buts. Risk world is not comparable to chess world

[cold hill]

I disagree with your opinion. It is a game played across the world. Chess ranks their players as does risk. Their top rank is called a Grandmaster. Both games players use strategy to win. Sometime people make blunders during the game… oh, there are tons of comparisons. You may not like it or agree, however, yours is just one opinion. Last time I checked, and while you can try to make your opinion a fact, it remains just an opinion.

[plucky mural]

But everyone is an international player

So it wouldnt make sense to have international just for the master rank

Unlessss

We had a regional matchmaker

like we have the settings and filters

we could also filter it by eu/us/asia players

Making it feel more international

but apart from that the game doesnt feel international

that fact that you get a disproportionate amount of people from bhutan

[cold hill]

Are you joking here Centauri? I sometimes miss humor… so I just want to see if you are messin with me. You get a disproportional representation from Bhutan as those are many of Pete’s followers trying to emulate their Risk hero. I’ve personally noticed that there is are a disproportionate number of people playing from Antarctic . Heck, we even had someone create a Risk-idea to change the flag for the Antarctic!

[plucky mural]

Exactly, basically what i am saying is that in risk, the word "international" has no meaning. Meaning it is not an applicable name to use in a rank.

[plucky mural]

I think you misunderstood what i was saying

[cold hill]

It’s all good.

[cold hill]

Well, it is a war game, one could make the suggestion to switch to a military ranking system. The number one player could be the only 5 star general or something along those lines.

[exotic bloom]

That would be tough. Every country uses different ranks. Even just in the US the different branches call ranks different things. I guess if you just called it “5 Star General”, “4 Star General”, “3 Star General”, etc. And most have Colonel and Lieutenant Colonel. If you kept it simple like that I could see it working