#Completely remove ELO from the game

1 messages · Page 1 of 1 (latest)

vast rivet
#

serves no purpose. mjd added it for fun. warbrokers is not like cs2 or val or r6 its minecraft with guns and its not serious enough for elo the way its framed in this community and it makes people ragequit and just play for elo/stats and not for fun

#

pomps mod has effectively ruined the game

sour kiln
#

Can you elaborate a little more on the why behind this?

vast rivet
# sour kiln Can you elaborate a little more on the why behind this?

people get too attached to the number next to their name and either play so stagnantly that it makes the game boring or they just dont play at all and end up smurfing or going on other accounts as to not lose their elo. elo is only meant for balancing teams, it isnt even real elo. and also, seeing your elo change in real time like most can/do is another huge issue because it only strengthens the effects i mentioned because people die once and see their kelo go from 1960 to 1958 or something then just leave

#

people treat their elo like an artifact at a museum they just get it then leave it and never touch it again ^

pearl obsidian
# sour kiln Can you elaborate a little more on the why behind this?

1.people use guest accounts and mess up balancing anyway

  1. casual game - you can join and leave lobbies at a whim. why have a competitive metric (for kelo specifically). Countless times you join a game, people leave instantly because they dont want to lose kelo. Elo is just stupid.

arugment against: "its something to work for"
what about medals, ribbons, etc.. add more stats progression instead of this metric. It doesn't do waht its supposed to do (balancing) and only serves to make quality of games worse. People adapt extremely corny playstyles and play passively even in gamemodes that require aggression (Capture point, missile launch) beacuse they dont want to lose kelo

vast rivet
#

when youre not so caught up in your elo you actually start playing with more guts and play more alive and less stagnant and you literally just end up having more fun because you arent cowering and desperately clinging to a number that dictates your happiness in game

clear bone
#

Not everyone has to play that way though

sour kiln
#

So is this an issue with it being easy to see (i.e. via pomp's mod) or is this more broadly about it being visible?

#

Like would removing it from pomp's mod, but keeping on, say, the stats site, be fine?

vast rivet
#

it can still reflect wider trends

vast rivet
clear bone
vast rivet
#

who said anything about top players

clear bone
#

it is a pretty good indicator of skill

vast rivet
#

if you got the feeling i was talking specifically about top players cuz of the 1960 --> 1958 thing then change the 19 to 18 or 17

#

it still applies

clear bone
vast rivet
#

I am in favour of that, as long as it's not as easily manipulable as KELO

#

Even GELO is better

#

At least you can't farm gelo like you can farm KELO

clear bone
#

higher KELO makes your matches harder so you lose GELO

#

its kinda funny

clear bone
vast rivet
#

much harder

#

you can just smurf on an alt and open steam then farm

#

or open two tabs

#

to farm kelo

clear bone
#

A lot of the top kelo guys were caught doing that

vast rivet
#

by who? the ghosts that roam the servers?

#

70 active players bro

clear bone
#

if you wonder where they vanished to its their accounts got banned for boosting

vast rivet
#

u can get ur account zapped into the abyss for boosting?

clear bone
vast rivet
#

wym ur house is "ip locked"

native fossil
#

Atleast remove it or Kelo visably i say

#

Or massively nerf how much you can view it

#

updated once every day kinda thing

#

People trying to farm Gelo honestly sounds kinda amazing in a way

#

either people would go for gelo

#

or go back to KD

rustic thorn
#

I agree with Vortex. GElo is not better. KElo and GElo compaired to one another is a much better way to determine if a player is good at the game. Second best is probably KElo by it's self. Then maybe GElo or K/D.
KElo is hard to get much of if you're not a good player, simply by the way it's earned. While a person with 1800 KElo may only have the skill of an average 1700s player, you won't find one with the skill of a 1500s or 1600s player. It does take some level of skill.
From my personal experience, I know that GElo is simply trash on it's own. When I was terrible at the game and had a sub-neutral K/D, I still had a 2300 GElo. I've never had KElo that skewed.
K/D is obviously no good on it's own either due to the existence of BR, and of bot farming. Plus, it's something that changes incredibly slow if you've been playing for any substantial amount of time. So those that joined the game as a noob to these types of games, will take an extremely long time to have their skill truly represented by their K/D.

KElo changes per kill or death, (most of the time,) so it will tend to go up as one gets better, and go down if they get worse. There are people who farm it, as you'll have with any stat. That doesn't mean it is inherently bad. If it is taken out, people will start farming other things instead. The visibility of KElo is up to each individual player, and they choose to see it, which means that they want to be competitive in some way, and some choose KElo as the thing to work on. KElo can encourage growth in the game as well. It embodies an achievable goal, giving players something to look towards.

#

Heh, enjoy my wall of text. 🙃

deep night
#

While a person with 1800 KElo may only have the skill of an average 1700s player, you won't find one with the skill of a 1500s or 1600s player. It does take some level of skill.

Not the burst the bubble... but the only "skill" here is the skill to farm.

We've seen cases for players reaching high values, without much proper skill. One large example is a player that reached 2200+ with a skill that players reflected as "1600 range".

#

So forget an 1800s with "1600 skill"... we've had 2200 with "1600 skill".

rustic thorn
#

I meant without cheating. If including the potential cheating, yes, the skill required to get a certain stat will always be different.

#

You wouldn't say it's easy to win a game simply because there's a pontential for cheating.

deep night
#

No cheating involved, as based on the rules of the game. Just pure farming.

rustic thorn
#

Was this not the.. thinking "Lucky" guy that was cross teaming?

#

Maybe that was a different person.

deep night
#

pepe_what
This was Black Stone, who iirc was shown NOT to be lucky.
There aren't any rules about leaving matches, so if players leave to preserve elo and select specific matches that allow them to only gain elo, then even if it's a little slow at times, it will be farmed to a high level.
Even if a single death sent them back 2 whole points, if they leave and get 10 to 20 easier kills in a no-risk environment, they can reverse that... as long as they left straight away (which happens).

rustic thorn
#

But yes, camping will also gain one much KElo. That is an intentional playstyle though. I got 2300 GElo while doing so bad that I was actually helping the other team being I had a lower than 1 K/D.

#

So what I'm saying is: Generally, KElo, on it's own, is the best predictor of a player's skill, when compaired to the alternative stats given.

deep night
#

I get to be annoying? 🥺

#

you wrote:

...KElo, on [it is] own, is the...

rustic thorn
#

Heh, yep, I habitually do it's being it is more often used in my vocabulary than its.

deep night
#

the best predictor of what skill?
all it shows is how much you kill, and should tie back to your K/D as well

#

in WB, a skilled player as far as I am concerned... wins

#

so if they can kill, but they have no clue how to control map points and let the enemies destroy their team... doesn't matter if they... idk... sat in the back of Office and sniped down the tunnel for some kills

#

they can be 10/0, only gaining

#

but they lost

rustic thorn
deep night
#

I don't consider them that skilled

deep night
rustic thorn
#

Nah, all good.

rustic thorn
deep night
rustic thorn
#

Yea, that's why I specified "on it's own" and " when compaired to the alternative stats given"

#

I don't think KElo should be used alone.

deep night
# rustic thorn Agreed. But would you say most do that?

a lot of the higher kills Elo guys, yes to call them out, do not have the best game sense these days, and are: more open to flanks, don't understand map positions so well, too scared or incapable of flushing out enemies in active combat areas that may require deaths, can only play TDM, can struggle on a large number of maps due to avoidance purely on performance on them... not actual 'fun' preference, etc.

rustic thorn
deep night
#

the funny thing with 1800 is that they might play at the "higher levels" but are less because they play in places that get them killed

#

which is not bad

#

that's just part of taking control of key spots in some maps

#

the higher numbers are sometimes less skilled because they are more of a sniper only

#

and a sniper can't do everything

rustic thorn
#

Agreed.

deep night
#

to call more people out, no name calling... 4v4 is the "I'm too scared to lose elo" hub

#

players reach a high level, become too caught up in the number as the 'proof' of their 'skill'

#

it becomes so superficial and so to keep playing, but not risk a little number, I notice these particular players refuse to do anything except 4v4

#

other maps only happening in wars, where Elo also isn't lost

deep night
rustic thorn
deep night
#

sniping for example doesn't really need any adjustment, the maps aren't big enough for it

#

so it's basically just aim and click

#

which means players can show they aim a big + onto a player quickly... but that's not all of a sniper

#

make it Remagen or something and uh oh... now we miss

#

anyway

#

to get to the suggestion

#

only kills elo was added for fun

#

games elo was added with purpose

deep night
rustic thorn
#

I find different ranges more fun anyway. I enjoy popping someone in Northwest standing 600m away in 2 seconds. Tis satisfying.

deep night
#

a lot of the talking seems to go to kills elo, but the suggestion doesn't specifiy that games elo is desired to be kept or removed too

clear bone
#

Every regular tribute match is just 1900s baiting their teammates

#

And they clean up the easy kills

clear bone
slim zealot
#

I think the problem here is at its core one of definition. What is a good player? What does a good player do, really? What is the acceptable false positive rate for Elo to be considered a useful metric? Could it even be considered useful if intentional false positives are possible? And to answer that Goodhart's law comes into play, as well as the question of the average integrity of humans. And in defining Elo as a metric that people aim for there is the inevitable argument that people can simply not aim for it, which discounts both individual and crowd psychology. So the interlocutors jump back and forth between different semantic layers, which in itself is not a bad thing, but since the layers flow into each other over the course of the discussion each layer bears accumulated malice from the entire previous discussion when it should start with a clean slate, and as the amount of layers and branches grow people start to lose track of the plot.
I don't really know where I was going with this.

deep night
#

I can tell... a lot of it was fluff (words that fill out space, rather than clear and concise)

#

kills elo specifically is the primary cause of complaints for a percentage of the community, and despite mathematically being incompatible as Elo (it's literally not Elo at its core), players discuss whether it is valuable for skill definition or not

#

because we have so many stats, I don't think kills Elo being removed would actually end the world for the people that argue it is useful... there are other things to look at

#

those that argue that it's not that useful to gauge skill would be pleased with it removed

#

just on that alone, the player reception to the removal of it specific to having it judge skill, can be argued as a net positive

#

_ _
we have old and new data that shows that at bare minimum, the direct presence of Elo in visibility has and does create unhealthy habits and play, which actively poison the experience of the victim... AND actively has second-hand consequences on other players around them

#

it leads to some hatred to some alternate game modes which require more than just "kill, don't die"

#

players use the complaint "they're playing TDM in non-TDM" to describe these players, which have a focus on either or both K/D and kills Elo

#

players get driven into individualistic play, rather than teamplay

#

elo was removed from view years ago because of psychological reasons mostly, with how it affected the mind having it there

#

it was visible in comp mode shortly after

#

then comp mode, the last way to see, was removed

#

POMP's mod was the drastic "we want it back" from a small percentage of players, which has led once more to the very reason it was removed from sight years ago and creating a repeat... arguably WORSE than before due to the displayed Elo on names being kills Elo instead of games Elo (further promoting play which can affect matches for those aiming to play a team game and win)

#

_ _
_ _
I don't have any heavy opinion on if the kills Elo should exist at all

#

but I am heavily in favour of it being taken out of the active view it's in... which as long as the Elo does exist, should be made available to see somewhere so people have the option OUTSIDE of the matches in a live-view

#

players that used POMP's Mod actively and managed to quit, voice an instant increase in enjoyment in the game once that toxic pressure was lifted

#

if it were to be removed entirely, an instant thing to point out is: match balance is tweaked... how?

deep night
# slim zealot I think the problem here is at its core one of definition. What is a good player...

-# btw the Goodhart's Law point is a good addition to the topic if the value of Elo had to be discussed, considering that Elo did seem to create an image of what skill was... people have an idea of "this is 1800s skill and this is 1900s skill and this is 1600s skill", but most of any accuracy it did have came from when it was invisible and changed between when players checked it...

-# now with it so actively visible and becoming glorified more and more circa 2021 onward, and people aiming towards it, it lost most of any value it did have... which if that is relevant to if the Elo is kept... yeah, good point

#

slightly off topic, but this is partly why a more mathematically sound Elo for squad ratings in a planned comp scene would be used, but not disclosed...
If I show the rankings, and not the Elo, is resists Goodhart effects (on top of having a few other variables at play in the back to avoid 'gaming' the system and focusing on ranks being the goal).

slim zealot
#

also how is kills elo even calculated?

#

like you said it clearly cant be the same as like in chess

deep night
#

the formula is from Elo, so it's:

image.png
#

Qa = 10^(2000/400) = 10^5 = 100,000
Qb = 10^(1900/400) = 10^4.75 = 56,234.1325...

Expected score for player A:

  • Qa / (Qa + Qb)
  • 100,000 / (100,000 + 56,234)
  • 100,000 / 156,234
  • 0.64

Expected score for player B:

  • Qb / (Qa + Qb)
  • 56,234 / (100,000 + 56,234)
  • 56,234 / 156,234
  • 0.36

Elo change for player A:

  • new rating = current rating + K × (score - expected score)
  • new rating = 2000 + 2 × (1 - 0.64)
  • new rating = 2000 + 2 × 0.36
  • new rating = 2000 + 0.72
  • new rating = 2000.72

Elo change for player B:

  • new rating = current rating + K × (score - expected score)
  • new rating = 1900 + 2 × (0 - 0.36)
  • new rating = 1900 + 2 × -0.36
  • new rating = 1900 + -0.72
  • new rating = 1899.28
#

actual score: 1 = win, 0 = loss, 0.5 = draw

#

E = expected score based on values

#

Expected score is defined as the chance of winning + half the chance of a draw.
So the player is at most expected to have a 64% chance of winning, 0% draw, 36% lose. On the low end they are expected to have ~42.6% chance of winning, ~42.6% draw, 0% lose.

Those are not the only outcomes, they are just the extreme ends.
As an example in the middle, it could be:
50% win, 28% draw, 22% lose.

note that with something like kills Elo, IT IS IMPOSSIBLE TO DRAW... there are two separate kills which provide two different win/loss calculations

vast rivet
deep night
#

For WB here and kills Elo, two things matter:

  • there is no draw condition, a very key part of the entire maths for the Elo formula
  • WB is not a zero-sum game, as Elo has been designed for
#

btw, if you are in a kill trade where you kill one another, you want to be the person that was shot slightly later

#

it's better to have the loss, then win, than win, then loss

clear bone
clear bone
slim zealot