All reports
#11297: "Implemented ELO ranking system not adequate for team games (Belote)"
What is this report about?
BGA website
What happened ? Please select from below
Something didn't function correctly while using the website (not during a game)
Detailed description
-
• Please copy/paste the error message you see on your screen, if any.
none
-
• Please explains what you wanted to do, what you do and what happened
When you play team games (e.g. Belote), BGA considers that you did tie with your partner. So, more you're partner have a low ELO ranking, more your own ELO ranking will degrade (whatever you win or lose).
However, it should be considered as more difficult to win with a lower ranked partner, so your ELO should grow (or at least not being affected by your partner ranking). • What is your browser?
Google Chrome v69
Report history
30. Oct 2019 19:06 • Robo65 • Bug has not been reproduced by delopers yet:
It's totally wrong to show the result as a tie. The team has won, so the platform should report the result as a win (unless the team scores tie, of course :) )
31. Jul 2020 15:36 • Romain672 • Bug has not been reproduced by delopers yet:
Wanted to report the same bug.
Just played a game with 161/104/35 elo in my team, and 1 elo in the ennemy team.
We won, and I gained: +5.7, -4.3, -1.74, which is negative. The system put that to 0, but that's still not enough. I should have won +5.7.
Just played a game with 161/104/35 elo in my team, and 1 elo in the ennemy team.
We won, and I gained: +5.7, -4.3, -1.74, which is negative. The system put that to 0, but that's still not enough. I should have won +5.7.
19. Aug 2020 3:57 • ParadoxicallyAiriest • Bug has not been reproduced by delopers yet:
I had 162 and played with 79 and 40. We won, but I lost 8 points, and they did not even gain any! Maybe this is why high rank players don't want to play with low rank players: even if the game is won, they will lose hard earned points! This is bad, we should encourage experienced players to play with new players, and help to teach them like I did. But this will not happen if the games take away their points like this.
boardgamearena.com/5/hanabi?table=107647644
boardgamearena.com/5/hanabi?table=107647644
5. Oct 2020 2:06 • R o d n e y • Bug has not been reproduced by delopers yet:
The ELO calculation for teams is wrong for all BGA games. In all team games on BGA, the partners tie. So, whether you win or lose, the higher ranked partner is always penalized, and the lower ranked partner is always rewarded. The ELO is supposed to be a reflection of how well a person plays, not their skill in choosing partners.
imgur.com/a/n8ldvnS
In the above game, you can see that the lower ranked partner receives an extra 9.48 points for tying his partner, while his partner loses -9.48 points. Despite both players winning the game.
On the losing team, the lower ranked partner gains 3.51 points for tying his partner, while the higher ranked partner loses an additional -3.86 points, despite both players losing the game.
imgur.com/a/n8ldvnS
In the above game, you can see that the lower ranked partner receives an extra 9.48 points for tying his partner, while his partner loses -9.48 points. Despite both players winning the game.
On the losing team, the lower ranked partner gains 3.51 points for tying his partner, while the higher ranked partner loses an additional -3.86 points, despite both players losing the game.
5. Oct 2020 8:38 • Pollyor • Bug has not been reproduced by delopers yet:
I think this needs to be updated also but the link will not let me like it. And yeah the game should not be a tie either it's a win for the team not a tie. I think it's great to have new players join in games but you should not be over penalised because of it.
5. Oct 2020 23:48 • ElrikMelny • Bug has not been reproduced by delopers yet:
C’est vrai que dans un jeu comme la belote, il semblerait logique que les deux partenaires obtiennent la même perte ou le même gain. A moins que je n’ai pas bien compris le principe du Elo. En tout cas merci aux équipes pour le plaisir que nous avons :) à jouer à tout moment et sur n’importe quel support, c’est magique.
7. Oct 2020 22:03 • AYoshiVader • Bug has not been reproduced by delopers yet:
All team games have problems like this and it ruins the experiece to some high ranked players.
7. Oct 2020 22:32 • NoLuckAndNoSkill • Bug has not been reproduced by delopers yet:
This error still seems to be present
7. Oct 2020 23:02 • Jokerrz0r • Bug has not been reproduced by delopers yet:
The error is still, also for Klaverjassen
7. Oct 2020 23:26 • Viaken • Bug has not been reproduced by delopers yet:
I didn't intend to vote for this. "R O D N E Y" spammed a bunch of my tables with a URL that auto-voted it. I don't particularly care about this "bug".
8. Oct 2020 7:39 • wolmot • Bug has not been reproduced by delopers yet:
Indeed an issue, also received bug report specific on Klaverjassen that winning team is considered a tie.
8. Oct 2020 12:59 • aleman • Bug has not been reproduced by delopers yet:
Please fix, otherwise people will not like to play team games.
9. Oct 2020 8:06 • er_mondezz • Bug has not been reproduced by delopers yet:
same problem for Armadora....
12. Oct 2020 22:48 • board game nerd • Bug has not been reproduced by delopers yet:
The ELO calculation is wrong for all teams games on BGA. Win or lose, one partner is penalized, and the other benefits. The bug was reported over 2 years ago, and hasn't been fixed. Will you please vote to fix this bug? And if possible, leave a comment. /bug?id=11297&vote got this wile playing sasntorine
12. Oct 2020 23:04 • BoB Special • Bug has not been reproduced by delopers yet:
Issue exists for years now. Goes for all team games. I love bga, but i think this is the worst error in the system. It really takes away the fun from team games
Please fix asap!!!
Please fix asap!!!
12. Oct 2020 23:07 • BoB Special • Bug has not been reproduced by delopers yet:
This is not a game issue, but a rating issue. Seems really easy to solve - computer scientist myself :)
Think it's a quickwin
Think it's a quickwin
12. Oct 2020 23:19 • board game nerd • Bug has not been reproduced by delopers yet:
didn't yet finish the game
16. Oct 2020 7:44 • kllll • Bug has not been reproduced by delopers yet:
Bug has not been reproduced by developers yet:
All team games
All team games
9. Nov 2020 8:59 • Cenine • Bug has not been reproduced by delopers yet:
Same issue with HANABI : the players have ties wether we finish or lose.
9. Nov 2020 21:27 • ReneSauz • Bug has not been reproduced by delopers yet:
ITS NOT A BUG...its an intentional and silly programming error. In team games just don't add or subtract ELO for the partner. There has to be an easy way to implement this instead of ignoring it.
30. Dec 2020 16:49 • mickalot • Bug has not been reproduced by delopers yet:
This seems to be especially bad for cooperative games -- it makes it impossible to get enough points to be eligible for arena mode.
20. Jan 2021 12:24 • Tussengas • Bug has not been reproduced by delopers yet:
also, you don't get your 5 gift points for these games
26. Jan 2021 16:13 • ptitours de lyon • Bug has not been reproduced by delopers yet:
In a team game , the two parteners are considered as "equality" and if you play whit a novice , you lose point because of probality of win, against a partener , which is impossible.
in first correction, you may suppress these points of equality
in second correction, you should calculate probapility of win by team, for example by compare the sum of ELO
in first correction, you may suppress these points of equality
in second correction, you should calculate probapility of win by team, for example by compare the sum of ELO
15. Feb 2021 22:06 • unclesamb • Bug has not been reproduced by delopers yet:
To summarize the requests as I read them:
1. In a cooperative game in which all players win or lose together (e.g. The Crew, Hanabi, Bandido, etc), the ELO gains/losses should be calculated based on whether the group as a whole won or lost ("tie" should not be an option). Then the ELO gains/losses for each player should take into account the experience of each individual player.
2. In a team game (e.g. Kami), the ELO gains/losses for all members of a team should be based on whether the team as a whole won or lost ("tie" should not be an option). Then the ELO gain/losses for each player on each team should take into account the experience of each individual player.
In team or cooperative games, there should be no ties--there should only be winners and losers, and points should be awarded accordingly.
1. In a cooperative game in which all players win or lose together (e.g. The Crew, Hanabi, Bandido, etc), the ELO gains/losses should be calculated based on whether the group as a whole won or lost ("tie" should not be an option). Then the ELO gains/losses for each player should take into account the experience of each individual player.
2. In a team game (e.g. Kami), the ELO gains/losses for all members of a team should be based on whether the team as a whole won or lost ("tie" should not be an option). Then the ELO gain/losses for each player on each team should take into account the experience of each individual player.
In team or cooperative games, there should be no ties--there should only be winners and losers, and points should be awarded accordingly.
19. Apr 2021 0:34 • Silene • Bug has not been reproduced by delopers yet:
More relevant than ever now with Tichu in Beta.
3. May 2021 12:10 • JSnova • Bug has not been reproduced by delopers yet:
as for hanabi, based on a game theory study on hanabi, only 82% of the games are winnable (max score), based on the most known tricks, finesse, reverse finesse, gentelman discard, bluff etc.
At the moment, you get punished if you play with lower elo players, instead of rewarded.
Also how higher the difficulcy specially multi 5, reduces the max. win chance.
this is not corrected within the current elo system
also all people agreeing to leave, should result in a default pushingment or only be aloud 1 time in the 4-5 games or so.
At the moment, you get punished if you play with lower elo players, instead of rewarded.
Also how higher the difficulcy specially multi 5, reduces the max. win chance.
this is not corrected within the current elo system
also all people agreeing to leave, should result in a default pushingment or only be aloud 1 time in the 4-5 games or so.
4. May 2021 18:46 • Just Keep Climbing • Bug has not been reproduced by delopers yet:
same bug in spades
30. May 2021 0:52 • BrenBarn • Bug has not been reproduced by delopers yet:
It's frustrating that this issue still exists, as it makes ratings kind of meaningless for team games.
The problem is this: ELO is based on rating two players who COMPETE AGAINST one another. Currently, BGA considers any two players in the same game to be competing, and adjusts their ELO based on their relative standing. Two players on the same team are considered to be tied.
But this is wrong. Two players on the same team do not COMPETE AGAINST one another at all. So pairs of players on the same team should simply be ignored for ELO calculations.
BGA already computes the overall ELO adjustment by adding up the ELO adjustments for each pair of players. For instance, in a non-team 5-player game, if you come in third place, your overall ELO is adjusted based on the ELO effect of losing to the two players above you and losing to the two below you. This is all fine.
All that needs to be done is to exclude PAIRS of players on the same team from these computations. If players A and B are on the same team, their ELO standings relative to one another should not even be considered, as they were never in competition. Player A did not win AGAINST player B, nor did player A lose AGAINST player B, not did player A tie AGAINST player B. There was simply no game "between" these two players, so the ELO rating should reflect that and not compare them in any way. They should gain or lose ELO points only by comparison with players on other teams.
The problem is this: ELO is based on rating two players who COMPETE AGAINST one another. Currently, BGA considers any two players in the same game to be competing, and adjusts their ELO based on their relative standing. Two players on the same team are considered to be tied.
But this is wrong. Two players on the same team do not COMPETE AGAINST one another at all. So pairs of players on the same team should simply be ignored for ELO calculations.
BGA already computes the overall ELO adjustment by adding up the ELO adjustments for each pair of players. For instance, in a non-team 5-player game, if you come in third place, your overall ELO is adjusted based on the ELO effect of losing to the two players above you and losing to the two below you. This is all fine.
All that needs to be done is to exclude PAIRS of players on the same team from these computations. If players A and B are on the same team, their ELO standings relative to one another should not even be considered, as they were never in competition. Player A did not win AGAINST player B, nor did player A lose AGAINST player B, not did player A tie AGAINST player B. There was simply no game "between" these two players, so the ELO rating should reflect that and not compare them in any way. They should gain or lose ELO points only by comparison with players on other teams.
7. Jun 2021 18:25 • xjk_cn • Bug has not been reproduced by delopers yet:
boardgamearena.com/table?table=178503292
Losing ELO even if I did my best in yokai. Maybe I need to try level 2 with high Elo players?
Losing ELO even if I did my best in yokai. Maybe I need to try level 2 with high Elo players?
19. Jun 2021 0:00 • alextingle • Bug has not been reproduced by delopers yet:
Just to add to this. ELO isn't just about bragging rights, it helps to arrange games between players of approximately equal skill.
A "Not Alone" game can be completely ruined by a single inexperienced player who makes bad moves. The ELO system makes it impossible to identify those poor players. It's made even worse by the "bonus" K-factor that new players get... a couple of lucky wins can propel a new player to ELO of 150..200, which is pretty high for Not Alone. So they *look* experienced, even if they have no idea how to play.
It discourages me from playing the game at all. I often have time for one game, including waiting for 20 minutes - it's pretty frustrating to have it ruined by a weak player who makes multiple unforced errors. I'd rather play a different game where I know I'll have some fun...
A "Not Alone" game can be completely ruined by a single inexperienced player who makes bad moves. The ELO system makes it impossible to identify those poor players. It's made even worse by the "bonus" K-factor that new players get... a couple of lucky wins can propel a new player to ELO of 150..200, which is pretty high for Not Alone. So they *look* experienced, even if they have no idea how to play.
It discourages me from playing the game at all. I often have time for one game, including waiting for 20 minutes - it's pretty frustrating to have it ruined by a weak player who makes multiple unforced errors. I'd rather play a different game where I know I'll have some fun...
25. Jul 2021 14:30 • Ax109 • Bug has not been reproduced by delopers yet:
I have not experienced this with the game Pandemic, but this might be because this is a game you don't play 'against' anyone, or the fact I play with low-ranked players.
30. Jul 2021 19:16 • Jellby • Bug has not been reproduced by delopers yet:
As far as I understand it, Hanabi gets it "right":
* All players win or lose together.
* The ELO change is calculated based on the average ELO of all the players and a "difficulty" level depending on the settings.
* All players gain or lose the same ELO (except possible differences from rounding and possibly k-factors).
I believe other cooperative games should use the same scheme, although deciding what the "difficulty" level is not trivial and may need adjustments. Hopefully that can be more easily done one a decent amount of finished games is available.
For other team games, it would be even easier: compare the average ELO of each team, and have each member of the team gain or lose the same amount.
* All players win or lose together.
* The ELO change is calculated based on the average ELO of all the players and a "difficulty" level depending on the settings.
* All players gain or lose the same ELO (except possible differences from rounding and possibly k-factors).
I believe other cooperative games should use the same scheme, although deciding what the "difficulty" level is not trivial and may need adjustments. Hopefully that can be more easily done one a decent amount of finished games is available.
For other team games, it would be even easier: compare the average ELO of each team, and have each member of the team gain or lose the same amount.
31. Jul 2021 12:18 • joueur ooo • Bug has not been reproduced by delopers yet:
Je n’ai pas pu cliquer pour valider mes choix au cours d’une partie.
Cela ne vient pas de mon appareil c’est arrivé A deux parties d’affilée sur votre site.
Cela ne vient pas de mon appareil c’est arrivé A deux parties d’affilée sur votre site.
11. Aug 2021 3:36 • Pumpkinthekingofgame • Bug has not been reproduced by delopers yet:
競技場即時制有時會搜尋不到對手，導致等很久
即便按了取消且關掉BGA網站
仍然進入了一場桌子，且沒有確認玩家是否還在場就開始遊戲
玩家已將網站關閉且人不在現場
仍然讓玩家進入桌子
就領到了超時以及離場
:(
即便按了取消且關掉BGA網站
仍然進入了一場桌子，且沒有確認玩家是否還在場就開始遊戲
玩家已將網站關閉且人不在現場
仍然讓玩家進入桌子
就領到了超時以及離場
:(
25. Oct 2021 13:19 • Joseph2302 • Bug has not been reproduced by delopers yet:
We shouldn't be using the standard win/loss against people ELO for co-operative or team games (like Ticu, which is 2v2). Because you can do well/win and still lose rating because you "tied" against someone with lower rating than you
For team games, needs to ignore people that were on your team
For co-operative games, needs a benchmark for good/winning score, and work out the ELOs based on how close you got to that score
For team games, needs to ignore people that were on your team
For co-operative games, needs a benchmark for good/winning score, and work out the ELOs based on how close you got to that score
8. Nov 2021 3:16 • Maester_RJ • Bug has not been reproduced by delopers yet:
A pagina fica carregando e não entra oponentes, mesmo recarregando, já alguns dias vem ocorrendo esse problema para mim. Por qual motivo, eu iria deixar um jogo, se, eu entro aqui, para jogar?
22. Nov 2021 17:57 • Skeptic Dave • Bug has not been reproduced by delopers yet:
Coop games like Tranquility definitely suffer from this broken system. To date,I've played > 90 games of Tranquility, winning > 60 of them, yet . . . my ELO is 64! Many times, I gain zero ELO points even when playing with a similarly-ELOd partner and while winning solidly on medium difficult settings.
As others have stated, the ELOs are meaningless when trying to decide whether to play games with others. I've had other people in Tranquility games with ELOs > 100 who clearly do not understand basic gameplay concepts. I enjoy the game tremendously, but would prefer not to have to teach it while playing (most of the time.) The current ELOs make that impossible to know ahead of time.
This may not be a "bug" in the normal sense, but the end result is no different.
As others have stated, the ELOs are meaningless when trying to decide whether to play games with others. I've had other people in Tranquility games with ELOs > 100 who clearly do not understand basic gameplay concepts. I enjoy the game tremendously, but would prefer not to have to teach it while playing (most of the time.) The current ELOs make that impossible to know ahead of time.
This may not be a "bug" in the normal sense, but the end result is no different.
24. Nov 2021 10:54 • Philoupino • Bug has not been reproduced by delopers yet:
Could you include alos the game Klaverjassen
boardgamearena.com/gamepanel?game=klaverjassen
Like belote, there are always 2 Winners & 2 Loosers.
boardgamearena.com/gamepanel?game=klaverjassen
Like belote, there are always 2 Winners & 2 Loosers.
30. Dec 2021 6:59 • youngathearttx • Bug has not been reproduced by delopers yet:
In Tichu, if my partner calls Tichu or Grand Tichu, and I support his call with my play, so he can go out first, why should I receive less ELO than my winning team partner ??
4. Feb 2022 7:09 • Kanan Jarrus • Bug has not been reproduced by delopers yet:
Impossible de quitter la partie hier alors qu'elle était finie. J'ai essayé de concéder la victoire puis j'ai dû quitter. Le bug est connu apparemment
Cordialement
Cordialement
11. Feb 2022 16:24 • Romain672 • Bug has not been reproduced by delopers yet:
To try to find a solution, I find this reddit subject: www.reddit.com/r/math/comments/fdz879/using_elo_to_rank_players_in_1v1_and_2v2_game/ , but there isn't much things about it.
It look like not counting the partner tie is usually done, but it's still not a solution which resolve all problems (every top elo would want to play with new players, but which are good).
It look like not counting the partner tie is usually done, but it's still not a solution which resolve all problems (every top elo would want to play with new players, but which are good).
11. Feb 2022 16:25 • Romain672 • Bug has not been reproduced by delopers yet:
And they said to just take the average elo of both teams.
24. Feb 2022 21:12 • Shiny_Charizard • Bug has not been reproduced by delopers yet:
Similar for BANG, the winner ELO should only be compared against the loser ELO and the loser ELO against the winner ELO. Ties amongst team members are just simply wrong.
Average ELO for the entire team also makes sense
Average ELO for the entire team also makes sense
5. Apr 2022 0:25 • ptitours de lyon • Bug has not been reproduced by delopers yet:
You should realy look up this ticket. Or create a "don't look up" team
20. Apr 2022 6:03 • JustinGames00 • Bug has not been reproduced by delopers yet:
I just lost 27 ELO points for a cooperative game of Hanabi. My dad is new to the game, and I used a normal mode to help him gain some points. I pretty much only play with friends, so ELO doesn't matter much to me, but I'd like to be able to hop on a game with a random opponent every now and then. It would be nice if ELO was a reasonable estimation of someone's skill in the game rather than a random number based on whether you play with people better or worse than you often. If you're not going to implement it well, just drop the ELO ranking for coop games.
3. May 2022 18:08 • lordmcfuzz • Bug has not been reproduced by delopers yet:
4 years later... not even acknowledged.
30. May 2022 4:28 • BillyBullet • Bug has not been reproduced by delopers yet:
I can't report anything, always the error message. what is the tech support link??? how can they want someone to be a paying user with a site with bad technical support?
30. May 2022 4:28 • BillyBullet • Bug has not been reproduced by delopers yet:
I can't report anything, always the error message. what is the tech support link??? how can they want someone to be a paying user with a site with bad technical support?
22. Aug 2022 14:41 • Juxta13 • Bug has not been reproduced by delopers yet:
ELO does seem broken. Players should never lose points for a win (including a tie, including if someone left the game and the others are considered winners). It just encourages people to be more selective (discriminative) about who they will play with. In team/partner games, someone always gets hurt.
1. Sep 2022 7:57 • strufle • Bug has not been reproduced by delopers yet:
Exclu dans les temps au cour d'une partie avec la n long temps dans les tours
5. Sep 2022 12:14 • tycy • Bug has not been reproduced by delopers yet:
We can get a new one for bag
16. Oct 2022 18:57 • 81Valerii • Bug has not been reproduced by delopers yet:
Quand je joue sur le site, ça coupe, plus d'internet.
Je me retrouve déconnecté du jeu, d'un coup.
Je ne peux rien faire, en plein milieu de la partie.
Cela mais arrivé plusieurs fois.
Ça ne m'arrive que sur ce site.
Je suis pénalisé alors que se n'est pas ma faute.
Cordialement.
Je me retrouve déconnecté du jeu, d'un coup.
Je ne peux rien faire, en plein milieu de la partie.
Cela mais arrivé plusieurs fois.
Ça ne m'arrive que sur ce site.
Je suis pénalisé alors que se n'est pas ma faute.
Cordialement.
26. Dec 2022 20:06 • EleaTyrell • Bug has not been reproduced by delopers yet:
I also encountered this unpleasantness, in a game of warchest in 2 vs 2 players; I simply don't want to play in team mode any more...
29. Dec 2022 23:39 • yaron • Bug has not been reproduced by delopers yet:
Points awarded/deducted should be based on the average score of each team.
That way, a low ranked player would get more points for winning, but so would their partner (since winning with a weak partner is harder).
That way, a low ranked player would get more points for winning, but so would their partner (since winning with a weak partner is harder).
2. Jan 2023 11:23 • Onitama Königin Jo • Bug has not been reproduced by delopers yet:
Yes,it often happens that someone has a Partner with who he plays often. The name of the player is Sussuarana Onitama. He plays a lot against beginners and have some second accounts. If he lost against the players who are probably his second accounts,the players gave the game with him up or let him win.And this problem was often in the past and sometimes Sussuarana Onitama does it do today also.
6. Jun 2023 8:13 • fishdapotato • Bug has not been reproduced by delopers yet:
i played with other 3 players. they have higher ELO than me. At last, one win and we other 3 is first-runner (all same marks). Yet, turn out they reduce 6 and 7 ELO while deducted me 18 ELO. (their ELO is ordinary higher than me already)
20. Jul 2023 15:27 • Hagardunor • Bug has not been reproduced by delopers yet:
Completely agree. ELO system has been built for a face to face game (Chess) and is totaly inadequate for team games
15. Aug 2023 10:43 • Malo77 • Bug has not been reproduced by delopers yet:
My 2 cents about what a good Score system could be, and looks like ELO:
ELO System for all team games and/or collab games:
Maybee a system like:
- at avery game start: every player bet 1% of his ELO point into the game pot
- if the game is lost the points are lost (and BGA "keeps" thems)
- if the game is won: take the pot point x RATIO, split and and give them back to the winning players.
- the RATIO is automatically adjusted by BGA ... if the stats show that an hanabi game is won 33% of the time .. 1 time over 3 ... the RATIO multiplier would be x2 (3 games-1win=2x) .. either each players gets back his points x2 .. or split them another way ...
- this way the total ELO points of players is about the same, without points growing without limit
ELO System for all team games and/or collab games:
Maybee a system like:
- at avery game start: every player bet 1% of his ELO point into the game pot
- if the game is lost the points are lost (and BGA "keeps" thems)
- if the game is won: take the pot point x RATIO, split and and give them back to the winning players.
- the RATIO is automatically adjusted by BGA ... if the stats show that an hanabi game is won 33% of the time .. 1 time over 3 ... the RATIO multiplier would be x2 (3 games-1win=2x) .. either each players gets back his points x2 .. or split them another way ...
- this way the total ELO points of players is about the same, without points growing without limit
28. Aug 2023 22:23 • Hagardunor • Bug has not been reproduced by delopers yet:
this demand was opened 4 years ago and nothing has appened since then des pitehundreds of signatures
13. Oct 2023 18:14 • tcarlaw • Bug has not been reproduced by delopers yet:
Ideally, ELO ratings should be calculated based on the lowest ELO ranking from each partnership. It would motivate people to play with lower-ranked players, instead of penalizing people for being social. The partner matching algorithm could be tweaked to prevent gaming the system. Still, it would allow people to find matches in less popular games and take the pressure off low-ranked players who may worry about hurting their partner's ELO rating. I felt a little bit like a kingmaker when my ranking tanked close to the end of the Arena season knowing higher ranked players would take a big hit for being paired with me, which would be hard to avoid in a small game community.
30. Nov 2023 19:55 • ChiefPointThief • Bug has not been reproduced by delopers yet:
5yrs and counting
14. Dec 2023 13:39 • penguin-EM • Bug has not been reproduced by delopers yet:
Not playing the game but reducing ELO592→552 for no reason
16. Dec 2023 21:03 • RicardoRix • Bug has not been reproduced by delopers yet:
The bug has been fixed.
But it still exists for games where someone quits or is time expelled.
ELO is exchanged between all players that didn't quit. NOT RIGHT!
But it still exists for games where someone quits or is time expelled.
ELO is exchanged between all players that didn't quit. NOT RIGHT!
22. Dec 2023 11:36 • ptitours de lyon • Bug has not been reproduced by delopers yet:
The bug is not fixed in Belote
in first correction, you may suppress points of equality between partners
in second correction, you should calculate probapility of win by team, for example by compare the sum of ELO from team
in first correction, you may suppress points of equality between partners
in second correction, you should calculate probapility of win by team, for example by compare the sum of ELO from team
21. Feb 2024 22:49 • ErikLevin • Bug has not been reproduced by delopers yet:
I'm not sure why people are saying this is fixed. Was it fixed very briefly and then broken again? I still see in team games, every time, that members within the teams exchange Elo points with one another for being tied.
Also, I don't think any complicated solutions are necessary. It is very simple: Players within the same team should not exchange Elo points with one another. Just remove team members from the calculation.
This seems like an easy fix for an obvious flaw, which has a very high demand from the user base (fourth highest voted bug/suggestion). It is very unfortunate that BGA did not roll an improvement in with the recent changes to the Elo system. You'd think that when you rework a site feature like Elo you take the chance to fix some longstanding issues with that feature while you're there. Let alone that BGA hasn't even responded or acknowledged the ticket.
Also, I don't think any complicated solutions are necessary. It is very simple: Players within the same team should not exchange Elo points with one another. Just remove team members from the calculation.
This seems like an easy fix for an obvious flaw, which has a very high demand from the user base (fourth highest voted bug/suggestion). It is very unfortunate that BGA did not roll an improvement in with the recent changes to the Elo system. You'd think that when you rework a site feature like Elo you take the chance to fix some longstanding issues with that feature while you're there. Let alone that BGA hasn't even responded or acknowledged the ticket.
17. Jul 2024 6:04 • Camargo_Silva • Bug has not been reproduced by delopers yet:
The bug is not fixed in Belote
26. Jul 2024 17:07 • Mit55 • Bug has not been reproduced by delopers yet:
Bonjour, depuis ce jour je ne parviens pas à jouer correctement alors que je suis bien connecté, j'ai donc été pénalisé à plusieurs reprises alors qu'il n' y a pas lieu!
joueur Mit 55
joueur Mit 55
12. Oct 2024 4:39 • w41k3r13 • Bug has not been reproduced by delopers yet:
This is not a bug, but simply a reminder that BGA does not care about it's players. It's crucial to understand their business model, you see, the consumer is the company that needs people to play their game, the product is you. A broken ELO system in single and Multi-player is the drug that keeps you around for the abuse. If BGA actually cared about the players, ELO would never have been applied to... and this is the truly crazy thing...
BOARD GAMES
BOARD GAMES
15. Nov 2024 16:21 • Hisoka_Onnanoko • Bug has not been reproduced by delopers yet:
Bonjour, depuis ce jour je ne parviens pas à jouer correctement alors que je suis bien connecté, j'ai donc été pénalisé à plusieurs reprises alors qu'il n' y a pas lieu!
6. Dec 2024 22:11 • Handsome the first • Bug has not been reproduced by delopers yet:
The same thing should applies for when the time runs out for one of the player, every other player should only win against the quitter to count towards the ELO and not also calculate the tie amongst themselves.
7. Dec 2024 21:47 • EcoRim2017 • Bug has not been reproduced by delopers yet:
Bug... to me showed other player was out of time
20. Jan 2025 14:20 • BurgerDragon • Bug has not been reproduced by delopers yet:
To address this, I suggest a system where
1. each player contributes some ELO (with higher ELO players contributing more) before the match begins, and
2. the total ELO is distributed equally among the winners.
1. each player contributes some ELO (with higher ELO players contributing more) before the match begins, and
2. the total ELO is distributed equally among the winners.
8. Feb 2025 5:08 • mnemonicker • Bug has not been reproduced by delopers yet:
Your ELO calculation doesn't make sense in most of the games. You are using too high K factor. It is like 60, 40, 20. It should be like 40, 20, 10.
14. Apr 2025 2:57 • nalka • Bug has not been reproduced by delopers yet:
Some changes have been made to the system since this bug has been reported. From what I see this is fixed in some games (like Coinche) but not all (for example, you still have this problem in Setup, Toc, Warchest...). I couldn't figure what this is based on but feel free to share if you know.
It looks like there is a second system that has been implemented to help fight this problem. I don't really understand its point given a clean solution apparently exists (cf. Coinche) but it seems to prevent elo losses for players that are declared "Winner" at the end of a game. For example, it triggered in these games:
• boardgamearena.com/table?table=657971566 (2 players, perfectly tied)
• boardgamearena.com/table?table=658333650 (3 players, first 2 are perfectly tied, the other one just lost).
The thing is Living Forest is NOT a team game. This means a higher elo player tying a lower elo one doesn't lose elo while they should. To summarize, this second system not only doesn't really fix anything for team games but it also bugs some solo games' elo calculation when there are ties.
It looks like there is a second system that has been implemented to help fight this problem. I don't really understand its point given a clean solution apparently exists (cf. Coinche) but it seems to prevent elo losses for players that are declared "Winner" at the end of a game. For example, it triggered in these games:
• boardgamearena.com/table?table=657971566 (2 players, perfectly tied)
• boardgamearena.com/table?table=658333650 (3 players, first 2 are perfectly tied, the other one just lost).
The thing is Living Forest is NOT a team game. This means a higher elo player tying a lower elo one doesn't lose elo while they should. To summarize, this second system not only doesn't really fix anything for team games but it also bugs some solo games' elo calculation when there are ties.
14. Apr 2025 10:09 • Phoxtrot • Bug has not been reproduced by delopers yet:
Probably related to this is, the "losers_not_ranked" option that is notably bugged and not doing what would actually make sense.
boardgamearena.com/bug?id=111437
boardgamearena.com/bug?id=161683
boardgamearena.com/bug?id=111442
boardgamearena.com/bug?id=111437
boardgamearena.com/bug?id=161683
boardgamearena.com/bug?id=111442
15. May 2025 8:49 • Arzololi • Bug has not been reproduced by delopers yet:
Le système ne fonctionne pas et il est impossible de jouer...je sors pour rentrer à nouveau et ça bloque la page BGA ne s'ouvre plus
C'est régulier ces dernieres semaines et je suis ensuite pénalisée.
C'est régulier ces dernieres semaines et je suis ensuite pénalisée.
Add something to this report
Please add here anything that seems relevant to reproduce this bug or understand your suggestion:
- Another table ID / move ID
- Did F5 solve the problem?
- Did the problem appears several time? Everytime? Randomly?
- If you have a screenshot of this bug (good practice), you can use Imgur.com to upload it and copy/paste the link here.
