Can you claim back points, if you discover that a player has been cheating?
ELO & ARENA POINTS
Forum rules
Please DO NOT POST BUGS on this forum. Please report (and vote) bugs on : https://boardgamearena.com/bugs
Please DO NOT POST BUGS on this forum. Please report (and vote) bugs on : https://boardgamearena.com/bugs
Re: ELO & ARENA POINTS
That would be great, but not sure. There is a cheater sitting atop the rankings currently. They have been reported by multiple players. Waiting to see what BGA moderators do.
Re: ELO & ARENA POINTS
It is good to see the moderators have penalized the ELO rating of the player using dummy accounts to boost their ELO into the top 5. It is hard to say if this was a short-term strategy over the past 2 weeks or a longer pattern but it did give some food for thought about how the ELO rankings are calculated and gamesmanship in Arena.
It took a large number of matches to creep up that high in the ELO rankings as the Dummy accounts had low rankings but the players who lost to the dummy accounts were penalized a larger number of ELO points. In some ways, the strategy wasn’t a great deal more effective than how some players self-sabotage by being distracted or tired. One thing that did happen was high high-ranked players who are currently motivated to just sit at the top of Arena once they reach the number 1 position were motivated to play.
ELO ratings are an excellent tool for zero-sum games between 2 individuals where there is complete information and low chance/luck. In the case of Spades, there is a high chance level, and a great deal of variability in partnerships which just adds to variability caused by incomplete information.
The strategy of deploying dummy accounts could be possibly negated by encouraging more games between players with similar ELO/ranking and limiting the penalty for higher players when there is a large ranking difference.
Increasing the number of matches between players with similar rankings where ELO excels would increase the reliability of the ranking. It would limit the possibility of employing the strategy this player used by making it time-consuming to increase the ranking of each dummy account to the point where top players would be vulnerable to having their points syphoned off by this sort of partnership.
While ELO is not that reliable for a game like Spades it does give an indication when a result is unlikely due to skill and more likely due to distribution of cards or a weak partner. By discounting only the penalty for the player with high ELO at the end of the match it gives opportunities for players with lower ranks to move up while also giving high-ranked players the opportunity to defend their ranking by playing games with lower-ranked players. If the top player isn’t playing the lower rank players could gradually creep up and pass them in the same way the player who used dummy accounts to overtake the top-ranked player who wasn’t playing.
Most games would still be scored the same as the current system but the matching algorithm would need to be tweaked to create tables and partnerships with similar ELO scores. People might end up playing the same players more often but that also results in more reliable ELO ratings. Players with a lucky streak would still move up very quickly but would face more difficult competition as their ELO increased and the higher-ranked players would be more likely to play them directly if their ELO rating didn’t reflect their ability.
I also think it shows that BGA has a really good opportunity to build stronger communities by creating ladders and groups with the same statistics that are generated for Arena on a smaller scale. People would be encouraged to recruit their friends into small clubs and people could arrange their own leagues like Arena with the settings they prefer to play with. These smaller groups could self-police and competitions between teams from different clubs would provide social opportunities. BGA could take the statistics these games generate to create leader boards for different formats.
BGA is in a difficult position because the more actively they police the more people are likely to complain and try to game the system in another way. THe The time and money spent on moderation also means fewer resources to build the community. I understand their policy on chess, but I think it misses the opportunity to provide chess clubs with a great tool for community building. People can and do use red-thumbs strategically to game the system but they can also be a powerful tool and if people have the opportunity to create clubs and build ladders using their own ideas it helps BGA to beta test the best formats and empower the users. People would join the clubs and groups that offer the best experience and those ideas would propagate.
What I would propose:
1. The ELO penalties to players with higher ELO in matches with a high difference would be decreased as there is a good chance it is a result of lucky deals, or a bad partnership.
2. The matching algorithm prioritizes matching players with high ELO together so that number 1 would be a relatively unlikely occurrence.
3. Generalize the tools BGA uses for Arena so people can create smaller events similar to tournaments using different formats over seasons, or whatever length of time they choose with something similar for a group as the All-time rankings and ladder rankings. Players could participate in multiple groups and formats. Popular formats would lead to larger communities.
What worries me:
1. People will lose interest in a great game because of conflict.
2. BGA will move away from providing statistics and leagues to avoid complaints and conflict.
3. BGA will miss the opportunity to innovate and improve its product and market size.
The BGA idea of keeping these forums positive is a very good one, but it might be worthwhile to give players the chance to avoid players who have moderation complaints against them or a history of cheating. It would protect younger players, and help maintain the integrity of gameplay. Letting people form their own groups and appoint moderators would take the onus off BGA for this kind of gamesmanship.
It took a large number of matches to creep up that high in the ELO rankings as the Dummy accounts had low rankings but the players who lost to the dummy accounts were penalized a larger number of ELO points. In some ways, the strategy wasn’t a great deal more effective than how some players self-sabotage by being distracted or tired. One thing that did happen was high high-ranked players who are currently motivated to just sit at the top of Arena once they reach the number 1 position were motivated to play.
ELO ratings are an excellent tool for zero-sum games between 2 individuals where there is complete information and low chance/luck. In the case of Spades, there is a high chance level, and a great deal of variability in partnerships which just adds to variability caused by incomplete information.
The strategy of deploying dummy accounts could be possibly negated by encouraging more games between players with similar ELO/ranking and limiting the penalty for higher players when there is a large ranking difference.
Increasing the number of matches between players with similar rankings where ELO excels would increase the reliability of the ranking. It would limit the possibility of employing the strategy this player used by making it time-consuming to increase the ranking of each dummy account to the point where top players would be vulnerable to having their points syphoned off by this sort of partnership.
While ELO is not that reliable for a game like Spades it does give an indication when a result is unlikely due to skill and more likely due to distribution of cards or a weak partner. By discounting only the penalty for the player with high ELO at the end of the match it gives opportunities for players with lower ranks to move up while also giving high-ranked players the opportunity to defend their ranking by playing games with lower-ranked players. If the top player isn’t playing the lower rank players could gradually creep up and pass them in the same way the player who used dummy accounts to overtake the top-ranked player who wasn’t playing.
Most games would still be scored the same as the current system but the matching algorithm would need to be tweaked to create tables and partnerships with similar ELO scores. People might end up playing the same players more often but that also results in more reliable ELO ratings. Players with a lucky streak would still move up very quickly but would face more difficult competition as their ELO increased and the higher-ranked players would be more likely to play them directly if their ELO rating didn’t reflect their ability.
I also think it shows that BGA has a really good opportunity to build stronger communities by creating ladders and groups with the same statistics that are generated for Arena on a smaller scale. People would be encouraged to recruit their friends into small clubs and people could arrange their own leagues like Arena with the settings they prefer to play with. These smaller groups could self-police and competitions between teams from different clubs would provide social opportunities. BGA could take the statistics these games generate to create leader boards for different formats.
BGA is in a difficult position because the more actively they police the more people are likely to complain and try to game the system in another way. THe The time and money spent on moderation also means fewer resources to build the community. I understand their policy on chess, but I think it misses the opportunity to provide chess clubs with a great tool for community building. People can and do use red-thumbs strategically to game the system but they can also be a powerful tool and if people have the opportunity to create clubs and build ladders using their own ideas it helps BGA to beta test the best formats and empower the users. People would join the clubs and groups that offer the best experience and those ideas would propagate.
What I would propose:
1. The ELO penalties to players with higher ELO in matches with a high difference would be decreased as there is a good chance it is a result of lucky deals, or a bad partnership.
2. The matching algorithm prioritizes matching players with high ELO together so that number 1 would be a relatively unlikely occurrence.
3. Generalize the tools BGA uses for Arena so people can create smaller events similar to tournaments using different formats over seasons, or whatever length of time they choose with something similar for a group as the All-time rankings and ladder rankings. Players could participate in multiple groups and formats. Popular formats would lead to larger communities.
What worries me:
1. People will lose interest in a great game because of conflict.
2. BGA will move away from providing statistics and leagues to avoid complaints and conflict.
3. BGA will miss the opportunity to innovate and improve its product and market size.
The BGA idea of keeping these forums positive is a very good one, but it might be worthwhile to give players the chance to avoid players who have moderation complaints against them or a history of cheating. It would protect younger players, and help maintain the integrity of gameplay. Letting people form their own groups and appoint moderators would take the onus off BGA for this kind of gamesmanship.
-
ChiefPointThief
- Posts: 486
- Joined: 14 August 2020, 22:27
Re: ELO & ARENA POINTS
Players camping at the top is a sitewide problem because as mentioned this system doesn't account for games with a luck factor. At a certain point it becomes nonsensical to participate further in the competition if your goal is to win. It is bad for solo games and for team games it is even worse. There is a bug report asking for change dating back to 2018.tcarlaw wrote: ↑30 November 2023, 12:59 One thing that did happen was high high-ranked players who are currently motivated to just sit at the top of Arena once they reach the number 1 position were motivated to play.
ELO ratings are an excellent tool for zero-sum games between 2 individuals where there is complete information and low chance/luck. In the case of Spades, there is a high chance level, and a great deal of variability in partnerships which just adds to variability caused by incomplete information.
The strategy of deploying dummy accounts could be possibly negated by encouraging more games between players with similar ELO/ranking and limiting the penalty for higher players when there is a large ranking difference.
Increasing the number of matches between players with similar rankings where ELO excels would increase the reliability of the ranking. It would limit the possibility of employing the strategy this player used by making it time-consuming to increase the ranking of each dummy account to the point where top players would be vulnerable to having their points syphoned off by this sort of partnership.
I also think it would be great if top players were forced to play against each other but there are a lot of difficulties in enforcing this. As far as matchmaking I don't think the player pool for spades and many other games is strong enough. Your idea to limit the possibility of using dummy accounts sounds good in theory. Unfortunately, there are people who will still go the extra mile to cheat.
Re: ELO & ARENA POINTS
At least BGA admins are trying to implement something new about rating. Don't know when it will arrive BGA main site though.
Re: ELO & ARENA POINTS
As far as camping its like the song the Gambler, you need to know when to walk away. An arena ranking mechanism that allowed the kind of ranking inflation that the player was using to game the system but with legitimate games would help to defeat this tactic but it would only be valid if it would require high quality/ELO rankings to implement. It would also lend itself to a different style of cheating but would require much more effort to create groups of people with high rankings. BGA's policy of large K factors for new players should probably be re-examined as it is easier to create a new account and rise through the rankings quickly with the large K while long term members are handicapped with low K values. I don't know if Arena follow the same pattern as ELO with a different K value but I would argue the K should be the same for all Arena players for the entire season.
I think after the halfway point of the season anyone achieving the number 1 ranking should receive a grandmaster title, and top 10 a master title, with maybe a season ending tournament for Masters, followed by another for the winner of that tournament, grandmasters, and previous title holders. Doing some sort of identity verification before these badges are rewarded might not be too onerous and could provide a great opportunity for a press release to promote BGA.
I think after the halfway point of the season anyone achieving the number 1 ranking should receive a grandmaster title, and top 10 a master title, with maybe a season ending tournament for Masters, followed by another for the winner of that tournament, grandmasters, and previous title holders. Doing some sort of identity verification before these badges are rewarded might not be too onerous and could provide a great opportunity for a press release to promote BGA.
Re: ELO & ARENA POINTS
Just with regard to ELO and matching games to people with similar: I was actually Arena "champion" a few seasons back (goodness knows how). Since then my ELO has been as low as low 100's, currently 227 so it is very varaible. I am the same player with the same skills and flaws so the ELO doesn't reflect accurately the game that the individual plays. I can think of one very good player who had consistently very high ELO, they are currnetly mid 100's.It does not accurately reflect the skill of the player. it is just a "snapshot in time".
-
ChiefPointThief
- Posts: 486
- Joined: 14 August 2020, 22:27
Re: ELO & ARENA POINTS
I share similar ideas with you. What you are proposing is heading towards an actual real competition but bga wants to have its cake and eat it too. They market arena as the "competitive" format but they created a watered down "competitive" mode in hopes that players who don't have a chance of actually winning the competition will continue to stay engaged. So while you have a great idea it is unlikely to probably ever be considered.tcarlaw wrote: ↑03 December 2023, 10:32 As far as camping its like the song the Gambler, you need to know when to walk away. An arena ranking mechanism that allowed the kind of ranking inflation that the player was using to game the system but with legitimate games would help to defeat this tactic but it would only be valid if it would require high quality/ELO rankings to implement. It would also lend itself to a different style of cheating but would require much more effort to create groups of people with high rankings. BGA's policy of large K factors for new players should probably be re-examined as it is easier to create a new account and rise through the rankings quickly with the large K while long term members are handicapped with low K values. I don't know if Arena follow the same pattern as ELO with a different K value but I would argue the K should be the same for all Arena players for the entire season.
I think after the halfway point of the season anyone achieving the number 1 ranking should receive a grandmaster title, and top 10 a master title, with maybe a season ending tournament for Masters, followed by another for the winner of that tournament, grandmasters, and previous title holders. Doing some sort of identity verification before these badges are rewarded might not be too onerous and could provide a great opportunity for a press release to promote BGA.
Don't be modest Silver. You won because you're awesomesilvertop wrote: ↑03 December 2023, 13:19 Just with regard to ELO and matching games to people with similar: I was actually Arena "champion" a few seasons back (goodness knows how). Since then my ELO has been as low as low 100's, currently 227 so it is very varaible. I am the same player with the same skills and flaws so the ELO doesn't reflect accurately the game that the individual plays. I can think of one very good player who had consistently very high ELO, they are currnetly mid 100's.It does not accurately reflect the skill of the player. it is just a "snapshot in time".
The purpose of the elo system is to gauge a players strength but the system bga chose for team games by design keeps players level. I'm actually confused how someone thought it was a good idea to use this system if the actual purpose was to rate player strength. Stronger players are constantly giving their elo to their weaker teammate. This is why camping at the top in arena is even more prevalent in spades. The elo leader in some games can take as much as 85% of the risk and only 15% of the reward for the team . If the elo was split 50/50 Silver it would better reflect players skill level.
-
MANNY-OLIVIA
- Posts: 57
- Joined: 21 April 2023, 00:02
Re: ELO & ARENA POINTS
Makes me laugh when CHEATS are moaning and commenting.
We now have 3 Arena Winners who are CHEATS!
We now have 3 Arena Winners who are CHEATS!
Re: ELO & ARENA POINTS
I think Silvertop gives a really valuable point of view. I have looked a little into the developer options for Board Game Arena and there arent' many options as far as game statistics which I can partly understand because it involves access to a lot of valuable proprietary information with some privacy issues. In Simple mode I have a lot of really goood game options that give me the chance to curate my opponents for better games, and I can understand that Arena creates a lot of excitement in new players as the K factor lets them rise through the rankings quickly and whets their appetite while accumalating badges. I really enjoy playing with Silvertop because there is no giant ego and genuine game enjoyment. What worries me is when players who bring less enjoyment to the table start creating dummy accounts and creating a toxic atmosphere. A lot of people won't take the time to look at player profiels and try deciphering the complicated web of deceit in the player logs and just move on to something more enjoyable. What I propose is BGA look at ways to make it easier to create small manageable clubs and encourage people tor recruit new genuine players.