New mechanism to prevent skipping in tournaments

[nik592]

Multiplayer elimination format will always be riddled with issues for handling player elimination with timeouts and inactive players I would make a Multiplayer Elimination specific option for how to handle timeouts. Just eliminate out of time players one by one based on most time used. If all eliminations are done, carry on. If more eliminations are required but the players stayed within the time limit, don't eliminate them.

I think the issue here is that if you're not eliminating at a consistent rate, this will make a tournament drag on longer. If I have a tournament with 20 players, 5 tables of 4, that's supposed to eliminate 2 each round, round 2 will have 10 players, round 3 should have 6, round 4 will have 4, and now we can find our winner. If half (rounded up) of the tables only eliminate one because of a timeout issue, round 2 now has 13, round 3 has 9, round 4 has 6, round 5 has 5, round 6 has 4, we're taking two extra rounds to get down to a final four.

I don't disagree that more flexibility in options would be good, but I assume there's work needed to code these additional options, and given that player matching already has some fairly significant bugs, I personally think it may be better to fix those before introducing more complexity. But I also don't run and don't play in a lot of tournaments (though I might play in more if I can find ones that disallow skipping).

[Andrew Davidson]

So, if the option to prevent skipping is chosen, the individual player time limits have no meaning, right? All that matters is the overall time allowed for the game, right? And if you hit that limit, the player who has used least time wins, right? That could turn games into a scramble like speed chess and so, perversely less relaxed.

[Mathew5000]

So, if the option to prevent skipping is chosen, the individual player time limits have no meaning, right? All that matters is the overall time allowed for the game, right? And if you hit that limit, the player who has used least time wins, right? That could turn games into a scramble like speed chess and so, perversely less relaxed.

I can imagine a situation where someone would strategically not play (thus letting the game time out) rather than playing a move to advance the game to its "normal" conclusion. But hopefully those situations will be rare. One question I have, is there any karma penalty for running out of time in a tournament game? Or does the karma penalty come only if another player skips you?

[nik592]

I can imagine a situation where someone would strategically not play (thus letting the game time out) rather than playing a move to advance the game to its "normal" conclusion. But hopefully those situations will be rare. One question I have, is there any karma penalty for running out of time in a tournament game? Or does the karma penalty come only if another player skips you?

You do get a time penalty on your account for going over time, but only get karma penalties for being skipped - no penalties if a game reaches maximum time. I mean, this can already happen, and would be preferred in some cases (eg. where a user has a relatively new account and a game is multi-elimination - skipping would be dangerous for them). But you've got to be in a scenario where another person at the table has so much less time than you that waiting out the remainder of the game would still leave them as the person who has used the most time. That's a pretty hard trick to pull off, and relies on another person being super slow before you decide to do this (or where a game is already pretty close to its max time anyway).

[Shivaware]

I'm playing my first tournaments with the new option to prevent skipping, and it's working great. Thanks BGA, for making this change! It will drastically improve most tournaments with more than 2 players.

[BananaHarambe]

I am actually quite torn on the prevent skipping option for tournaments. I have used it on a few 2p tournaments and everything went fine except that players that ran out of time never got a karma penalty like they do for skipping.

Choosing the prevent skipping option for a tournament is effectively saying "disable any reputation/karma loss" . I set my tournaments with very reasonable time settings so I would prefer overtime players to get a karma penalty. I put the reputation requirements for all my tournaments at maximum 85% because I only want the most reliable and respectful players in my tournaments. If players can keep timing out in tournaments without penalty then how will I be able to filter them out?

Example would be someone with 100% rep a while ago signs up for many tournaments and decides to stop BGA entirely. Normally if they joined 4+ games and got skipped their reputation would tank to 84% or lower which means that they could be removed from the tournament signup automatically 24 hours before the tournament starts and stop being a problem. If they joined a bunch of prevent skip tournaments we can no longer filter them.

Also for 3p and 4p games, there is no way I would disable skips because I think skips are the lesser evil of how to handle incomplete games with its current handling, which is to use the old "placement by thinking time" method. There is no way I want a 3p or 4p game result to be dictated by thinking time rather than split than split the win with all players who played fast enough.

The only situation I can recommend preventing skips in my personal opinion: Testing new games and tournament settings when you are unsure if the game duration is long enough and you do not want players to get penalized.

If BGA changed how they handle placement in maximum duration games to something more fair such as "on time= winners, overtime=loser" I would be more in favor of it. I would also prefer karma penalties for going overtime in tournament games to remain with meaningful penalties for being unable to complete a game by maximum duration.

[Meeplelowda]

If BGA changed how they handle placement in maximum duration games to something more fair such as "on time= winners, overtime=loser" I would be more in favor of it. I would also prefer karma penalties for going overtime in tournament games to remain with meaningful penalties for being unable to complete a game by maximum duration.

I haven't disabled skipping in any of my tournaments, but I exclusively run 2p tournaments. It feels like it isn't needed until you have to deal with higher player counts.

But does on time = winner really work for 3p or more when often it's just one bad apple timing out and everyone else is on time?

[BananaHarambe]

I haven't disabled skipping in any of my tournaments, but I exclusively run 2p tournaments. It feels like it isn't needed until you have to deal with higher player counts.
But does on time = winner really work for 3p or more when often it's just one bad apple timing out and everyone else is on time?

For most 3p matches on time= winner works exactly how skipping currently works with 1 player being the culprit most of the time. The skipped player gets last place and everyone else tie for 1st.

My proposal for on time= winner would make this same result translate to the prevent skip settings and make it more flexible overall.

In the same 3 player game example with 5 days thinking time per player (like my Vale tournaments) what happens often is after about 9 days one player runs over time and going to the negative with other players having 72 hours each remaining. With skipping enabled there is a good chance that the game will still finish on time if everyone is putting effort to make frequent moves, but neither player with time remaining is at risk of going over time themselves because if they ever get somewhat low on time (<8 hours) the optimal move would be to skip the other player first. I personally would not risk getting skipped myself to extend the game.

With skips prevented in the above scenario there would not be any drama regarding when to skip a player. If one player goes over time each other player should still try to make their moves fast enough to having thinking time remaining after 15 days. If the other 2 players go quickly there is a good chance that the game finishes on time and everyone is content. If one of the other two players plays too slowly they would be at risk of going overtime themselves and also getting a loss. The slow player also has an incentive to complete the game to avoid the auto loss. I feel like everyone has enough incentive to keep the game moving with this system.

What if a 3p game has 2 slow players and 1 fast player? What happens after 11 days if the fast player has nearly 96/120 hours left and the other two players have only a few hours left or one has already gone negative? With skipping both slow players would be incentivized to skip the other player to avoid getting skipped which is some tragedy of the commons - Game Theory scenario. With my "on time / over time" system the game outcome would more fairly reflect players time management. If the game doesn't finish and 2 players were over time limit, both of them are shared losers and the remaining 1 player is the winner. This is most fair since that player tried much harder to play on time. For both players out of time, both of them are incentivized to finish the game to avoid a guaranteed loss so the game should continue moving.

Also consider higher player counts. A 4p game with 30 day duration will quite often have 2 slow players that require over 7.5 days thinking time and 2 fast players that are helping move the game along with no risk of ever timing out. If the game is incomplete after 30 days and the thinking time was 11-11-4-4 days among the players I think it would be most fair to decide the two fast players with 4 days time winners that share 1st place and the two slow players that both ran out of time as losers with shared 3rd place (in the case of this 4p scenario). Trying to sort players by exact time spent like the current system doesn't really feel fair to me. If I played quickly and stayed within my time limit for a full 30 days why should I get penalized with 2nd place because I used more thinking time? The thinking time was entirely out of my control and decided by player order and time zone/play hours more so than being active.

[Meeplelowda]

I haven't disabled skipping in any of my tournaments, but I exclusively run 2p tournaments. It feels like it isn't needed until you have to deal with higher player counts.
But does on time = winner really work for 3p or more when often it's just one bad apple timing out and everyone else is on time?

For most 3p matches on time= winner works exactly how skipping currently works with 1 player being the culprit most of the time. The skipped player gets last place and everyone else tie for 1st.

My proposal for on time= winner would make this same result translate to the prevent skip settings and make it more flexible overall.

In the same 3 player game example with 5 days thinking time per player (like my Vale tournaments) what happens often is after about 9 days one player runs over time and going to the negative with other players having 72 hours each remaining. With skipping enabled there is a good chance that the game will still finish on time if everyone is putting effort to make frequent moves, but neither player with time remaining is at risk of going over time themselves because if they ever get somewhat low on time (<8 hours) the optimal move would be to skip the other player first. I personally would not risk getting skipped myself to extend the game.

With skips prevented in the above scenario there would not be any drama regarding when to skip a player. If one player goes over time each other player should still try to make their moves fast enough to having thinking time remaining after 15 days. If the other 2 players go quickly there is a good chance that the game finishes on time and everyone is content. If one of the other two players plays too slowly they would be at risk of going overtime themselves and also getting a loss. The slow player also has an incentive to complete the game to avoid the auto loss. I feel like everyone has enough incentive to keep the game moving with this system.

What if a 3p game has 2 slow players and 1 fast player? What happens after 11 days if the fast player has nearly 96/120 hours left and the other two players have only a few hours left or one has already gone negative? With skipping both slow players would be incentivized to skip the other player to avoid getting skipped which is some tragedy of the commons - Game Theory scenario. With my "on time / over time" system the game outcome would more fairly reflect players time management. If the game doesn't finish and 2 players were over time limit, both of them are shared losers and the remaining 1 player is the winner. This is most fair since that player tried much harder to play on time. For both players out of time, both of them are incentivized to finish the game to avoid a guaranteed loss so the game should continue moving.

Also consider higher player counts. A 4p game with 30 day duration will quite often have 2 slow players that require over 7.5 days thinking time and 2 fast players that are helping move the game along with no risk of ever timing out. If the game is incomplete after 30 days and the thinking time was 11-11-4-4 days among the players I think it would be most fair to decide the two fast players with 4 days time winners that share 1st place and the two slow players that both ran out of time as losers with shared 3rd place (in the case of this 4p scenario). Trying to sort players by exact time spent like the current system doesn't really feel fair to me. If I played quickly and stayed within my time limit for a full 30 days why should I get penalized with 2nd place because I used more thinking time? The thinking time was entirely out of my control and decided by player order and time zone/play hours more so than being active.

Is this all assuming round robin and/or Swiss? If so, that's something that would be worth trying. But in a bracket tournament with fewer than all remaining players advancing, we still have the old problem of deciding which of the "winners" is actually also going to be a loser.

[nik592]

Players only get a karma penalty if the other players at the table decide to skip them, which doesn't always happen. If you feel your players should be getting through the tournament in the given number of days, then probably not disabling skipping (ie. allowing skipping) is appropriate, but it will still rely on the whims of the players. You seem to think that people who go over time should be skipped/penalised/etc., but I guarantee some tables will and some tables won't and those that don't will have rankings determined by time.

As someone who almost never times out in non-tournament play, but frequently in tournaments, despite checking in on BGA probably half a dozen times during my waking hours, I don't think feel it's fair to say that players who are over time deserve to lose. I'm not sure I've played in one of your tournaments (I don't play a lot of tournaments at present), and maybe your timings are such that people who are on regularly won't go over time, but that hasn't been my experience with almost any tournament I've played. It really depends on how much overlap (or lack of it) there is between players online time.