Board Game Arena

smattathias wrote: ↑22 December 2021, 14:46 I think players who do not play in good faith are going to do whatever they want to do. They are not considering the finer points of who gets what percentage of ELO points. They are going to abuse the system no matter what the system is. That is why I propose, especially in 2-player games, that game abandonment end in the gain of 100% of the available ELO points for the player who stays, whether it occurs on move 2 or move 200, as well as the loss of 100% of the available ELO points for the player who leaves. Connectivity errors, as someone brought up, are just part of the system. People make the wrong moves in online play, people drop their connections in online play, people cheat in online play, and yet everything continues. Gaining or losing 100% of ELO points for a game isn't going to make a huge difference in any of our game experiences, and it will be more equitable for all.

I agree for two player games, this would make much more sense than the current system. It is especially frustrating when someone ceases to play when their opponent has obviously won, at that point they should either play their final move or resign, and if they don't the winner should still gain full ELO.

King Hrothbeort wrote: ↑22 December 2021, 18:01 I agree for two player games, this would make much more sense than the current system. It is especially frustrating when someone ceases to play when their opponent has obviously won, at that point they should either play their final move or resign, and if they don't the winner should still gain full ELO.

well, i still ain't agreeing, especially on 2-players table where you can play whit more than 2, someone could get master just by constantly kicking the one who is timed out, that's where this system become broken, do it in 2 players? 100%. In a 3+ players of the same game? Tied whit the others and only get whatever value of percentage done before that occured.

So, here all the games that could be broken by this method :

-Saboteur (extention since the extention allow a 2 players game)
-Rallyman GT
-Downforce
-Red 7
-Corridor
-Probably many more that allow a range of 2 or more players per table

TheClaudeQc wrote: ↑22 December 2021, 19:27 well, i still ain't agreeing, especially on 2-players table where you can play whit more than 2, someone could get master just by constantly kicking the one who is timed out,...

You make it sound like it was the normal case that someone is out of time. But it still is the exception - or at least it should be. And even then there's a 20sec-delay for kicking.

Yes, the system is not ideal for a 2-player game. However, changing it will introduce many more issues. (And, please, don't suggest one system for 2-player games, and one for 2+-player games. The potential for problems with that configuration is off the scale.)

Consider a 3-player game where there is no adjustment for number of players, and players have little interaction (such as Can't Stop). Players A, B, and C are playing a game, when player C quits with a significant portion of the game remaining. Call it point X. Players A and B continue to almost the end of the game, when player B quits (point Y).

How much ELO adjustment should A get? X%, Y%, or 100%? If changes are made for 2 players, all 3 values could be argued. In which case, how much adjustment should B get, and does that mean that there could easily be a situation where each player's ELO is being adjusted by a different amount for the same game? This breaks the ELO system more than people think it is already.

Additionally, the arguments on this thread have been based on "the game was close to being finished." While that is sometimes easy to determine by a human, it is much more difficult for a computer. See the previous references to game progression being non-ideal. But, any change to the system must be able to be implemented strictly by computers, with no human interaction.

Honestly, I think that a 'resign' button on 2-player games would make a lot of sense.

IRL, I might play out a game if we don't know it well enough to know how it'll go. But I think that abandoning games is a reasonable thing, specially in perfect-information abstract strategy games.

At a high level, most games of chess/Go are won by resignation. And I think that it often makes sense to resign a game of quoridor or whatever.

Resignation should lose the ELO but no karma imo.

Stuff by Bez wrote: ↑26 December 2021, 15:10 Honestly, I think that a 'resign' button on 2-player games would make a lot of sense.

Already implemented Upper right corner menu, wave the white flag. Conceding a game is only available for 2-player games, and the system treats it as if the game finished normally; Full ELO is awarded and no karma penalty.

Jest Phulin wrote: ↑26 December 2021, 15:36

Stuff by Bez wrote: ↑26 December 2021, 15:10 Honestly, I think that a 'resign' button on 2-player games would make a lot of sense.

Already implemented Upper right corner menu, wave the white flag. Conceding a game is only available for 2-player games, and the system treats it as if the game finished normally; Full ELO is awarded and no karma penalty.

Oh wow! Thanks!

If a player doesn’t finish their game they get an additional loss of 10 ELO points. It’s always better to concede or just lose than to not finish.