Board Game Arena

Ursula1972 wrote: ↑14 March 2023, 12:07 I like to open Lucky Number tournaments for 2 players in tournament modus (2 games) with the knock-outsystem. This means that a match could end up after 2 games in a draw. In this situation it is random who continues in the knock-outsystem. I’m thinking about ELO-ranking to decide that. So the player with the lowest ELO ranking after the game continues in the tournament. Maybe there are other games with the same problem (tournaments could end up in a draw), maybe also in games with 3 or 4 players. ELO ranking could apply to all of them?

lucy williams wrote: ↑18 March 2023, 10:16 It encourages more multi-accounting where people set up a second low-ELO account to enter the tournament. Then the law-abiding "single account" players have to play some very good players on low ELO accounts.

JoysJoy wrote: ↑19 March 2023, 14:26 There is not a good way of doing this but using ELO would not fix the problem you described. Just as many times you would find yourself the lowest ELO person in the tie and then think that is an unfair measure.

JoysJoy wrote: ↑19 March 2023, 14:26

Ursula1972 wrote: ↑14 March 2023, 12:07 I like to open Lucky Number tournaments for 2 players in tournament modus (2 games) with the knock-outsystem. This means that a match could end up after 2 games in a draw. In this situation it is random who continues in the knock-outsystem. I’m thinking about ELO-ranking to decide that. So the player with the lowest ELO ranking after the game continues in the tournament. Maybe there are other games with the same problem (tournaments could end up in a draw), maybe also in games with 3 or 4 players. ELO ranking could apply to all of them?

ELO is meaninglyless when it comes to ties. It is based, ultimately, on who has played more on BGA, and is not some objective measure of who is a better player- They could just as well choose whomever joined BGA first or which was the first person to play the game on BGA as the winner since both these events are a huge driver of ELO difference. Someone could be a world master in the game and still have a lower ELO. How about who spends the most number of hours playing games online? Both skill and luck take a backseat to sheer volume of hours of your life spent playing games online. If we had more to do IRL, we would all have lower ELO lol.

For Lucky Numbers, which is won largely by luck of the draw, they could maximize on who had the luckiest draws and discarded the least number of tiles? Or who picked up the least number of tiles from the discard, a clear measure of luck. Who took the least thinking time would choose the one born the fastest thinker.

There is not a good way of doing this but using ELO would not fix the problem you described. Just as many times you would find yourself the lowest ELO person in the tie and then think that is an unfair measure.

Mathew5000 wrote: ↑20 March 2023, 00:26

lucy williams wrote: ↑18 March 2023, 10:16 It encourages more multi-accounting where people set up a second low-ELO account to enter the tournament. Then the law-abiding "single account" players have to play some very good players on low ELO accounts.

The OP is talking about Single Elimination tournaments, in which having a low Elo is a huge disadvantage because of the bracketing. In this tournament format on BGA, in the first match the entrant with the lowest Elo faces the entrant with the highest Elo, and the second-lowest Elo faces the second-highest, and so on. Therefore it would be counterproductive for someone to cheat by creating a new low-ELO account in order to play Single Elimination (a.k.a. "knockout") tournaments.

JoysJoy wrote: ↑19 March 2023, 14:26 There is not a good way of doing this but using ELO would not fix the problem you described. Just as many times you would find yourself the lowest ELO person in the tie and then think that is an unfair measure.

I think maybe you have it backwards? The OP proposed that a tie would go to the player with the lowest Elo.

For a game like chess, the OP's proposal would be unsatisfactory; it would really change the game because one player would be able to strategize knowing that a stalemate is just as good as a win. (The same is true of a tiebreaker such as "most time remaining" because toward the end of the game, even if both players have lots of time left, the player with more time left can "play to tie" knowing that a tie is as good as a win.)

In a game like 2-player Kingdomino, the OP's proposal would work well for Single Elimination tournaments because the game itself already has two tiebreakers, so true ties are rare. "Playing to tie" for the player with lower Elo would almost never be a factor.

I don't know about Lucky Numbers (I've never played).

Ideally, in case of a tie the same two players would have a rematch, repeatedly until one or the other won the game. But that would leave other players in the tournament with a delay, waiting for the tie to be resolved (especially in a game like chess where you could frequently have two or three draws in a row). There is an active "Suggestion" about this in the BGA bug-reporting system: https://boardgamearena.com/bug?id=26098

See also: https://boardgamearena.com/forum/viewto ... 25#p137025 (listing some other threads on the topic of ties in Single Elimination tournaments on BGA).

In my view, I'd like to see the tournament creator have a choice (with Single-Elimination tournaments) for how to deal with ties, after the game's own prescribed tiebreakers are exhausted:
- coinflip (should be the default)
- rematch (with a warning that this may greatly extend the runtime of the tournament)
- player with most time remaining proceeds to next round
- player with lower Elo proceeds to next round
(with warnings on the last two options, that game strategy may be affected in an unbalanced way).

In any event, the tournament page on BGA ought to make clear, when there is a tie in a tournament, why one player proceeds to the next round while the other is eliminated.