Single Elimination: Two-Player Game Finished in a Tie and Newer Account Advanced

[Meeplelowda]

Is it possible for the players to "abandon game" by mutual consent in a single-elimination tournament? I can't remember whether that's possible, and then if so what happens; does one of those players advance to the next round, or are they both ruled to have lost so that their would-be opponent in the next round gets a bye?

Okay I created a tournament to test this out: https://boardgamearena.com/tournament?id=434629 In my first match I'm going to propose to collectively abandon game, and I'm hoping that whoever I'm against, will agree to that. Anybody interested please sign up; I can invite you to the tournament if you don't have sufficient Elo in Reversi.

Are we abandoning both games, or only one and letting the other play out to see what it does in that situation?

[Mathew5000]

Let's try abandoning both. (I'm not sure whether my opponent will agree to abandon.)

[Meeplelowda]

Let's try abandoning both. (I'm not sure whether my opponent will agree to abandon.)

Done. middledragon advanced. FWIW, they have the younger account by 10 days.

[Mathew5000]

Done. middledragon advanced. FWIW, they have the younger account by 10 days.

Ah, so that's the answer: BGA regards both players as winners when you agree to abandon (you both are expressly denoted as "1st" in your game: https://boardgamearena.com/tournament?i ... ge_display ). In my view it would make more sense if both players are regarded as losers, i.e. both eliminated from the tournament with their would-be opponent in the next round gaining a bye.

I would conjecture that for single-elimination format tournaments, the rule BGA applies in case of mutual abandonment is the same rule as if the game were played and drawn. That rule appears to be, a random coin toss, with each player having a 0.5 probability of being chosen to advance. I think it's clear that age-of-account plays no role in what happens in single-elimination tournaments, although it does in multiplayer-elimination, round-robin, and Swiss.

Worth mentioning that in a 2017 post, someone suggested that the rule might be "elo-ponderated random value": viewtopic.php?p=39436#p39436 What they mean is that it's randomly chosen but not a 50-50 coin toss; rather, the player with lower Elo has a greater chance of being selected to advance. We'd need a lot of data to test that hypothesis; probably dozens of data points before we could say whether it appears that hypothesis is true or false, and maybe several hundred before we could be reasonably sure of it.

[Mathew5000]

Hopefully people actually pay attention to the name of the tournament.

Well one player out of the four actually wanted to play Reversi and he won the tournament. (Of course he did have every right not to agree to abandon.) We've seen before, that if the final round is a tie, BGA still arbitrarily chooses one player to "finish 1st" in the tournament: https://boardgamearena.com/tournament?id=365523 which was discussed at viewtopic.php?p=209466#p209466

The thing is, if a "Groups Stage" format tournament has single-elimination for phase 1, there has to be some way to decide which player advances to phase 2. I think a coin toss is better than age-of-account and probably even better than lower-Elo. But I'd rather see the game replayed (even though it would prolong the total tournament length).

I searched the BGA bug-reporting system on the keywords SINGLE ELIMINATION and found this proposal, which I don't think we've discussed before: https://boardgamearena.com/bug?id=132546 "Please document (and improve ?) tie resolution in single elimination tournaments" I have not upvoted it because I think it duplicates one or two other proposals.

I have upvoted this proposal which calls for a replay of any tied game in single-elimination tournaments: https://boardgamearena.com/bug?id=26098

[Meeplelowda]

It just occurred to me that single elimination tournaments award a third place even though the two semi-final losers (I'm only thinking of 2p games) don't play each other for that honor like you would in a football/soccer tournament. Is that also randomly awarded? That seems like a big oversight that should have been dealt with even when tournaments were only meant for 2p games.

[Mathew5000]

In single-elimination, BGA doesn't award a single third place; the two semifinal losers are reported as having tied for 3rd; the four quarterfinal losers are listed as tied for 5th; etc. See for example https://boardgamearena.com/tournament?id=431832 (and tangentially, I always smile at the 32 players tied for "33th"; see https://boardgamearena.com/bug?id=17100 ).

Now, what if you needed a single third-place finisher, in a Groups format tournament where three advance from each group. I don't know whether BGA would choose the third advancer by a coin toss or by seniority-of-account. I'm not really inclined to test it, because if phase 1 is single elimination then it would be pretty silly to have more than one player from each group advancing: what's the point of each group having a final match, and then regardless of who wins, both finalists advance.

[Meeplelowda]

In single-elimination, BGA doesn't award a single third place; the two semifinal losers are reported as having tied for 3rd

Ah. I recently got a third place but didn't check to see what the other semi-final loser got. This makes sense.