Competitive auction (e.g. Name That Tune) seems clearly better, but I'm assuming there's a reason that they don't want to deal with that hassle. Regardless, the idea of bidding for time in armageddon is a big improvement- far bigger than the improvement from sealed bids to competitive bids.

I'd rather go with buchholz (sum of opponents' points) than an armageddon game. Average opponent rating would also be OK but has some issues (e.g. Larry Kaufman only won the World Seniors Championship because he was rated 120 points lower than Suba, because he dragged down Suba's average).

If a playoff is preferred, Curtains' idea has some merit, but i would still prefer 2 regular rapid games before this armageddon bid game.

Huh? But if he wanted to avoid a coinflip he could have just bid lower to begin with. Honestly the idea that it's not fair to randomize on a tied bid really mystifies me. Any one of the participants had their chance to avoid it by bidding one second lower to begin with. He is the one that tried to use some kind of weird psychology to make his bid and because of that he may get punished with a coin flip. It's fair in the sense that the players made their own bed by getting too strategical with their bids, and they know well in advance that on a tied bid there is a coin flip.

It's completely and totally fair to flip a coin on a tie and I have yet to see any logical/rational argument that is even remotely convincing that it's not. I don't even state this as a matter of opinion because I think it's simply fact.

I agree that this is a fair system, and I don't understand the objections to the coin flip.

I think you are incorrect about "strategic play" though. This is just an auction now, and one definitely should play this part strategically in the manner that RT mentioned. One might argue that there are better auctions to use in this case that would be more transparent, by giving better incentives for people to bid their true private value. (I.e., would a second price auction or something work better?)

In either case, the system of bidding seems pretty fair to me, but maybe could be improved on the margins. It'll be interesting to see how it plays out if this happens.

Thinking about it more, I don't really like paying extra money based on the game W/L/D result as opposed to the tournament result- incentivizing the game result messes with the tournament. In lots of sports situations, and even in chess matches, one side will simply not care about losing- it's either winning or nothing. When a hockey team pulls the goalie down 1, they absolutely don't care if they "lose" the rest of the game. When basketball teams foul, they absolutely don't care about "losing" the last minute. In plenty of soccer competitions, one side needs a draw and the other side needs a win to advance.

Fans understand that all of these situations are natural- unless they have a bet on the hockey team to not lose by more than 1 (or the under), they're not sitting there complaining that the team isn't even trying to play defense, or that their soccer team isn't trying to win. They know what the objective is. It's not corrupting the game to play for a nonstandard objective, as long as the players (and fans) know what it is.

I'm not saying that people shouldn't bid strategically, I'm just saying that it's absurd to call it unfair because their strategic bids resulted in the same bid as their opponent, and then claiming that they shouldn't have to be subjected to random selection at that point.

Maybe you are right. Here is the flip side:

1. If you are white, you are now going to be slightly more careful before going completely insane trying to win some absurd/lost position.
2. If you are black you will be less blase about just giving perpetual the moment you can, even if you are up a ton of material and totally winning.

Lots of chess purists and fans may be a bit offended when any of the above occur, as they will certainly make the game look more absurd. I'm not sure what the right choice is, but I wanted to err on making the game that decides the US Championship seem as ....normal as possible.

It's a value judgment- the right choice is the one that achieves what you want it to achieve The k-k and Kramnik-Leko finales were all fine even though they were draw-odds games.

Yeah tough to say, the other way definitely has merit.

wow i rly hope hess wins the tourney

Naka impressive victory.

He's gonna win it all.

He was playing somebody quit a bit weaker than him. Why is it an impressive victory?

Because of the ease of how he won vs. a strong IM with the Black pieces.

What is the pros of a swiss system (is this what is being used?) over round robin?

At first I thought it is that you get to the top players to constantly play each other, but that is wrong since you can't play someone twice?

Fewer rounds. In order to do this year's US Championship as a round robin, they would have had to either make it 23 rounds, or else reduce the number of invited players. Obviously the first option isn't viable for time constraints, so it comes down to whether you'd prefer to emphasize maximum accuracy in the tournament results (which RR emphasizes), or emphasize the human interest stories that a larger field allows for.

Whether all 24 players deserved invitations to the US Championship is a matter for debate (Curtains has made his opinion clear on this in a couple of his videos, hehe), but once the decision was made to invite all 24, a swiss tourney became the only viable way to resolve things within a reasonable number of games. It's not perfectly fair, but it's close enough; at least in the opinion of the people making the decisions.

My idea of a swiss modified which I mention in some of my videos is:

1. First 7 rounds of the swiss are normal
2. Final 2 rounds there can be rematches

Anyway it's an interesting thought I think, some downsides of course but some definite upsides with the key players facing each other in the final 2 rounds.

That's pretty interesting curtains.

I was already thinking to myself how the US Champ. is kinda like Buzzkill in the last 2 rounds. That tournament idea would definitely remove it.

Checkers lets the last round repeat pairings, and I think in the past, they let more than that repeat. It's not a bad idea to let the final 1/3 or 1/4 of the tournament repeat pairings in an over-long swiss (Rounds - int(log2(n)) >= 2 or something). Yesterday would have been Kamsky-Onischuk, Hess-Nakamura, and Shulman-Akobian. Today (ignoring colors) would probably have been Nakamura(6)-Kamsky(5.5) and Onischuk(5.5) vs. Shulman winner(5.5), or Shulman (5) if they drew. That looks a lot better than a tournament leader getting paired 2 score groups down, and not even to the top player in that group (Friedel was viable only looking to board 3).

On a completely different note, why not extend your bidding system to normal chess games? Time control is 120 + 30 or something, and players simply bid for white (well, choice, but without draw odds, you have to be on crack to pay time to choose black) instead of getting assigned colors. It's a different game, but I don't think it's a bad one.

I can't see how this doesn't just come down to an unusual handicapping system? The goal would obviously be to find your equilibrium bid which will mean more time needed versus stronger players, less against weaker players, etc.. In fact you could mathematically determine the proper bid if you could roughly estimate the elo impact of having 'x' minutes less than your opponent.

If so then I think it would be improper for tournament (and especially swiss pairing) play.

Why? Assuming the players bid and evaluate properly, then the expected score is the same with either color and depends only on the skill differential. In, say, a 5 round tournament where everybody is equal, normally half the players- the ones who get 3 whites- are at an advantage due to sheer luck. In the bidding system, the luck of the colors is removed completely.

Yes I've definitely thought of this, it's maybe not necessary perhaps only in the very last round when colors are 4-4 then it makes sense. Because I do think it's quite a bit unfair for naka and hess to get the extra white in the key round based on pure chance.

So my recommendation is that whenever 2 opponents face off in the final round, both having had 4 whites and 4 blacks, or an equal number or whatever, then it makes a lot of sense to use such a system.

Congrats Hikaru! And I love you so much for playing the Two Knights defense with Ng5 in a must win game!!! (I regret not getting to see the Fried Liver attack, but of course 5. ... Na5 is better so I'll take what I can get) fried liver

ggs

sigh, was hoping for some sort of tiebreak, but congrats to nakaaaaaa.

Well without getting into arguments about the theoretical basis of this all - I'm just considering the final result. You're going to end up with games exactly like you have now with one difference. When two players of a substantially different playing ability are paired up, the weaker player will either get to play as white or get substantial time odds. In closely matched games, nothing much will happen as it'd be quite dangerous to give large time odds.

I think this is only unclear because time is an intangible. Imagine if the players were bidding for how many pawns they were willing to give up to play as white - then it's quite clear that it's nothing more than an unusual handicapping system.