(1) Way to go, Ted Castro! I hope you aren’t too disappointed by losing your last match. I can give you nineteen thousand reasons not to be disappointed.
(2) The commentators pointed this out, so I will too: You never saw so many relaxed, happy faces before the last round of a chess tournament. Although there was so much money at stake, everybody who was still playing was already guaranteed of making a bundle. Hikaru Nakamura and Le Quang Liem were guaranteed $50,000 for second in the Open section. Ted Castro and his opponent, P.P. Prachura, were guaranteed $19,000 for second in the under-2200 section. At that point it’s like playing with house money. You’ve already had the best tournament of your life, and you have a 50 percent chance of doubling your income at no risk. Who wouldn’t be happy?
(3) Speaking of the commentators, I give them two thumbs up. Specifically the thumbs up go to Robert Hess and Lawrence Trent. Robert was so much better than I’ve seen him before that I was amazed. Previously I saw him commenting on the U.S. Championship, I think, with Jennifer Shahade, and it seemed as if they were both half-asleep. This time Robert was awake and on point, with lots of energy and good observations. I always like Trent, but he’s got to cut it out with the puns. I mean, we all get it. P.P.’s name sounds pretty silly if you’re in kindergarten. We’re not in kindergarten any more.
I’m not sold yet on the third commentator, Tania Sachdev. I can’t say why exactly. I’d have to watch the tape and think about it, but I’d just say that it seems as if she’s trying too hard.
(4) It was really cool to see the games of 1200 players and 1600 players being given the same treatment by the commentators as the games of the grandmasters. That, of course, is intentional; it’s the whole point of the Millionaire Chess concept.
(5) Bizarre fact, also pointed out by Hess. At the same time as the playoffs were going on, the Open section was still playing its Rounds 8 and 9. You had Fabiano Caruana, Wesley So, Gata Kamsky, and Luke McShane playing on the top two boards — and absolutely no one cared. Well, the players cared, obviously, and there was still a $10,000 fifth prize to play for… but the players who weren’t in the playoff might as well have been invisible on the broadcast, except for one bizarre moment when Kamsky came barging in for some unknown reason. Maybe to protest the lack of TV coverage?
(6) The move of the day came in the first game of the match between Holden Hernandez and Marcin Tazbir for the top prize in the 2400-2549 rating group. First of all, let me just pause to let that sink in. We have two grandmasters playing for a class prize. How sick is that? That shows you how strong this tournament was.
Anyway, they got to this position, with Hernandez (playing Black) to move. Can you see what he did?
Black to move.
FEN: 6k1/1bpp2pp/1p6/3PnP2/1PP2R1P/r3q1P1/2Q3BK/7N b – – 0 1
(7) The missed opportunity of the day — well, there were lots of candidates, but this one stands out for me. Third game of the match between Farai Mandizha and Kaiqi Yang for top under-2400. It’s a 15-minute game. Mandizha, as White, has been suffering the whole game, but he has stirred up a little counterplay. Suddenly Yang gives him an unbelievable gift.
Black to move.
FEN: 1k6/6R1/2P5/p3p3/3r2P1/2KB2r1/P1P5/8 b – – 0 1
Black played 1. … R4xg4?? Can you see how White should capitalize on this mistake?
With less than a minute on his clock, Mandizha couldn’t figure it out and he played 2. Rb7+?? He eventually drew, but he could have won in this position, and it was arguably his last chance to win the match. They proceeded to draw two more games and then Yang won game 6 and the match.
The answer is 2. c7+! If 2. … Kc8 3. Rxg4 Rxg4 the bishop is unpinned and White wins with 4. Bf5+. Or if 2. … Kb7 3. c8Q+! once again forces Black’s king to the bad c8 square. Obviously the players must have missed one of those two possibilities. Looking at their faces after the game, it was clear that Yang knew he had dodged a bullet. I’m not sure whether Mandizha realized he had missed a win, but I’m sure that somebody told him pretty soon.
(8) There’s a little bit more on the controversy over the 9-move draw in the Nakamura-McShane game. First, I agree that the no-short-draws clause only forbade agreeing to a draw before move 30, not playing to a draw by threefold repetition. So you can’t say that Nakamura and McShane broke the contract.
The new wrinkle is that David Smerdon posted a mathematical analysis of what Nakamura and McShane should have done as rational economic agents — taking all questions of morality, sportsmanship, etc. out of the picture. He concluded that because the prize fund was so hugely skewed towards first-place winners, they both made the wrong decision economically. It’s pretty simple, especially for Nakamura. If he plays on, he has let’s say a 33 percent chance of winning the game and making the playoffs. By agreeing to a draw, he put himself in the position of having to win a nine-man playoff just to get to the “real” playoffs. He has only an 11 percent chance of winning that playoff. Even if you think Hikaru was the strongest of the nine, still his chances might be, say, 20 percent at best. (He was, in fact, pretty lucky to beat Wesley S0.) In any case, he had better odds if he kept playing against McShane. I’m not sure that the case is so clear-cut for McShane; his chances were pretty low either way.
Smerdon’s conclusion is that Maurice Ashley, the organizer, has done the right thing by making such huge prizes for first place. But the players haven’t caught up. Nakamura was still in the frame of mind for most tournaments he plays in, where the difference in prize money between first and second is too small to justify taking a chance. If he had rationally worked out his odds in this case, he would have seen that the prize money made it worth taking a risk.
I don’t completely agree with Smerdon’s conclusion. One of his commenters pointed out the problem — people don’t act as rational economic agents. Especially under tournament pressure. You see it a lot on game shows. Suppose you’ve won $500,000 on a game show and now you have to choose between taking the money or flipping a coin for $2,000,000. But if you lose the coin flip you get nothing.
Economic theory says you should flip the coin. But there’s no way I would do it. I have $500,000 in my pocket, a ton of money, and the additional million and a half just isn’t worth that much to me. I can’t even conceive of what I would do with it. The biggest consideration for me is that I don’t want to give away the half-million I’ve already won and end up with nothing.
So I think a realistic economic analysis would have to take into account risk aversion and it would have to take into account that the second half-million means much less to me than the first half-million, the third half-million means less than the second, and so on.
(9) Would I ever want to play in the Millionaire Chess tournament? I think the answer is no, although it was fun to watch. I still hate the idea that you play seven games of classical chess just to win $16,000 (talking about the open section now; other sections proportionately less) and then you play rapid and blitz playoffs to win $84,000. This sends the message that 25-minute chess is five times more important than classical chess. I do not agree.
Nevertheless, Maurice Ashley’s tournament format is great for TV, and that is a big part of his scheme. If we could ever, somehow, persuade a network like ESPN to broadcast the Millionaire Chess tournament, what a game-changer that would be.
(10) Speaking of mainstream sports, the title of this post is stolen brazenly from Peter King’s column that appears on the Sports Illustrated website every Monday during the football season. It always ends with a section called “10 Things I Think I Think.”
Answer to quiz: Hernandez won with the queen sacrifice 1. … Qxf4! Tazbir resigned, but the rest of the combination goes 2. gf Ng4+ 3. Kg1 Ra1+ 4. Bf1 Rxf1+! 5. Kxf1 Ne3+ and Black wins the queen and ends up a piece ahead.
{ 15 comments… read them below or add one }
(3) I give the commentary a B+. I wish Lawrence Trent would dial it back a notch, and I got a little tired of IMs (plus one GM, admittedly) aided by computers being so agog at how terrible the GMs’ blunders were. All three of them also need to learn not to automatically say “Absolutely!” all the time, too. But, my gripes aside, it was an entertaining broadcast. Unfortunately, it kept crapping out on me, but the fault may have been on my end.
(4) Agreed; I thought it was really charming to see the commentators demonstrating such excitement about the class games.
(5) I didn’t need video, but I couldn’t even find the moves for the last two rounds of the open anywhere! It was really odd how they completely disappeared.
(8) I thought the big problem with the draw rule was that there wasn’t even a clear rule (and I’m not sure there could be). The regulations say “We will not be forcing a player to walk into checkmate or lose a pawn to avoid a draw.” OK, will you force a player to move into a position with an eval of -0.9? How about -0.8? -0.7? -0.1? And which player do you force to end the repetition? And as far as penalties go, you can give both players 0 points but then they could just end up in an endless game of chicken, continuing to repeat moves, neither one claiming a draw but both refusing to move on. You could start adding rules like “If a player makes a move that repeats a position for the third time, s/he loses the game”, but that’s not chess anymore.
I thought Ashley did more to deter sponsors by his rant about short draws being a “scourge” and a “stain on our game” than the players did by agreeing to a short draw. There were plenty of other exciting games that round anyway…
(9) Yeah, it was weird how the most important games were all rapid (or quicker). The World Cup matches often get decided by fast games, but at least they have the opportunity to play slow ones first. The knockout format certainly avoided the grandmaster draw problem, though!
“Economic theory says you should flip the coin.” Actually, economic theory doesn’t tell you categorically what one should do, precisely for the reason you state later: “So I think a realistic economic analysis would have to take into account risk aversion.”
That’s exactly what economic theory does do. It says you should maximize expected utility, not expected payoff. For flipping a coin to maximize your expected utility requires that your coefficient of risk aversion be below a certain threshold. (If you’re risk neutral, you’d flip; if you’re very risk averse, you wouldn’t. At some intermediate level of risk aversion, you’re indifferent between flipping and not.)
Yes, my statement was a little bit too simplistic. However, my point is that Smerdon’s approach was likely this same simplistic one, where the utility and the expected payoff are the same. (So a 50-50 chance at a million dollars has equal utility to a guaranteed half-million.) I don’t know for sure if that’s what he assumed, but from his blog post I suspect so.
Hi Dana,
Great analysis.
I liked Robert Hess. I agree that Trent should probably dial it back a bit.
The thing that bothered me about Tania’s commentary was that she kept saying things like, “Wesley So has just wrapped up his game” or “And now it’s completely over.” Ten seconds later, the match would continue to go on. So she’d tell us that it was over, when in fact it was still very competitive. I think once she stops doing that, her commentary will improve a lot.
I completely disagree with David Smerdon. McShane was on fire this entire tournament. It would not be an easy win for Hikaru–not even an easy draw. By agreeing to the draw, Hikaru guaranteed that he’d be playing rapid or blitz for the rest of the tournament, and as everyone pointed out, with the possible exception of Le Quang Liem (who was already past the playoffs and was therefore no longer a threat), Hikaru had no equal in blitz or rapid. Wesley almost won, but if you look at their across-the-board history, Hikaru usually wins. So yes, I think Hikaru had better chances of winning a nine-man rapid/blitz playoff than a single classical game (as black, no less) against Luke.
Even if Hikaru is the best rapid/blitz player, I think it’s tough to give him more than a 30% chance of surviving a 9-player playoff. (Even that 30% basically means that he has 4 times the chance that anyone else has.) So he would need to have a better than 30% expected score in his game with Luke for it to make sense to play on, and I think in reality his expected score is probably not less than 50%, even playing Black (he does outrate him by over 100 points). So I think the most rational decision was still to keep playing, although I agree that it is closer than it may seem at first.
Hi Dan,
Thanks for your comment.
I think it’s far more than 30% that he would win the 9-person playoff. I’m guessing that before his match with McShane, Nakamura probably already had decided whether he was playing for a win or a draw. If he was going for a draw, it’s probably because he thought his chances in rapid/blitz were greater than 30%–and that he also thought it was unlikely that there would be decisive scores in 100% of the other matches between people who were tied with him. Considering his record in rapid/blitz, I think it’s a fair assessment of his own skills.
Think about this: Magnus Carlsen just won the World Rapid Championship over a field that was a lot greater than 8 other people. He was the Rapid champion last year too. Before he won, would you say that Magnus’s chances were only 30%? Sure, Nakamura, So, and Liem weren’t there, but Carlsen still had to beat MVL, Kramnik, and Nepo. Yet he still won, repeating his performance from last year. I agree that Magnus’s chances of winning were not guaranteed, but I’d still say that it was greater than 50%.
I think Nakamura’s dominance in short time controls is similar, which is why he went for it. I remember watching an interview with Nakamura where he was asked why he was better than everyone at blitz/rapid. He said that his chess skill might be similar, but he was better than most people at guessing his opponents’ next moves. This probably plays a bigger role when the time control is short.
Yeah, it’s a tough job to come up with a good estimate. One reason that I downgraded Nakamura’s chances is that there was likely to be at least one knockout component to the tiebreak. I think it is much easier for the top player to win in a round robin or long Swiss than in a knockout, where there is no room for error.
That’s a very good point– there is more luck involved when it’s there is (or is likely to be) a single-game knockout involved!
(8) Your analysis implies that Nakamura had to make his decision under pressure. But he knew his score after the previous round – he would have had all afternoon (or all night) to think it over. Given the experienced poker player that he is (and poker is ALL about calculating odds and computing payoffs), making the wrong decision is surprising. Perhaps he ran the same calculation but estimated his rapid/blitz chances higher than Smerdon did.
I think you completely miss Maurice Ashley’s objection, which is about the spirit rather than the letter of the law. 9-move draws destroy the game’s credibility in the eyes of the non-expert public, who must ultimately fund chess if it is to reach the big leagues.
Ashish: “I think you completely miss Maurice Ashley’s objection, which is about the spirit rather than the letter of the law. 9-move draws destroy the game’s credibility in the eyes of the non-expert public, who must ultimately fund chess if it is to reach the big leagues.”
If only Maurice had left it at this. Instead, he falsely asserted that Naka/Shane broke the rule. That’s defamatory. He had no basis to seek to reverse their draw. He then came back from his consultations with external arbiters, saying that they said Maurice has no basis for a reversal because he can’t prove either “intent” or that they had an agreement.
This highlights two different notions of “agreement” with respect to a draw. There is a typical “draw by agreement,” which isn’t relevant here. You can also imagine that two players collude before the game to play a particular sequence ending in a three-fold repetition. Clearly this would be illegal. (Though I can’t offhand point to the FIDE rule that governs this.) That Ashley said he can’t prove an agreement, almost suggests that he suspected this, and merely doesn’t have sufficient evidence, which I think would be outrageous.
To Naka’s and Ashley’s credit, all of this was put aside by the time of the award ceremony. Maurice spoke highly of Naka and Naka made a very sincere call for thanks/applause for Ashley and Amy Lee for all they’ve done for the tournament and chess.
Great reporting and opinionizing, Dana – many thanks!
I attended the event on Sunday and Monday, and there was indeed a lot of excitement. A few comments and a correction seem in order.
First, Hikaru did indeed risk missing a playoff entirely: had Bareev found the win of a piece early in his game, he would have qualified for Millionaire Monday, and Hikaru would have been tied for 5th, instead of 4th. (Also, it hardly looks like he saved time and energy by not giving himself a chance to outplay McShane. None of the other three finalists had to play after round seven, while Hikaru had to play late into Sunday night, under tremendous tension.)
Maurice is developing the sporting value of chess as he sees it, so his denunciation of the 9-move draw is based on the idea that this kills the value of the event for the audience, and thus for potential sponsors, to at least some degree. My suggestion for a new rule to discourage this, to add for next time, is to introduce a financial reduction in prize money (I suggest 10%) to be applied to any player whose game ends in a draw in less than 30 moves for any reason other than by agreement (which is already prohibited).
And yes, Maurice and Amy are clearly aiming to put chess in ESPN, if not on a major network!
I second the idea in your second-to-last paragraph. I think that the remedy is to be found in financial incentives, not in rules. One virtue of this approach is that the decision can be made by the organizer, without consulting FIDE arbiters.
I would propose one friendly amendment. The length of a game with a 3-fold repetition should be determined by the move when the repeated position *first* appears, not when the draw is claimed. So Nakamura and McShane are not allowed to just keep repeating the position 11 more times until they get to move 30, because the position first appeared on move 5.
In addition, I’d like to combine my proposal with yours. In the event of a short draw, you and your opponent will face a 10 percent prize penalty, *unless* you agree to replay the game with colors reversed and time already used deducted from the clocks. If the second game also ends in a short draw, then you will have to pay the 10 percent penalty (no more second chances). If it ends in a long draw or a victory for one of the players, then the players are “forgiven” and there is no penalty.
Now we’re getting somewhere! I like the “soft penalty” of reducing prizes, which makes it easier to apply across the board no matter how small the violation, rather than the hard penalty of reducing points, which makes it hard to draw a bright line between cases that are and aren’t worthy of being penalized. Counting the repetition from the first instance of the position is also a brilliantly simple idea that I’m ashamed I didn’t think of independently, removing all the “repeat N times” loopholes at a single stroke.
I like that in the current case, Nakamura and McShane could have gone ahead and drawn just like they did, and Ashley could just publicly shrug and say “Well, okay, they chose to be cowards and throw away money,” with everyone being able to walk away feeling like they got a reasonable result.
This proposal wouldn’t mess with the rules of chess, which is a big plus, not mention that as Dana points out it has the benefit of not crossing FIDE. Hopefully Maurice is reading!
This idea is GREAT. I like this idea, alone, better than Dana’s suggestion to replay the game with reversed colors and remaining times, because it is just simpler. For example, supposed I spend 30 minutes thinking before accepting an early repetition. Is it legal to FIDE rate a game with uneven times?
It’s worth noting that any new rules about 3-fold repetition early in a game would also affect draws by perpetual check. (Perpetual check isn’t an independent standalone basis for a draw. It implicates either the 3-fold-repetition rule or the 50-move rule, whichever comes first.) I don’t think that a draw by perpetual check would likely be a boring-for-spectators game, even if it’s a short game.
{ 1 trackback }