Today’s twelfth game of the world championship match — the last “slow” game — ended abruptly with Magnus Carlsen’s offer of a draw to Fabiano Caruana, in a position where Carlsen had (a) good attacking chances, (b) almost no chance of losing, and (c) a substantial time advantage of 30 minutes versus 10 minutes for 10 moves.
Until now, I have not joined in the complaints about a chess match where all the games had been drawn. If you watched the games, you could tell that they were serious battles. Game ten especially was an exhilarating high-wire act and near-death experience for both players, which I think may have had a strong effect on their psychology in the last two games. I don’t mind a draw when the game is hard-fought and is allowed to run its normal course. But in game twelve, that isn’t what happened. The fun was just getting started. We were all looking forward to seeing Carlsen play a genuine attack, and waiting to see if Caruana could pull off another Houdini act.
And then, whoops! Carlsen pulled the plug. It was as if we had set up the Christmas tree for Santa’s arrival, and then the power went out and Christmas was cancelled.
From a strategic point of view, it’s possible Carlsen buys into the conventional wisdom that he is a better fast player than Caruana. So basically his objective today was to draw, not to win, thereby sending the match to a playoff (where the initial time control will be game in 25 minutes). If that was his strategy, it is a very dangerous one. In statistical terms, maybe his expected winning percentage is greater in a rapid game, but the variance is also much greater. If there are chess gods, they will punish Carlsen for his hubris.
(I don’t know if there are chess gods. In the television game show Survivor, there are definitely Survivor gods. Any time you see a player get too cocky about his winning chances, it’s almost a guarantee that he is going to be eliminated at the next Tribal Council. But that’s partly because the producers edit the show that way.)
The alternative is to believe that Carlsen really didn’t think he had very good winning chances. That’s what he seemed to say in the postgame interview. “I think my position is not as good as it looks,” he said. But this is astounding. First, what does it mean for a position to look good but not be good? What if it looks good after five minutes of analysis? What if is still looks good after an hour of analysis? What if, in addition, it looks good after Alpha Zero or your favorite supercomputer says that Black is ahead by 0.8 pawns?
According to an article on chess.com, the world computer championship will be suspended for a day in order to have all the computers play games against each other starting from the final position. I don’t know what the point of this exercise could possibly be, except to embarrass Magnus Carlsen. But perhaps he should be embarrassed. There is a standard we hold world champions too, and it’s higher than the standard for ordinary players. To be a true champion, he owes it to the chess world to be fearless, to show us how we should all play this kind of position.
This morning Gjon Feinstein and I hosted our fourth world championship watch party. I have been so thrilled by the interest and enthusiasm of our young listeners (today we had four kids and one adult). When the previous games ended in draws, we were able to explain to them why they were draws, and the kids learned from that. But this time, I couldn’t explain anything to them. It was the first time in this match that I felt embarrassed.
Of course, I understand the argument about pressure. None of us can imagine, supposedly, the pressure these two players are under, with the world championship depending on every move. But the argument cuts two ways. World Champion. Higher standard. What makes you worthy of the title is the fact that you can cope with pressure that ordinary chess players can’t even imagine.
Matches and games like this one always lead to a chorus of people on the Internet suggesting gimmicky and contrived ways to cure chess of its supposed curse of draws. I almost always ignore these, because they’re worse than the disease they are trying to cure. The problem isn’t with chess, it’s with players who don’t want to fight, and no gimmick will solve that problem. And it hasn’t even been a problem in this match, until game twelve.
Still, if you want to talk about changes, here are my thoughts.
- Give the champion the draw advantage. No, no, no, no, no! This is what Bobby Fischer fought against. It creates an unfair fight, where one player only has to score 6 points and the other has to score 6½. Unfairness is a worse sin than boringness. Some people say that the champion has “earned” the right to have draw odds — but that’s nonsense. Botvinnik shouldn’t get draw odds against Fischer because Botvinnik beat Tal. Different opponent, different time.
- Speed playoffs followed, if necessary, by Armageddon. That is the current setup for this match. I see the speed playoffs as a necessary evil, but Armageddon is a completely unnecessary evil. It once again sacrifices fairness to obtain a falsely decisive result. We try to even the odds in the Armageddon game — we give White a time advantage and (of course) a move advantage, but give Black the advantage of playing for a draw. This is wrong because, first of all, it distorts the strategy of chess, and second, because we have no idea what is the right balance to ensure fairness. A 5-minute to 4-minute time advantage? Or 5 minutes to 3 minutes? The balance may be different for different players. And what if, as in this match, Black actually has equal or better chances to win? This throws off the whole Black/White calculus. This actually happened when I played against Juande Perea for the Santa Cruz Cup many years ago. We played six playoff games at increasing speeds, and Black won every one. Under the circumstances, whoever drew Black for the Armageddon game would be a huge favorite. Juande and I decided to “just say no.” We didn’t play the Armageddon game and we agreed to be co-champions. Which brings up another possibility…
- Just declare Magnus and Fabiano to be co-champions. You can call this the Mackenzie-Perea solution. It’s completely fair for now, but the trouble comes in the next world championship cycle. Who does the next challenger play against? Should we have a three-way round robin? [Actually I kind of like this, except from the logistical point of view — a three-way round robin takes three times as long as a two-player match.]
- Go ahead and have a speed playoff, but with the stipulation that no player can repeat the same opening move as White that he has used before. Of the suggestions I’ve read, this seems the most intriguing to me. The idea is to force players out of their preparation — eventually they are going to be forced into unorthodox openings like 1. b4 or 1. g4. The good thing about this idea is that it is not in any way distorting the game of chess. It also tests the players on the whole opening repertoire. The bad thing is that it doesn’t address the fundamental issue, the players’ willingness to fight.
- In case of a tie, decide the world championship by a vote of the living ex-world champions. Okay, I can hear Fischer screaming over this one, because in his era the Russians would always win the vote. Probably you should exclude ex-world champions who come from the same country as one of the two competitors. This solution might in fact address the real issue better than any of the others, because if the players knew that their games were being judged by a jury consisting of Anand, Kramnik, Kasparov, and Karpov, they might try a little bit harder to play beautiful and fighting chess.
In sum, all of these “solutions” have substantial problems. The one change I would definitely recommend, which as far as I can see has no downside, is to lengthen the classical part of the match. Even a 16-game match would, I think, feel quite different. It gives the players more time to experiment with their openings, and you wouldn’t have the sense that a single loss might decide the whole match. Plus, chess fans would get more games to watch!
For now, all of this discussion is academic. On Wednesday, we will have a very exciting round of playoffs, and at the end of the day we will at least have a consensus “first-among-equals.”
{ 19 comments… read them below or add one }
The debate about what happens if two players are tied at the end of a World Championship match has gone on for years. I am not convinced there is ANY perfect solution. More games increases the chances we’ll have a decisive result but it’s still possible we won’t. Rapid, blitz and Armageddon playoffs produce a decisive result but not in the format the match was designed around. It’s meant to be a classical match and there’s a reason we have different ratings for classical, rapid and blitz. It’s because some players are better at one or more types of time control than others. Indeed, Caruana is the classic example. He is relatively weak at blitz by Super-GM standards.
So, what other solutions are there? One that Peter Svidler brought up is intriguing: have the playoff at the START of the match. That way, both players know who will be declared winner if the match is a tie. I really like this idea and it would add a lot of intrigue to the match with players forced to take more risks (but still under classical time control) as the match progresses. Does it unfairly influence the play of the player who loses the pre-match playoff? Maybe but it’s no less unfair than most of the other suggestions I have seen people banding around. Also, I would not have Armageddon be involved in any way. Just have the playoff be four rapid games followed by a “sudden death” blitz out similar to a penalty shoot out in soccer. Two games will be played and each player gets one white and one black. If it’s tied at 1-1 after those two games, we play two more. And we keep doing it until it there is a result. It’s blitz… it won’t take that long.
All of us chess players are way too smart to do anything simple – and we love to be clever. But the one thing we all know is proven is the 24 game match. So why quibble about a few fewer games? And why be clever about moving things around to change where pressure is from a playoff at the end to at the beginning (when no one is warmed up and worn down and in a state of flow, anyway)?
And about Magnus’ intention to simply shut down the game and move to the tiebreaks, I attended the final game in NY two years ago, in which he did that in the opening as white (as he did in game 11 here, too), and he explained afterwards that in the tension of a single winner-takes-all game, a single blunder could throw it all away; he chose instead the pairs of rapid and perhaps after that blitz games, in which a loss in the first game could still be made up by winning the second game of the white/black pair. I should mention that the organizer was so embarrassed by the non-game of game 12 in NY that he committed to admitting all game 12 ticket holders to the playoffs two days later, so I was able to attend the travesty of seeing the WC title up for grabs in fast tiebreak games.
Moving the playoff to the start of the match isn’t intended to be “clever”, it’s intended to produce a winner at the end of the normal time control games. That it would add intrigue to the match itself is a bonus in my opinion.
I have no issue with the 24 game format but it still doesn’t address the fundamental question of what happens if the match is tied at the end? We could have them play 100 games and, although far less likely, a tie could still result. By playing the tiebreaks at the beginning you are greatly reducing the chances of the classical game score being tied. That just makes common sense because the player who loses the tiebreaks will have to play for a win. I don’t see anything wrong with that and surely it’s better than the current system.
What alternative suggestion do you have that would resolve the question of what to do in the event of a tie?
We’ve lived with ties in WC matches before – from Botvinnik-Bronstein to Kasparov-Karpov. If two players are so evenly matched that even a long match doesn’t separate them, we’ve accepted that in the past. The issue here is how to ensure a very good chance that enough of a difference manifests itself and we can get a decisive result without compromising the quality of play, which should be able to be of the highest order. The games between the world’s best players should be timeless, not blunder-filled blitz throwaways.
And if you are so enamored of the tie-breaker at the beginning, how about proposing to FIFA and the IOC that penalty kicks be conducted prior to the start of soccer games. Or shootouts prior to the start of NHL games. (Or maybe, in between halves of soccer or periods in hockey.)
Except we haven’t entirely accepted that, have we? If we had, this issue would have never been a discussion to begin with. In the event of a tie in the past, the Champion always retained his title. We basically told the challenger that sorry, although you proved you are equal to the Champion you can now get lost because we’re declaring that you lost the match. Even though you didn’t.
This has nothing to do with one-off soccer or hockey games that are decided by penalties. The Svidler (or Kramnik or whomever) suggestion is that we have a mechanism in place to decide who the winner will be after ALL the games are completed and if the score is tied. It’s not at all comparable to soccer or hockey.
If you are so enamored of the past practice of accepting a tie but declaring the current Champion the winner, how about proposing to FIFA and the IOC that in the event of a tie/draw at the end of normal time the higher ranked side is declared the winner? You can see how ludicrous this idea is when you flip the coin to the other side, eh?
Having one side with a tiebreak advantage before the match starts gives that side an incentive to draw the match – which in 24 games is harder to control than in 12. My soccer and ice hockey analogies were intended to highlight the difficulty of letting one side be favored going into the game, and now both of us have been able to put up a straw man and say “wouldn’t that be ludicrous.”
OK, I now understand that you object even in the case of 24 games to let the champ keep the title, whereas it seems to me that at some length of match, we reach the point of diminishing returns – or of absurdity, as in the five and a half months in Moscow that caused reversion to 24 game matches again, post-Fischer.
I did see one funny suggestion, which Emil Sutovsky attributed to Grischuk: unlimited match to six wins, but if the following cycle produces the next challenger before the match is over, both players are out. 🙂
That’s true that having one side with a tiebreak advantage before the match starts does give them the incentive to draw the match. But there’s no getting around that, at least not with a two player system. If we allow the Champion to retain his title in the event of a tie, we give the Champion an incentive to draw the match. If we go to rapid/blitz/Armageddon tiebreaks at the end, we may give one player an incentive to draw the classic portion of the match (which is what’s happened in the Carlsen-Caruana match) if they fancy their chances at rapid/blitz more than classical chess. We would have to eliminate draws in chess to properly resolve this issue, something that I have seen suggested by others (and I am not personally in favor of) but that’s another topic for another time.
Got to admit, I do feel a little smug today…
https://www.chess.com/news/view/batavia-amsterdam-chess-tournament-fights-against-draws-with-blitz-before-classical
This is exactly the kind of idea I was talking about. Let’s see if it’s successful, albeit this is on a per game basis.
a jury of world champions would viscerally know a simple fact that negates the beautiful chess argument: each game is a single round of the overall fight and cannot be disconnected from the overall match situation. this is the same BS of why tournament directors get their underwear in a bunch over last round draws that lock in a first place finish. what all of us chess players know is that a wcc match is a different animal from an elite round robin, which is different yet again from a continental FIDE swiss with 200+ players.
that you can’t explain why the draw happened is because you’re thinking one game at a time. that’s not thinking like a champion or a tournament winner, that’s thinking like a chess player. we get so wrapped up in finding the perfect move that we’ll lose the damn game…or match, or tournament. classic analysis paralysis. magnus is playing a bigger game than we are.
I heard that idea of doing the tie-break before the match (I actually think it was Kramnik’s idea). At first it sounded too weird but it made more sense as I thought about it.
But I’m starting to think the idea of a single match to determine the world champion is obsolete in this age of super-computer preparation for a known opponent. In the last few WC matches, we’ve seen this trend of avoiding risks, extreme opening preparation and lack of decisive games. Maybe the WC should be the winner of a series of strong tournaments around the world, similar to the current Grand Chess Tour. That way, we see the player who’s the most versatile (since they’d be playing against many different types of opponents), and the strong players would have to take some risks to win the tournaments. Just a thought…
I didn’t realize it was Kramnik’s idea. Svidler was the first person I heard mention it. Regardless, I think having a tiebreak before the match would eliminate the avoiding of risks surely? The person who lost the tiebreak won’t be able to play that way, which will force his opponent to be more adventurous too. After all, there is nothing more dangerous in chess than playing for a draw.
I’m starting to think more seriously about my initially half-baked proposal of a three-way tournament. The culmination of every cycle would be a three-way tournament between (1) the current champion, (2) the most recent past champion, and (3) a challenger who qualifies from the Candidates Tournament, and is from a different country than (1) and (2). Here are the main points.
1. TIES WOULD NOT BE BROKEN. In case of a tie we would have co-champions, and in the next cycle we would have a three-way tournament with the two co-champions and a new challenger. A three-way tie is extraordinarily unlikely, but if it happens then we have a four-way tournament in the next cycle (the three co-champions and a new challenger).
2. This system brings back, in an organic way, the idea of deposed champions getting a rematch.
3. The greater time duration of a three-way tournament seemed like less of a problem to me, once I realized that the days off can be rotated. So each player will have one day off out of three (just as in the current system) — but spectators will get one game every day! This is kind of nice, for keeping up interest in the match, as there are no off days. If we had 8 games between each pair of players, the tournament would take 24 days, the same as a 16-game match between two players. Alternatively we could have 12 games between each pair of players in 36 days, the same duration as a traditional 24-game match between two players.
4. Another benefit for spectators is a little bit more variety. Instead of repeating the same two openings (Petroff and Sveshnikov) over and over, we would get different openings and different styles each day.
5. The one danger in a three-way tournament is the possibility of collusion. That’s one reason for specifying that the challenger must come from a different country. But I really think we have to rely on champion’s honor here. Playing for a draw is one thing — but a champion or challenger who lost a game on purpose would become a pariah for life.
6. Finally, let me emphasize: no tiebreaks, no speed chess, no Armageddon. The world title would be decided by honest classical chess.
Well, the 3-way final is fun idea, except for (at least) one clear counterargument, and it is a strong one. Consider this: under the new system, there could essentially not be 2 top-3 players in the world from same country. In other words, what if a top challenger, perhaps even a top rated player, is automatically excluded of the final because of his nationality?! Surely that should be seen as unfair to the invidual. After all, the playing system is only serving everyone for deciding who the best invidual player at any moment is.
Yes, the more I thought about it the more I agree. It certainly could happen that a new WC-quality player emerges in the same country as the current WC, and my restriction on the countries would prevent him or her from ever getting a match.
If we take out the country restriction, I wonder if there is any other way to reduce the incentive for player C (who has fallen too far behind to catch up) to lose to player A (his friend) deliberately in order to keep player B from winning. Maybe not.
I think Botwinnik was a 3-way finals advocate.
That’s very interesting. Thanks for sharing.
One idea I haven’t heard mentioned yet is, in the event of a tie match, to split the rewards, with the defending champion keeping the title and the challenger collecting the winner’s share of the purse.
I don’t remember if it is in his 2nd volume of Analytical and Critical Works (1942-1956) or in the 4th volume (which is an assortment of articles), but after capturing the title in 1948 he developed a detailed proposal for the title competition format.
The ideal format is the 24 game match, and the champion keeps the title in case of 12-12.
Considering the “tiebreak before the classical match”, I first thought it was an interesting idea. But then somebody somewhere online said it is like putting a carriage in front of the horse. Indeed. It´s like getting rid of one`s teeth in order not to have nightmares about it.
The reasoning that the match would then be decided on classical chess is ludicrous. Why? Because we already had the fast games partially deciding the match…
Think about this: what if the loser of the tiebreak-before-the-match should lose one game? Then he would be in the same (or even worse, since the amount of games would be limited) situation that Fischer demanded for himself against Karpov: that the challenger might have to win by 2 whole points! That would essentially kill the match considering the defensive skills of nowadays players.
There is another solution, or rather a version of the old that I haven`t seen yet.
How about a final match, where only wins would count for deciding the title. We have been there yes and it didn`t always turn out so well, right? Well just adjust it to our times. Instead of 6 wins needed there would be just 3 wins needed. Sounds better?
Maybe the fear would now be that the match would be too short! But surely there would still be plenty of draws available and probable end result would still be about 12 games at minimum (in an even match, if just half the games are drawn the match would last already 10 games). A possible 2-2 might produce a drawfest to follow, but this is hardly an issue, perhaps time limit could be reduced step by step etc.
As it is said, there is no perfect system. But that means the current one is hardly the best either as changes have always happened and will happen. Armageddon for one is not a great solution as mentioned, just slightly better than drawing of lots.
Hmm I guess the most obvious way of minimising both collusion and lack of interest from being too far behind is to create two prize pools: one for the general 1st, 2nd, & 3rd places; plus a prize awarded each round if there’s a win.
I would also reiterate that precluding certain players only by country doesn’t feel right. Its fairly easy to envision a Carlsen, Caruana, So matchup.
I like this idea though, there’s a lot to be said for it.