The Upcoming Candidates Tournament

The tournament I think everybody’s been looking forward to this year will start pretty soon. Information can be found all over the place, although as I’m partial to Wikipedia, I’ll link to their entry.

ChessBase has published some nice profiles of some players already: (1) Peter Svidler, (2) Vassily Ivanchuk, and (3) Alexander Grischuk. I’m not sure how they’ll get to everybody before the tournament actually starts on Friday, March 15th, but maybe they’ll start double-posting entries this coming week.

These profiles are nice both because they show the full head-to-head history,  but also because they remove those pesky rapid and blitz games that pop up in the simplest of database searches. That’s especially unimportant here I think, given that any rapid tiebreak is only if there’s a tie for first and a bunch of other mathematical tiebreaks come out even.

There were a couple surprises to me in these 3 profiles. While I knew Chucky was Magnus’s customer and that Grischuk has played the top guys pretty close (albeit without any/many decisive results in his favor), I was quite surprised about Svidler’s relatively good scores against almost the entire field. Some of that is probably because he hasn’t played as much recently with Aronian and Carlsen, but it’s still an achievement. I remember first reading about him after he beat Kasparov in their first two head-to-head encounters I think (Tilburg and some other tournament) in the mid-1990s – by most accounts, he certainly had the class, if not always the drive and work ethic, to hang with the guys at the very top. (That game can be replayed here, or read with Seirawan’s annotations here)

As for predictions, I think the smart money has to be on Magnus. In conversation with a couple other players, I put Magnus’s odds of first place (clear or shared) at about 45%. After that, I had Aronian at about 25%, Kramnik at 20%, and the rest of the field at 10%. That was before the Zurich event, but even with that, it’s hard to say what form Kramnik and Gelfand will really be in for the Candidates.

Still, I’m a numbers guy, both in inclination and in terms of my actual job these days, so I was interested in what some simulations might show. Last time around, I collected everybody’s classical game results against 2700+ opposition, took their color results, and simulated the candidates matches.

What was interesting then was that the short matches coupled with the randomness of tiebreaks meant that while I felt Aronian was a reasonable (but not strong) favorite going in, his odds of actually making it through the match cycle were much lower than I expected. With longer matches, maybe it’d be different.

In a tournament though, you aren’t eliminated after a drawn mini-match and some random blitz loss. To me, this means that somebody like Carlsen is likely to do much better in this format than in a match format; that’s because there’s no real issue if he draws a mini-match with Kramnik for example. It doesn’t come down to rapids or blitz, and meanwhile, he gets to snack on some of the lower-rateds where he is expected to do better than his expected competition.

A tournament is also easier to simulate than a series of matches with potential tiebreaks. But before I did it myself, I looked around online and I stumbled upon a site called Chess-DB, which has a tournament simulator. I plugged in Carlsen’s rating from the January 2013 and March 2013 rating lists (handily listed on Wikipedia!) and lo-and-behold, I got a distribution of likely finishes.

Amusingly, with the March 2013 ratings, this simulator gives him a 42% chance of taking first place. Not so far from what I had pulled out of the air. However, with Jan 2013 ratings, his odds drop to about 29%.

I’m pretty sure that when I said 45%, I took into account the  results that took him from 2835 to 2782. Still, thinking about it more now, I’m not sure if he’s really gotten that much better compared to this kind of field. Even if he wins, I don’t think he’ll exactly run away with it – at his dominant Corus/Tata Steel 2013 performance, 5 of his 7 wins came against sub-2700 players.

2 of these set of 3 numbers align pretty closely; the 3rd also makes sense given the input parameters. A 4th set of odds from Anish Giri’s comments at a German site (excerpted here, h/t TheChessMind) though is way out in left field by comparison – he seems to give Magnus odds of 80%!! Now maybe he truly believes this, but if you think about this prior probability as the odds you’d set as a betting line so as to be indifferent between either side of the bet, I’m quite sure Vegas would not agree. Nor would most people, I imagine.

7 responses to “The Upcoming Candidates Tournament

  1. The criticism of Magnus, if there is one, is that he is relatively a “fish-beater.” The opposite of Anand. So against a no-fish field, Magnus’ rating would be (relatively) overstated.

    It would be interesting to calculate the players’ ratings for the last few years against only 2740+ opposition and use that in the simulator.

  2. I think what most of us can agree is that if Anand (and, say, Karjakin) were added to the field, no one would expect Anand to win. Which is a sad commentary on the system.

  3. Yeah, that’d be interesting to see – maybe I’ll take a look at that later this week.

    And I agree that Anand’s chances in a similar tournament (if say you added him and Karjakin in place of Svidler + Ivanchuk) would be relatively small. It’s certainly non-zero, but he’d be the favorite only to his die-hard fans. He’s probably won more games against 2700+ opposition this year than he won in 2011/2012 combined, but he hasn’t won that many nor been that impressive overall. Still, I’d say his rating at 2780 or so is pretty fair and indicative of where he stands tournament-wise with his colleagues.

    In any case, chess fans seem to be a divided lot about what that all means. On the one hand, there’s a huge uproar when the title is determined by tournament play (e.g., the San Luis or Mexico City tournaments in 2005 and 2007) as opposed to tournament play (so these people, if they’re being consistent, would say that his chances in a hypothetical tournament doesn’t mean much). Extending that, if the candidates cycle produces a challenger who isn’t the pre-determined favorite or #1 seed (e.g, Leko against Kramnik, Gelfand against Anand), then these people also cry about the title match being somewhat illegitimate.

    I think the only result that would leave everybody satisfied (not necessarily thrilled) involves long, interesting candidates and world championship matches where the favorite ends up winning without any tiebreaks happening.

  4. I’m no GM (though I have a plus score against GMs ;-), but it seems to me that the tradition of matches predates the explosion of computer-generated opening theory.

    Doesn’t tournament play deemphasize the importance of openings (relative to matches), since it’s impossible to deeply study every system eight or ten opponent might play? Isn’t deemphasizing openings what we all say we want?

  5. I hadn’t thought about that, but I guess that’s true – tournaments force you to prepare more broadly (but maybe less deeply). A good portion of the top 10 is relatively limited in their repertoires.

    But while de-emphasizing super deep opening preparation is nice, the issue many people have with tournaments is that you no longer have to beat everybody you go up against. In a match system, you do have to win every match you play (ignoring time controls, or if somebody else beat the highest rated before you go to them, etc). That doesn’t hold in a tournament. In fact, you can lose mini-matches or head-to-head matchups against close competitors and still win the event.

  6. At least in a tournament you have to play everyone – if you can take first despite losing one or more mini-matches, you must have done something right. Especially when there are no “fish” – some would say Wijk aan Zee and the December London tournament are examples of doing it “wrong.”

    By contrast, in elimination matches, luck of the draw is so important. It doesn’t bother me that Gelfand qualified to play for the WC, but I’m sure the complaints would have been fewer if he had beaten Kramnik and Aronian along the way.

  7. You share interesting things here. I think that your page can go viral easily,
    but you must give it initial boost and i know how to
    do it, just search in google for – mundillo traffic increase go viral

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s