01 February 2014

It's Not About Short Draws, Garry

When the 13th World Champion talks, people listen. When he talks about chess960, I listen very carefully. Here are some of the things he's said, or that have been said about him, that I've noted in past posts on this blog.
  • Kasparov's Chess960 Proposal • GM Kasparov: 'I’m in favor of at least investigating doing one position per year from chess960.'
  • Why Not Announce Positions Beforehand? • GM Kasparov: 'I have always liked the idea of choosing a few decent positions. And, I don't think you need more than 15 to 20, out of the 960 possible random chess positions, many of which violate our sense for normal chess geometry.'
  • Not Everyone Likes Chess960 • GM Kasparov: 'As for Fischer Random or similar ideas, I'm very much in favor. Let's be very specific. Fischer Random in its purity is not such a great idea. It creates a mess at the chess board from the very beginning. Out of 960 positions, 95% are quite bad.'
  • Kasparov *Did* Play Chess960 • GM Shipov: 'I remember we played six games of Fischerandom chess, and there was no battle there at all! In completely unfamiliar positions, Kasparov's advantage over me was far greater than in normal chess.'

Kasparov recently visited the 2014 Tata tournament at Wijk aan Zee, where he conducted a press conference sitting next to Dirk Jan ten Geuzendam of New in Chess. The full session can be viewed on Youtube: Tata Steel Chess 2014 - Press conference - Garry Kasparov. At about 38:35 into the clip, ten Geuzendam asks him a question.

Q: When you became World Champion you created a revolution by the way you approached the game: big emphasis on preparation, you were incredibly well prepared. Now we have Magnus Carlsen and things have become slightly different. Where do you think chess is going at the moment?

A: Magnus' victory was very symbolic, and significantly, because four or five years ago, there was a big talk about chess being a drawish game. Sofia rules are very much a reflection of this fear because nobody believed that you can actually force players to go back to this fighting spirit. Magnus almost singlehandedly changed this perception.

I believe that the short draws are very much a psychological problem. The players feel that they are not obliged to play fighting chess. They have no responsibilities for the public and for the world of chess. The way Magnus played, and he's still playing, pushed them in this direction.

No one is talking any more about tournaments being too dull, too many draws. Look at this tournament [Tata] or look at any other tournament. The leader obviously influences the whole pack. The way he plays, it just offers us hope that -- without drastic reforms, this reshuffling of pieces in the opening position -- we still can go on for quite a few years with the classical chess and people can still enjoy it. I'm really grateful for Magnus for being such an inspiration for many players.

In this context, the mention of chess960, aka shuffling the pieces in the opening position, tells me that Kasparov believes it is intended to tackle the problem of short draws. While that might be a (welcome) side-effect, players are still free to agree a draw on the third move if they are so inclined (*). I really shouldn't be surprised by this. It's consistent with an observation I made in 'Kasparov's Chess960 Proposal'.

The purpose of [Kasparov's] suggestion seems to be letting professional players continue with the working methods they have used for decades -- preparing opening systems in advance, committing them to memory, and playing their ideas in important games against unsuspecting opponents.

The Sofia rules address the problem of short draws. Chess960 addresses the problem that the opening phase in chess is more and more about memorizing computer generated variations. It gives players a choice between studying opening theory (chess) or working things out themselves (chess960). The idea that chess960 is all about short draws is so prevalent that it deserves to be added to my list of Top 10 Myths About Chess960.

While I'm on the subject of GM Carlsen, I'll mention a previous post, Carlsen's First Chess960 Move? Perhaps he'll come back to it again some day.


(*) 'So inclined': Next Short Draw: 2750.


HarryO said...

Exactly Mark, it is not about short draws! When are people going to wake up to the domination of opening theory on Chess?

Nakamura's own distinctive genius is showing Carlsen how to play Chess, but each time Nakamura shows him, Carlsen adds that to his pre-existing opening knowledge and Nakamura looses another weapon. How many weapons are at his disposal that Carlsen won't be aware of? Well almost none because the players are starting from the same chess position every game and Carlsen's memory is massive. He simply absorbs it.

This would not happen in Chess960. The flame of natural risk taking players would not be extinguished by opening theory.

There is also a myth about Carlsen that he is showing that you don't need a massive opening repertoire to win. Carlsen's opening knowledge is huge don't be fooled by that.

GeneM said...

Mark is right that – chess960 FRC has nothing to do with the short draws problem. Chess960 has nothing to do with the high draw rate problem either (in long time control games between elite grandmasters). Perhaps Kasparov is correct that the short draw problem has decreased recently, but the draw rate problem has not been reduced.

Mark is mistaken about FIDE’s source of strength: I believe it comes primarily from FIDE’s rating system.

Mark is unjustified in asserting that the goal of chess960 is to remove the effect of at-home opening phrase preparation. Yes that was Fischer’s goal, but Fischer speaks only for himself. Excessive at home preparation is a problem, but zero at home prep is too far in the other extreme.
To me the goal of chess960 is to see the opening principles at work in any sensible start setup other than the only one start setup that has been studied and tested (in any depth). A close second goal is to witness the emergence of completely new families of opening systems, in the weeks and months when they first appear. In contrast, I was not around when Petroff’s Defense first occurred.

Yet my main comment today is to urge Mark to switch to a glass-half-full perception from his glass-half-empty perception. Kasparov has said repeatedly that he has some sympathy for the idea of one new start setup (from chess960) per year-ish: Mark should talk more about what a major step forward that would be.
If American chess sponsor Rex Sinquefield told Kasparov that Sinquefield would $ sponsor a small but elite chess960 tournament, with one FIDE-blessed start setup (and with long time controls), I presume that would motivate Kasparov to have a FIDE committee formally annoint one setup (Sinquefield could probably nominate the setup). Of course, Sinquefield should do with his own money only what he wants to do. But if Sinquefield wants to have an long *enduring* effect on the course of chess history, this chess960-related opportunity seems like an exceptional opportunity.

GeneM , 2014-Feb-09

HarryO said...

Something doesn't make sense. If you want chess to be a theory game rather than a mind sport (a practical contest over the board), you will generate a lot more theory if you accumulate games from randomly distributed starts, than if you play one position every year for 959 years....

Think about it, one position a year just means that you are deeply understanding the move sequence of one position, but there are 958 others that also have a story to tell. I'm talking about theory as in general principles, not theory as in memorising combinations!

Not only will it generate "theory" more slowly, one position a year will favour players who love to pre-arrange their games, rather than players that love over the board contests.

Your argument just does not stack up I'm sorry I don't get it. I also disagree that Fischer was thinking only of himself when he invented 960.