Friday, July 20, 2007

Checkers Solved

"Computers solve checkers" was the headline at ChessBase News yesterday. Turns out it's a draw. The article is worth a look, if only to read about the late Dr. Marion Tinsley's complete dominance of the game up until his death in 1994, and speculation about when, if ever, chess might be solved.

Funny, but knowing that checkers has been solved in no way affects my interest in playing the game....

6 comments:

Zweiblumen said...

I had the same feeling about interest in the game....knowing that chess or checkers is a draw with perfect play doesn't make it any less daunting and interesting for me to play perfectly.

I read the paper and I find their methods clever and intriguing as well, coming from a CS/AI background myself.

funkyfantom said...

Michael- this question is driving me nuts. I have the Mihail Marin "open Games" book, and I just read the Ruy Lopez Exchange chapter.

I have learned that strong chessplayers routinely play f6 as Black against the Exchange.

Why? It looks weak, by analogy to the Damiano defense, which is supposed to be positional suicide.

Why should the bishop-for-knight exchange change this?

Michael Goeller said...

I'm not the best person to answer this -- and this is not the best forum -- but I do have Marin's book and play the Open Games (though never the 3...a6 Lopez, since I prefer 3...Nf6, 3...g6 or 3...Nd4). My impression is that the point of ...f6 in the Exchange Lopez is to control the dark squares with pawns in a number of lines -- especially to control the ...e5 square. Following the exchange at d4 and the exchange of Queens, Black really has no worries about the weakening of the h5-e8 diagonal. Play might go 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Bxc6 dxc6 5.0-0 f6 6.d4 exd4 7.Nxd4 c5 8.Ne2 Qxd1 9.Rxd1 Bd7 10.Nbc3 Ne7 11.Be3 Ng6 12.Nd5 0-0-0 13.c4 Bc6 14.Rd2 Bd6 etc.

OR 6.d4 Bg4! (this looks like more fun) 7.dxe5 fxe5 8.Qxd8+ Rxd8 9.Be3 Bxf3! (giving back the two Bishops for structural compensation) 10.gxf3 Bd6 11.Nd2 Ne7 12.Rfd1 0-0 13.Kf1 Ng6 (controlling the key dark squares) -- in both cases with about equal chances.

Maybe you imagine that White can keep the Queens on with d3 rather than d4, but then he does not get to exercise his kingside majority and the ...f6 move helps to support a kingside initiative for Black after ...Bg4, ...Qmove, ...O-O-O, ...h5 and ...g5 etc.

Actually, I'll have to check out Marin's book some more on this, since this ...f6 move looks pretty good to me...

funkyfantom said...

Thanks for replying, Michael.
I will digest your full analysis once I understand the basics of the issue.

What I mean is, for example,
1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Bxc6 dxc6 5.Nc3 f6

Then why does Nxe5 not work? I assume it doesn't work since f6 is book.
But why is this not simply analogous to the Damiano defense, where f6 is just a bad move, and Nxe6 is the refutation?

I haven't used a computer on the position, since I want to understand it on a human level.

Michael Goeller said...

The other thing to note, which I did not mention in my previous response, is that White no longer has his light-squared Bishop, so exploiting any weakening of the light squares in Black's camp will be difficult for him to accomplish. One of the safer advantageous responses to the Damiano, for instance, is 1.e4 e5 2.Nf3 f6?! 3.Bc4! when Black is going to find castling difficult. And in the sac line with 3.Nxe5 fxe5? (better 3...Qe7) 4.Qh5+ Ke7 5.Qxe5+ Kf7 White needs the light-squared Bishop to continue his attack with 6.Bc4+ etc.

In the Ruy Lopez Exchange, White has given up the Bishop, so Black can safely settle his King on f7 in any sac lines. What's more, Black usually pins the Knight with ...Bg4 anyway -- a pin that cannot be easily broken without the light-squared Bishop.

You seem to assume that because ...f6 is bad in the Damiano, it is just a bad move generally. But as we are learning more and more, while some rules do apply, chess positions all have their specific requirements based on their unique features. I think if you looked over this with Fritz you'd feel better about it....

funkyfantom said...

I did a web search and found that a reader of Gary Lanes's ChessCafe column asked my exact question, which Gary answered, entitled "The Fixer") dated 1/3/2006.

From the lengthy analysis, the knight sac is unsound, but still there are many ways for Black to go wrong in a practical game, especially lower-rated players under faster time controls.

I think Marin should have included some discussion of this in his book. I don't know of any opening book that discusses it, and I would certainly want to be familiar with it playing the Black side of the Ruy.

By analogy, I have two books on the Petroff defense ( the Hooper and the Raetsky ) and both cover the Cochrane Gambit in only the most superficial way. This gambit is very dangerous for the average player.

Most opening books seem like they are only written for GMs.