The Chess Mind

In the chess glut that is the Mainz festival, even the computers aren't left out. Shredder was the defending Chess960 champ, but ubiquitous upstart Rybka took its place, defeating it 2.5-.5. More here.

In a post several days ago, I presented the following sequence of moves:

DM-Pocket "Fritz", g/10:

1.e4 c6 2.d4 d5 3.e5 Bf5 4.Nf3 e6 5.Be2 Ne7 6.O-O c5 7.c4 Nd7 8.Nc3 dxc4 9.d5 exd5 (novelty) 10.Nxd5 Nxe5 11.Nxe5 Qxd5 12.Bxc4 Qxe5 13.Bb5+ Nc6 14.Re1 Be4 15.f3 Rd8 16.Bxc6+ bxc6 17.Qe2 (eventually 1/2-1/2)

and challenged readers to find improvements for White. During the game I felt sure after 9.d5 that I should have at least a small edge, and I felt happy about my moves when playing them. Despite that, I "awakened" around move 15 or 16 to realize that Black was slightly better, and it was time to earn the draw. (Which I did.)

Here are my findings; readers are encouraged to offer further improvements.

A week or two I was out for a walk and decided to renew my old "friendship" with my Pocket Fritz*. Here's how the opening went in this G/10 encounter:

1.e4 c6 2.d4 d5 3.e5 Bf5 4.Nf3 e6 5.Be2 Ne7 6.O-O c5 7.c4 Nd7 8.Nc3 dxc4 9.d5 exd5 (novelty) 10.Nxd5 Nxe5 11.Nxe5 Qxd5 12.Bxc4 Qxe5 13.Bb5+ Nc6 14.Re1 Be4 15.f3 Rd8 16.Bxc6+ bxc6 17.Qe2 etc.

Black is slightly better here, but despite being a pawn down and short of time, I was able to hold the draw without too much trouble. (Black's lousy queenside pawn structure was what I counted on, rightly.) Nevertheless, I was a bit disgusted and certainly confused: I felt after 10.Nxd5 and my clever 12.Bxc4 that I simply must be better! Yet it's not at all obvious where White could have improved, especially without the assistance of a computer.

See what you can find (without using Fritz, Rybka, Shredder, etc.); I'll offer my thoughts in a few days.

*In fact it's a version of the Shredder program, written by Shredder's author, but is called "Fritz" for (presumably) marketing purposes.

In fact, it probably should have been 5-3, as the computer lost the first game quickly due to a gigantic programming error. It wasn't a bad result by Benjamin, but it's scary to think that a strong GM going -2 with pawn odds is a fair result! Could it ever reach the point where humans could receive knight odds from the machine? I'm inclined to think it couldn't, but can we rule it out completely? Some day when I have a couple of hours on my hands, I'll see if Rybka or Fritz can give me knight odds and win - if it can't take me, then there's no way it should ever be able to take down a GM with that handicap, at least not for the foreseeable future.

Link.

HT: Prime#