If God plays chess, he's a gambiteer. He'll play the King's Gambit. Maybe the Danish.

Here's my rather easy draw against the 'beast' computer:

1.e4 e5 2.Nf3 Nf6 3.d4 Nxe4 4.Bd3 d5 5.Nxe5 Nd7 6.Nxd7 Bxd7 7.O-O Bd6 8.Qh5 O-O 9.Qxd5 Bc6 10.Qh5 g6 11.Qh3 Ng5 12.Bxg5 Qxg5 13.Nc3 Rae8 14.a4 Qf4 15.Nb5 Bxb5 16.axb5 Qxd4 17.Qf3 Qxb2 18.Qxb7 Bc5 19.Ra4 Qe5 20.g3 f5 21.Kg2 Kg7 22.Rd1 1/2-1/2

Computer analysis from my side started on move 14.

I seem to recall some old analysis (c. 1920's) that said that Black could draw most of the time with the Petroff, which is why it went out of business. Marshall complained that the Petroff was his favorite defense with Black, and it was now dead.

To win a tournament it is almost a necessity to win with Black, or at least to have chances. Maybe that logical necessity will keep this from happening in centaur chess.

Perhaps 'all' the computers do is raise the level of play to that a 1920's GM. No small achievement, that.

I don't know about that. The Petroff is Kramnik's drawing weapon, and he still loses with it sometimes.

Humans can still lose from it as a result of mistakes. When you ensure no blunders with the aid of a computer, white gets NOTHING.

Oh, I believe you, I was just replying to Ed. Between humans 1. e4 is still a legitimate winning try. And 1920s Grandmasters would surely get kicked around pretty badly with black in the Petroff against modern players.

I should say for the record that solving chess really doesn't matter for us humans that enjoy playing it. I'm more speaking from the 'Centaur' aspect, where it seems the Petroff makes 1.e4 a dead opening for white if black so chooses. I compare it to Tic-Tac-Toe in that regard. One things for sure: I've adopted the policy of always playing the Petroff against higher rated 1.e4 players in CC. I'm guaranteed a gain in ratings points without any fear of losing.

Just the other day, I got into a match with a 2400 in a Centaur game (I was only 2000 provisional). He went 1.e4 and after I went into the Petroff, he began complaining bitterly as he knew I was going to shave some of his rating points off. I replied "Of course I am. I'm smart that way ;-)"

Is it smart? It seems to me that there are two downsides. The first is that it brings closer the day when this form of chess dies out. Do you want that? The other is a form of opportunity cost: would you not shave more points by winning?

Consider it this way: at a 400 point rating differential the lower player has a scoring expectation of something like 5%. So the lower rated player risks little by losing. On the other hand the gain is much greater if one wins than if one draws. Then there is the consideration of certitude: if you play to win you risk losing, but by playing to draw you foreclose any chance of winning.

I have a related question: why not play the Cochrane Gambit, or simply transpose into a Four Knights?

It can't be true that the Petroff provides an easy draw in corr., because then almost noone would play 1.e4 with white and probably all games against 1.e4 would be Petroffs. I'm sure most games against 1.e4 are Sicilians.

I have a database with 166,000 IECG games (1995-2009). I searched for Petroff games from 2007-2009 and found 988. This contains ~400 white wins. If I restrict this to ratings of 2200+, still 16 white wins remain.

[Event "ST.2007.0.00008"]
[Site "LSS"]
[Date "2007.01.01"]
[Round "?"]
[White "Daurelle, Herve"]
[Black "Da Costa Junior, Luiz Roberto"]
[Result "1-0"]
[ECO "C42"]
[WhiteElo "2354"]
[BlackElo "2359"]
[PlyCount "64"]
[EventDate "2007.01.01"]
[EventType "corr"]

1. e4 e5 2. Nf3 Nf6 3. Nxe5 d6 4. Nf3 Nxe4 5. d4 d5 6. Bd3 Bd6 7. O-O O-O 8. c4
c6 9. cxd5 cxd5 10. Nc3 Nxc3 11. bxc3 Bg4 12. Rb1 b6 13. Rb5 Bc7 14. h3 a6 15.
Rxd5 Qxd5 16. hxg4 Re8 17. Ng5 g6 18. Bc2 Qd6 19. g3 Qf6 20. Bb3 Rf8 21. Ne4
Qe7 22. Qf3 Nd7 23. Bg5 Qe8 24. Bh6 Qe7 25. Re1 Rae8 26. Re3 Qd8 27. Ng5 Re7
28. Bxf8 Qxf8 29. Nxf7 Kg7 30. Ng5 Qxf3 31. Rxe7+ Kf6 32. Rf7+ Kxg5 1-0

Any Cochrane Gambit games? (3.Nxf7)