I think this project is interesting, but if the engine used isn't specific to 960, wouldn't it omit issues related to castling? That would limit the accuracy of the analysis. I'm not familiar with this engine.
There's a decent article on 960 openings here: http://www.chess.com/groups/forumview/chess960-the-opening-makes-a-comeback
By the way, I've finally got my 960 rating above 1200! Made a great Christmas gift!
There is only a small database of 960 games so far, and I don't think that it includes any games of humans vs. computers. For such things to develop, it doesn't exactly grow like grass, 960 would need to gain considerably more popularity and have many more tournaments, ratings etc.
Also, the intention behind inventing Chess 960 (Fischer's modification of Benko's idea) was indeed to avoid opening theory, not that some day opening theory for 960 will be developed. (Which would need to be so large that only computers could memorize it.)
Chess960 doesn't avoid opening theory, it only avoids opening theory/history from Standard Chess.
Opening theory for Chess960 is developing as we speak. We are witnessing history!
While I would agree that computers will have no trouble memorizing 960 theory, humans - especially those who train for/play only 960 will be able to memorize theory as well.
Who exactly is developing "opening theory" for 960 ? Do you see Anand or Kramnik or Carlsen working on it ? Do you see players training for Chess 960 only ?
The history you are witnessing is largely happening in your imagination.
The OP seems to be misunderstanding engines. An engine can never tell you what opening is best. It's impossible. Unless you mean to tell me 1.Nf3 is best in standard chess, since Rybka says its the best move if it has no Opening Book loaded.
I think this project is interesting, but if the engine used isn't specific to 960, wouldn't it omit issues related to castling? That would limit the accuracy of the analysis. I'm not familiar with this engine.
There's a decent article on 960 openings here: http://www.chess.com/groups/forumview/chess960-the-opening-makes-a-comeback
By the way, I've finally got my 960 rating above 1200! Made a great Christmas gift!
Yes, indeed, it's not specifically chess960 engine, so I make the castling manually when I think that it's an appropriate move. Fortunately, there was only 1 reasonable castling in openings after 5 moves.
Thanx for the link, Merry Christmas, Igor
When you play Chess960, just shoot for a position you can play. Get your pieces developed and play the game. In general, central control with pawns are good such as 1. e4 or 1. d4. All the backrank pieces are randomised so tough luck.
Nope, it's not exactly like that. First of all, U should open the way to the queen and bishops. That's why I speak about dQB-, vQB-openings etc.
The OP seems to be misunderstanding engines. An engine can never tell you what opening is best. It's impossible. Unless you mean to tell me 1.Nf3 is best in standard chess, since Rybka says its the best move if it has no Opening Book loaded.
Any good engine will tell U that 1.e4, 1.d4 and 1.Nf3 are the best moves in the standard chess. So there are always several "best" moves. Usually, in all chess960 positions, they are dQB-, vQB- and knight openings and sometimes fianchettos.
I'm adding now estimations for Sicilian (=dQB/dQ) and Caro-Kann (=dQB/dQf) games.
Sorry if my earlier comment appeared trollish; I was just attempting to make the point that +0.02 doesn't mean that much at all in terms of an advantage. Considering that most engines give the initial position as +0.4 or so for white, yet white only scores about 54% even in grandmaster games, an advantage of +0.02 means nothing to most human players.
And, yes, as earlier posters pointed out, the point of 960 is to be able to play chess without memorizing openings.
[COMMENT DELETED] wrong thread
Nice...
Can you stop the rain.No.Its going to happen,this finding of openings for 960.Embrace it and enjoy the fact that you were there in the begining.(Dances with Pawns)
Hahaha
Kevin Costner dances with wolves but Richard Williams ... with pawns.
Thread starter is a neeeeerd
A little systematisation is usual and easy for me as a professional mathematician.
Systematisation of any kind including chess is a poison to my artistic mind.
Well, it might be in your particular case, if U feel like that, though it's far from being universally true. Many quite systematical scientists were talanted artists and vice versa.
26 positions so far; I'll add estimations for English (dQ-opening in dQB positions), Anglo-Scandinavian (dQ/vQB in dQB positions) and Pirc (dQB/vQBf in dQB positions) openings.
That Belgian guy had some stats about this, but I think that the number of high-level games played with each position is still too small to draw any definite conclusions. At any rate, this is dealt with by letting each player play one game with White and one with Black in the same setup.
I've always felt, but never bothered to try to prove, that there are quite a few chess960 starting positions that favor white so much that the game is almost unfair. Usually, it's positions with Bishops on one flank and the king on the other. The attacks are ready-made and brutally easy to carry out.
The defenses, if they exist, are much more difficult.
Has this been borne out by the attempt to create a chess960 opening theory?
And if I'm right, doesn't it weaken the case for chess960 if White can force a significant advantage (say, of a pawn or more) out of the opening?
So far, 27. BBQRKRNN (with 2 bishops on 1 side and 2 unprotected pawns on the other) seems to be the most advantageous for white. Though Sigma estimates it at only +1/3 of pawn.
That Belgian guy had some stats about this, but I think that the number of high-level games played with each position is still too small to draw any definite conclusions.
Do you mean this: Advantage in Chess960 Start Positions Revisited? It was done two years ago and could be redone using the current CCRL numbers. - Mark
Opening theory will never be as prominent in 960 OTB play as it is in conventional chess. There is simply too much to memorize. However, I wonder if there will eventually be a "My System" of 960 that provides guidelines for opening play in various types of piece configurations. How would such a book influence conventional chess theory even beyond openings?