Analysing with computer

Discussion of chess software programming and technical issues.

Moderators: hgm, Harvey Williamson, bob

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
User avatar
Marek Soszynski
Posts: 509
Joined: Wed May 10, 2006 5:28 pm
Location: Birmingham, England

Re: Analysing with computer

Post by Marek Soszynski » Sun Jul 15, 2007 12:10 am

While we're talking about analysing with a computer, I'd like to appeal to all UCI authors to implement "Next Best" and "Analyze Threat," which I believe are part of the UCI protocol. In ChessBase during infinite analysis, pressing the y-key should cause the engine (if it supports it) to analyse the second-best move, and pressing the x-key should cause the engine to show what move is best if it was the other side to move. Not all engines support either or both functions, but e.g. the latest Rybka and Hiarcs do - that's because I pestered their authors about it!
Marek Soszynski

Michael Sherwin
Posts: 3017
Joined: Fri May 26, 2006 1:00 am
Location: WY, USA
Full name: Michael Sherwin

Re: Analysing with computer

Post by Michael Sherwin » Sun Jul 15, 2007 1:03 am

Marek Soszynski wrote:While we're talking about analysing with a computer, I'd like to appeal to all UCI authors to implement "Next Best" and "Analyze Threat," which I believe are part of the UCI protocol. In ChessBase during infinite analysis, pressing the y-key should cause the engine (if it supports it) to analyse the second-best move, and pressing the x-key should cause the engine to show what move is best if it was the other side to move. Not all engines support either or both functions, but e.g. the latest Rybka and Hiarcs do - that's because I pestered their authors about it!
Good job. Keep pestering! 8-)
I hate if statements. Pawns demand if statements. Therefore I hate pawns.

glorscheid

Re: Analysing with computer

Post by glorscheid » Sun Jul 15, 2007 8:24 am

Fritz can analyse an existing game or variation tree. By doing it backwards it may be able to keep some information in the transposition cache when moving to the previous move. But it needs the game/tree in advance.
Fritz has also an option to analyse a position. You have to specify how many moves it should find for the first, second, third and fourth halfmove. It just analyses the base position, then in a new step the first of the alternatives and so on. If my idea may be slow, this implementation on top of existing engines is very slow and I don't know in advance how many interesting alternatives exist on which level.

Post Reply