Author: David Eppstein
Date: 13:31:56 12/16/98
Go up one level in this thread
On December 16, 1998 at 08:45:09, James T. Walker wrote: > It seem as though the only "Real sacrifices" are the ones that are unsound !! I disagree. If you are talking about the long term game-theoretic value, all moves are either correct or blunders, there is no middle ground. But in real life, we can't see that far, and we (or our computers) have to make decisions only based on what we (or they) can see, based on imperfect heuristics such as material evaluation. If a move gives up material, and we see that it later forces checkmate or the return of at least as much material, it's a combination not a sacrifice. The combination may be unsound (e.g. we missed a Zwischenzug) but it's still a combination. But, if a move gives up material, and all we see in exchange for it is positional compensation, it's a true sacrifice. The sacrifice may or may not be sound (may or may not have the appropriate game-theoretic value) but it's still a sacrifice. With computers, it's especially easy to distinguish between a sacrifice and a combination: look at the PV and see how the material balance at the end of the line compares with the current position. If the material balance is worse, but the overall eval is as good or better, then it's a true sac. But you need the PV of the machine actually playing the game, not someone else's post-game analysis. It's like the difference between a theorem and a conjecture in mathematics. Both are statements that may or may not be true (a proof can be wrong, of course) but one is something we think we've seen through to the end and the other isn't.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.