Author: Robert Hyatt
Date: 10:34:21 10/18/00
Go up one level in this thread
On October 17, 2000 at 00:53:58, Ratko V Tomic wrote: >> The only thing I was pointing out is that if a program evaluates >> something as +3, when material is even, then it will evaluate that >> same position as even, when it is a piece down. > >That doesn't mean at all it would give a piece in that position. >Only a simple linear/additive form of the evaluation function would >have such side-effect. But a nonlinear combination of separate evaluation >components (as exemplified by Samuels' checkers program) need not >have such drawback. > You miss the point. We _know_ it will sacrifice material. In positions where it is not clear that it wins. IE the Rc6 move. So it _does_ happen, and if it happens when it is good, it will happen when it is bad. And people will learn how to exploit it. If you want to see this happen, take your favorite fast program, run it on ICC to get a good rating, then set the contempt to -.5 so it hates draws. Then sit back and watch. And it will drop like a rock. Because everyone finds out that it is best to try to repeat, as it will avoid. In fact, it will go to -.5 to avoid repeating. And -.5 is nearly losing since it is a positional score. As to the non-linear combination, I believe _everybody_ does that already. I certainly do. But if I can get +4 positionally, I can certainly give up a knight to get to +1, or two pawns to get to +2, either of which had _better_ be right as I am committed, being material down. Tiger is simply going to force some to fix their king safety. Some already have it fixed. Once everyone is up to snuff there, the attacks will begin to fail, and the material sacrifices will turn into boring losses. >While Gambit Tiger might not go as far to use as general mapping >as Samuels program did (since the chess parameter space is vastly >larger, in extent & number of dimensions, than that of checkers, and >would thus take too long to tune), there are great many intermediate >levels of non-linearity, some of which might work better and be practically >tunable in chess. For example, the sum of squares of properly chosen >term deviations would give much greater weight to the term which stands >out in a given position, producing an effect of program concentrating on >that aspect/feature of the position. That is in fact much more humanlike >way of analyzing a position. A good player knows what is the most relevant >in a position and spends his computational resources on the relevant >aspects, ignoring the irrelevant ones. > As I said, I already do this. I don't use sums of squares... I indirect the scores thru an array so I can scale them _exactly_ as I see fit. But with the same effect... And that 'favored score' can override the loss of material if I am not careful. Just check to see how many programs will trade a pair of minor pieces for a rook and pawn + some minimal positional edge. Etc. Some fake with the values of pawns (Rebel does this, as does others) to prevent trading a piece for 3 pawns which normally loses. It is a difficult thing to get right. >The Botvinnik's scheme takes this concept (of over-emphasis of relevant) >further by explicitly guiding multiple alpha-beta searches from the same >position, but using different objectives (and evaluation functions) each >time. In that case you would see great jumps in evaluation, which >merely reflects different perspectives one might take in a given position. > I won't bite there. I don't know that "Botvinnik's scheme" would even play chess... >It is natural that if you analyze a position looking for, say, a king >attack, then analyze the same position looking for a better pawn structure >for endgame, you will have widely varying estimates, e.g. if king attack >doesn't seem useful, you may get a very low score in that search (due >to weakening/sacrifices needed to get the attack off the ground). At the >same time the pawn structure oriented search may give an even position. > True. But if you wreck one term to favor another, you had better be _right_. >On the other hand, if the king attack search looks promissing, triggering >many criteria for a strong (but beyond horizon) attack, you may get a high >evaluation, well beyond the material balance on the board. For example >you could get +3 value, but that doesn't mean you would get 0 value if >you take a piece off the board (e.g. that piece is contributing to >the king attack or defending against the break on the other side etc). >Without a piece, there would likely not be any triggering of the high >score for the king attack, any you might simply get -3 score. Or that piece is sitting on a8, and you are willing to give it up for the attack rather than extricating it. That could be fatal. > >Taking one more step in abstracting the meaning of such high evaluation >swings (the first two levels being nonlinearity and different perspectives), >one can view the large score (well beyond the simple additive variation) as a >reflection of a much longer term estimate. This longer term estimate comes >from specialized evaluations tuned for a given type of position, which >may not be available or usable in all positions. But when it is judged >usable by the program, it will give a value much closer to what the >regular evaluation may see 20 or 30 plies later. E.g. in getting >ready for a king side attack, no checkmate or large material gain is seen >in the nodes examined, but the specialized long term estimators indicate >that in such position the expected gain may be a piece by the time attack >is completed, some 20-30 plies later. The large score variation is thus a >kind of far extrapolation reaching much closer to the final game values >(which could be, say, +99 for a win) than what regular evaluation offers. > >Obviously, the trick is to come up with such long term evaluations. But, >observing human play, we know these do exist in some implicit form in the >mind of strong players. So, since they do exist, it isn't beyond possible that >someone (Christophe T., Chris W., David K., Marty H., ...) may have teased out >some of that knowledge. to be sure...
This page took 0.01 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.