Author: Bas Hamstra
Date: 08:25:32 11/25/99
Go up one level in this thread
On November 24, 1999 at 00:36:41, Michael Neish wrote: > >Hi, > >I started this thread, "The Limits of Positional Knowledge" about a couple of >weeks ago. I'm happy to see that it's still ticking over, although by now it's >evolved pretty far from what my original question was! > >A couple of the early replies to my question were on the mark, but >the rest have strayed into areas far more noble and sophisticated than my >lowly inquiry. But since no one's really answered my (naive) question -- >maybe there is no answer? -- I'll restate it. > >Having recently written my first program, which can search only up to 6-ply, >that's if you want it to move sometime before the big crunch, I thought what >would happen if I stubbornly kept to a 6-ply limit and just tried to improve >the program's playing strength by giving it increasingly better positional >sense? You probably can make a strong program if you have a superior eval, even if it searches less deep than the average program. In my opinion and to my experience so far, eval is the factor with the biggest impact on playing strenght. By far. If you want to solve problem suites you can do quite well with a material only eval, but *not* so for playing games. But let me say this: if you want to stick to 6 ply you make it very difficult for yourself. You will be outsearched deeply all of the time and can only survive with not just a superior, but only a *godly* eval. You mentioned it yourself already: you reach a good position, but fail to exploit/convert the advantage. That's exactly the point, you will reach many positions that are won in a master's eye, yet a game is long and one tactical error and you lose. And that is going to happen often. The deep searchers are sometimes incredibly tough defenders. So I think 6 ply is pretty extreme. Still it would be an interesting experiment to try to see if it is possible to compensate the extreme outsearching with a extremely good eval. Against humans: yes. There are a few *instantly* moving engines playing at FICS, that have achieved ratings over 2000. But I suspect they *do* use their full power in the oppo's time. So that might bias somewhat. Against computers: very hard, I think. >I presume that given perfect positional knowledge no lookahead at all would be >required, since all tactical opportunities would be nipped at the bud (although >how it would exploit tactical blunders by the opponent is a separate question >itself). So my question was "how strong can a program become just by >concentrating on the positional evaluation?". I will allow the use of >extensions if a tactical possibility is discovered. Very strong if you don't want to stick to 6 ply, but focus on a very smart eval, and are willing to sacrifice some depth. Of course if you would load the thing with all types of extensions, it wouldn't be a typical 6-ply searcher anymore. Take Genius: it takes forever to complete 9 ply. Still it sees very deep tactics faster than most 13 ply thinking nullmovers. That's not a real 7/8 ply searcher, tactically it sees much deeper. >I fiddled with my program with this in mind, and although I found that it seems >to ponder over better moves quicker with a better evaluation function, I seem >to be hitting a wall as far as performance improvement is concerned. For >instance, if I play the program against itself, with White adopting the old >parameters, and Black the new, it seems Black plays better, but after developing an advantage it cannot convert. There are much eval tricks that can help: better king safety eval, rooks at seventh rank, freepawn eval, knowing about won endgames, two connected passers at 6 rank, precise eval of pawn at seventh rank. They should help a lot. Example: if your oppo doesn't see 2 connected passers at 6 rank are worth at least a piece, your program will sac a couple of pawns and win the game. >I appreciate any replies. Thanks. > >Mike Still a fascinating question. Can ANY 6 ply searcher survive in a computer pool? Regards, Bas Hamstra.
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.