Author: Jeremiah Penery
Date: 17:05:39 10/13/99
Go up one level in this thread
Sorry to leave this much text, but not much response. :/ On October 13, 1999 at 19:19:53, Ratko V Tomic wrote: >>Look at programs that can [be set to] calculate only 2 plies. Your 10 year old >>kids can beat it. If we go to 5 plies, it is quite a good opponent. At 10 >>piles, they play brilliantly. At 18+ plies they would be "the gifted of the >>gifted of the gifted" as far as tactics are concerned, and I think such a >>machine would actually have strategic power from time to time (depending on the >>board condition). > >Even the 18 plies in complex middle game (which may be 2-3 years off) still >hinge on the same old simple-minded evaluation of the terminal nodes. You think DB uses 'the same old simple-minded evaluation'? >Unless you >find a clear cut advantage at that depth, the judgment of such position by a >strong human player is superior to material, square count and such simple >citeria. This depth eliminates only more of tactical shots (if there are many >left after certain depth but well before table-base level). So what the strength >gain is depends grat deal on the type of position. A human GM playing such >computer in a manner he would play another human would likely lose, since they >would generate typical kind of positions where the deeper search may find >something. But there arte positions where no tactical shots exist, or at least >no such that are beyond GM's vision. While these positions may be more rare, >when one plays in a normal manner (as if against another human), even the >relative bumbling patzers (like most of us) have gotten the programs into such >positions from time to time. > >If there was a strong enough incentive & motivation (and maybe there will be >some day) for the chess professionals community to work out the openings and >strategies which would steer the game with great probability into such >positions, the current programs, DB and the rest, would drop down to 2000 level >or worse (depending on how complex such strategy may be for humans to carry >out). If I (a mere 2100 player over a decade ago) can get Fritz 5.32 etc to just >shuffle rooks on the back rank with no clue what to do next, But if you let Fritz search more ply (15, 20, ...ply), it will eventually find something better to do than shuffle rooks. Other programs may not have this problem at all. >with no systematic >plan or technique for getting there on my part, the real pros, with real >motivation, and few years to perfect it, could turn it into an almost certain >game flow. You don't need a table base to win most endgames which are winnable. >Yes there are odd/paradoxical positions here and there, but they could chip off >at best a tiny fraction of GM's points. You don't need the best move to win, >just a good enough move (chess programmers don't seem to know this, as yet). So do you want the chess programs to say "Ok, I have 3 choices for moves, and all look like they're 'good enough'. I think I'll choose the one that I think is the worst of the 3, because it's still 'good enough'." ? How else would you suggest to do it other than to pick the 'best' move (Or at least what the program thinks is best.) ? >>Look how hard they would make an expert work. 17% of the time, a new ply gives >>you a better move. With 3 plies deeper, that is nearly 50% of the time you get >>a better move [(1-.83)^3] on average. If you are not prepared for it, you will >>have to think really hard about why a particular choice was made. > >I have seen few papers where they follow change of move choice with depth. But I >wouldn't say that a different move is automatically a "better move" as you >assume. The "better" here could mean one more square attacked by the program as >seen 12 plies deep (which often means nothing since fre moves later it can all >change). Even an apparent small material gain doesn't mean it is a better move. >If you look sometimes how a program goes queenside pawn hunting, seeing a >brilliant 10 ply combination to a win your b2 pawn, while you maybe see an >enemy Queen getting into his qeenside, likely to win a pawn or two, but counts >on his king side pawns to break open the program's castle and is readying his >pieces to take part in an attack 10-15 moves ahead. To draw programs queen into >such adventure, whenever chance occurs, I weaken my b2 or a2 pawns, knowing that >as soon as it sees a two-three move queen manouver to grab one or more of them >it won't be able to resist the temptation. A different move almost always is better if it's found at greater depth. I'll use your example: Say at depth 10, the program sees that it can do Qxb2, winning a pawn. If it searches to depth 11, say, it may see that it's getting attacked, since it doesn't have its queen to defend. So it picks a different move, that doesn't win the b2 pawn, but does prevent your attack. Is this not a better move, found because of greater depth? I think you're underestimating the evaluation functions of (most) chess programs. It's a lot more than simple material, piece-square tables, mobility, and attacked squares. Jeremiah
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.