Author: Danniel Corbit
Date: 11:21:09 05/12/98
Go up one level in this thread
On May 12, 1998 at 11:55:33, Bruce Moreland wrote: [snip] >Pick a position at random from a GM game, and figure out how many times >it has come up elsewhere. The answer will almost always be zero. >You'll get some endgame positions and a lot of opening positions, but >almost no middlegame positions. And even if you get one, you get X >minutes to search it, where X is not a lot. But when I use crafty to insert all GM moves into its database [including the whole game] the number of unique positions is greatly reduced. But I do suspect that it is openings and endgames, as you suggest. >Because if you start talking about *every* position from *every* high >quality game, you're talking about a lot more than the 500,000 positions >that you mentioned before, you're talking about maybe a hundred million. > I think it was 6000 computers for a week before, now it's 1.2 million >computers for a week. Exactly why you can only prepare for a single GM, not all of them. >And the result is almost meaningless. I don't think it would improve >strength measurably to occupy two billion dollars worth of hardware for >a week. At $10 an hour and one minute per install, it would cost you >$200,000 just to install the software on these computers. Yet things like this do get accomplished. Look at the crack of the 129 digit RSA number or the Great Internet Mersenne Prime Search. >If you want to talk about "siimilar" positions, then you have the same >idea that Graham Laight presented in r.g.c.c. a couple of times. The >mere act of identifying a similar position is very hard. Once you have >something that is capable of understanding that a position is similar to >another position, and verifying that a move made in the other position >is still a good move here, you might as well just let it play chess and >get rid of the pattern matching part, because it couldn't help but be >very good just by itself. I have searched through database tables of 24 gigabytes and found the answers in a fraction of a second. The answer is a unique, clustered index [which I would have on the FEN strings]. Even for a miss, a LIKE search would still be very fast if there is a large portion of the key that is the most significant characters. In any case, we are talking about seconds, not minutes. Each analyzed position that is superior to the current position can become a goal, like capturing a piece or improving position. If it is not reachable, we simply abandon that goal and operate normally. While the number of feasible chess positions is obviously astronomical, the number of excellent formations is considerably smaller, I should think. And if we store a billion of them, it should be possible to reach some of them without too much effort, at times.
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.