Author: Peter McKenzie
Date: 00:28:12 07/13/02
Go up one level in this thread
On July 12, 2002 at 13:31:31, Will Singleton wrote: >I didn't see any discussion of this game, the pgn was not available for awhile. >It appears to me that the game was drawn, but Warp picked the wrong move in the >following position (fxg5). If Bxa5 draws, then Junior was indeed lucky to win >the tournament. the game started about an hour late because the junior guys had a hard disk failure and were installing all their tablebases on the new machine. i was quite happy for this to happen. my random thoughts on the game, although i haven't analysed it yet and won't be for a while as i am now on holiday for a couple of weeks; the game itself was exciting. warp played the first dubious move with nfg5, but it wasn't too serious and the game was still about level. then junior played a series of suspect moves such as ra6, h5, h4. warp had a good game, and then i think junior unsoundly sacrificed a pawn. warp accepted it and defended well, winning another pawn in the process. so warp is 2 pawns up but junior has a nice blockade and there is no easy way for warp to make progress. probably warp should sacrifice the c-pawn, or even sacrifice the exchange. i think the position is objectively winning but my program doesn't see a good plan. eventually we get near the 50-move rule and also repetitions are an issue. warp has the idea of moving is bishop over to a3 to push the c-pawn. good idea, but not sound because of juniors wonderful f5. at this late stage of the game the thinking time is much reduced, so warp is not searching deep enough to see f5 coming. soon after f5 warp thinks it is a draw but a few moves later i think it thought it could avoid the draw and maintain a small plus which obviously was incorrect. junior was showing a draw score for a while too, so i suspect you are right and a draw was possible. in the end warp just wasn't up to the task, and junior played very well. i was satisfied with how my program handled the middlegame, but i must work on getting a bit more search depth to be fully competitive. of course, playing with 1.7ghz vs 2x2ghz doesn't make it easy either. i wonder how long programs need for f5, good test position i suspect. congratulations to the junior team, their program has very sharp teeth. you might guess the shift key doesn't work on this machine... > >[D]3q3k/5Q2/7r/p2N1Pp1/P1r1pPP1/4P2p/3B3P/6K1 w - - 0 67
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.