Author: Uri Blass
Date: 11:34:11 01/03/05
Go up one level in this thread
On January 03, 2005 at 14:17:45, Jason Kent wrote: >I'm talking about my personal experience with Pharaon. I think engines can be >good at one time control and relatively bad at others. I have played quite a >few games with it in tournaments of long and short time controls. Based on my >testing, I think pharaon is quite a bit better at long time controls. This is a >few of my results with pharaon 3.1 recently. > >4min games: 11th (last place): Pharaon 3.1 148.5/400 > >60min games: 8th place so far: Pharaon 3.1 16.0/39 > >40min per 40 moves: 4th place so far: Pharaon 3.1 7.5/12 The number of games at long time control is not significant and it may be statistical error. 148.5/400 is more than 37% 16/39 is less than 42% No significant difference. Comparison between x minutes/game and x minutes/40 moves may be misleading because of time management problems. I think that the only good time controls to compare are x minutes/40 moves in both cases or fischer time control in both cases. In games with no increasement engines may lose on time or alternatively play too fast and lose the game for different reasons(for example a program that play immediately when it has less than a minute on the clock may have big problem in blitz but not at long time control when it almost never get that situation). 4 minutes/40 moves usually give engines enough time not to get into time problems so it is a good time control to compare with 40 minutes/40 moves because the meaning of 40 minutes/40 moves is searching nearly 10 times more nodes that 4 minutes/40 moves when the meaning of 60 minutes/game is often not searching 15 times more nodes than 4 minutes/game. Uri
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.