Author: Russell Reagan
Date: 23:16:40 09/10/03
Just for fun, I created a bullet tournament today, and as I watched the games, I noticed a trend. I played many games between Ruffian and Yace, and I noticed in the evaluation graph of Arena that at some point in the game, Ruffian's evaluation would jump up (good for Ruffian), while Yace's evaluation stayed closer to an even score. Most of the time, after a few more moves had been played, Yace's evaluation would drop (bad for Yace). To me this says that Ruffian knows something that Yace doesn't, and would be a good way to test whether or not a change you made to your engine was good or not, and it should also help find weaknesses. For instance, you could find the "key positions" where Ruffian's evaluation (or whichever engine) jumps up while yours stays the same, and see what your engine doesn't understand about the position. Maybe you could see some trends. This gives me the impression that bullet games can be very valuable. Am I overestimating the importance of bullet games, or perhaps underestimating the value of games at longer time controls?
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.