Author: Daniel Clausen
Date: 07:36:04 10/22/02
Go up one level in this thread
On October 22, 2002 at 10:15:32, Uri Blass wrote: [snip] >The 8 games with big hash for Fritz7 only gives an impression. >Instead of playing hundreds of games to check if the impression is correct it is >faster to get into conclusion by finding if Fritz7 search less nodes with higher >hash tables. I agree. But in this case the NPS (or depth reached, and similar) should be posted/considered, not the game result. It's very well possible that the NPS was only 50% (because a potential clearing of the HT would take more relativ time time with short timecontrols, as you mention) but the game result wouldn't reflect it, because it is only 8 games. Or in other words: To conclude from a 5-3 result, that the NPS didn't change significantly is simply useless. (although you're probably right in about 50% of all cases :) Sargon
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.