Author: Dann Corbit
Date: 11:59:34 01/26/01
Go up one level in this thread
On January 26, 2001 at 11:21:46, Severi Salminen wrote: >Hi! > >Once upon a time there was a funny creature called Nellie the NPS... > > If I remove SEE from my engine I get a 2x speed up in terms of NPS. This of >course results in an increase in number of nodes to be searched. Overall it >seems that with SEE the strenght is not much higher than without. The funny >thing is also that removing all futility pruning code results another increase >in NPS (about 2x-3x). It is likely that removing it also decreases the strenght. > The point of this stupid message is that: DON'T MAKE ANY CONCLUSIONS OF >STRENGHT BASED ON NPS. > >I have spoken... You can go even faster if you remove checks to see that the moves are legal and just go with pseudo-legal moves. On the other hand, if you time your move GENERATOR, you can make some conclusions on that speed IF: 1. The opponent is using the exact same search algorithm 2. The opponent uses the exact same evaluation 3. The opponent uses the exact same extensions IOW, you are right, you cannot make _any_ assumptions based on NPS. MCHESS has demonstrated this for a decade or so, I think. On the other, other hand, a really, really fast move generator doesn't hurt a thing. ;-)
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.