Author: Jeremiah Penery
Date: 13:40:17 10/16/99
Go up one level in this thread
On October 16, 1999 at 10:56:08, Ratko V Tomic wrote: >> Of course, the effective branching factor will vary with the position. >> For some positions it will simply be fairly low. Others, higher. Or >> it can be low for many ply, and then suddenly be high at certain depths >> in the search. > > >Yes, I know it can vary. But B=3, with typical Crafty's nps (200 kN/s on Celeron >400Mhz) which I get in Fritz UI (which came with Crafty 16.6) didn't fit at all >the depths I saw it reach in the 15-30 seconds when I was trying it out. So I >took a position which had exactly 38 branching factor and measured the exact B >for depths 10, 11 and 12 plies, and indeed B was 4.5-4.7. Yes it varied with >depth, but just a little bit and nowhere near to bring the B down to 3. Now this >is in Fritz UI and what gets displayed as Nodes may not be the same what Crafty >diagnostics show as nodes, i.e. we're dealing with different definitions of >"total nodes." I simply go by the amount of time taken. It's easier. :) >The position I tested is the from the game below, after black's >6th move (6 ...Nc6) Is this correct? In the game below, this is not the move shown. >and while white is thinking for 7th move (7. Bb5). What do >you get with your Crafty for B here (and what are the exact "node totals" from >which B was computed) at depths 10, 11 and 12? >1. e4 c5 2. Nf3 e6 3.d4 cxd4 4. Nxd4 a6 5. c4 Nf6 6. Nc3 Bb4 7. Bd3 Nc6 8. Nxc6
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.