Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Averaged results.

Author: Dann Corbit

Date: 17:39:38 04/02/04

Go up one level in this thread


On March 31, 2004 at 21:43:09, Dan Andersson wrote:

>Averaged results over 1000 runs per depth:
>value    depth   nodes      relative minimax    relative a-b
>2        2       3.463      0.86575             1.1543333333333
>10       4       10.399     0.6499375           1.4855714285714
>42       6       28.148     0.4398125           1.8765333333333
>170      8       74.578     0.2913203125        2.4057419354839
>682      10      185.363    0.1810185546875     2.9422698412698
>2730     12      459.625    0.11221313476563    3.619094488189
>10922    14      1106.041   0.067507385253906   4.3374156862745
>43690    16      2691.694   0.041071990966797   5.2675029354207
>174762   18      6280.954   0.023959938049316   6.1397399804497
>699050   20      14867.996  0.014179225921631   7.2633102100635
>2796202  22      34813.976  0.0083002986907959  8.5015814407814
>11184810 24      80965.644  0.0048259284496307  9.8847080942498
>44739242 26      185048.509 0.0027574376612902  11.295154062138

I am guessing this is what the table means:
value = the number of nodes in the full tree
depth = search depth
nodes = nodes examined by alphabeta (but then the depth=2 makes no sense)
relative minimax = ratio of nodes examined compared to minimax

relative a-b I do not understand.

Can you explain to greater detail what this represents?



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.