Author: Ratko V Tomic
Date: 00:25:28 11/10/99
Go up one level in this thread
Errata:
>
> 3.6 Castle state was not included in the count, but it was estimated
> that it would add less than 0.13 bits to the position code, in
> the worst case (check earlier discussion on king pair + castle
> status encoding in 12 bits). The ep status also wasn't encoded,
> but since there could be at most 4 ep capture targets (white
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> assumed to move, thus 4 targets are black pawns on the 5-th rank),
> ^^^^^^^^^
> the ep would add at most log2(5)=2.32 bits
^^^^^^^^^^^^
This should say: "at most 5 ep capture targets" (e.g. the 5th rank could
be: pPppPppP with 5 black potential targets & 3 white pawns capable of
capturing any of the 5). Thus the code addition is log2(6)=2.58 bits.
This addition of 0.26 bits still doesn't bring the full position code
above the 157 bits.
This page took 0.01 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.