Author: Robert Hyatt
Date: 15:21:00 07/18/02
Go up one level in this thread
On July 18, 2002 at 13:36:58, Dann Corbit wrote:
>On July 18, 2002 at 13:02:42, James Swafford wrote:
>
>>On July 18, 2002 at 13:01:36, James Swafford wrote:
>>
>>>On July 17, 2002 at 17:29:49, Dann Corbit wrote:
>>>
>>>>On July 17, 2002 at 16:30:24, Patrik wrote:
>>>>
>>>>>Hello, Dr Hyatt.
>>>>>
>>>>> if (do_null && !tree->in_check[ply] && pieces && (pieces>5 || depth<421)) {
>>>>>
>>>>>Where does the depth 421 come from?
>>>>
>>>>It appears to be about 7 plies, since INCPLY is 60.
>>>
>>>So does null move break if you set INCPLY to something else?
>>>Seems ... || depth < INCPLY*7)) {
>>
>>depth < (INCPLY * 7 + 1))
>
>Yes, that would be a good idea.
>
>Since he is very good at using INCLPLY every where else, I suspect he was doing
>some sort of binary search for the best value and just left the constant in when
>he found it.
Correct. It took months to arrive at what was "optimal" for this. Whether it
is still optimal today is yet another question to be answered of course.
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.