Author: Peter McKenzie
Date: 22:20:17 09/03/00
Go up one level in this thread
On September 03, 2000 at 22:36:41, William Bryant wrote: >I/m working on my time allocation to improve my program's play. > >Given a time increment of Game/Unit_Time (or Game/Unit_Time + Increment per >move), how do you allocate the Unit_Time among individual searches? > >At present I'm doing the following. > >1. In the opening, (when out of book), I'm allocating 1/64th of the time left >for a search (time_left>>6). So you only want to do a short think in the opening? I don't see the logic behind that. The only thing I do differently in the opening is use a bit extra for the 1st move out of book (because hash tables will be empty then). The amount of extra time I use is randomised, which is nice because it makes it harder to repeat the same game vs my program (useful on the chess servers). > >2. When past move 16, I change this to 1/32nd of the time left (time_left>>5). This is similar to my normal strategy too, although I think I divide by either 30 or 25. > >3. If I fail low at the root, I will allocate up to an additional 5x the initial >time to resolve the fail low. It may take less, but this is the limit. If I fail low at the root, I just multiply the original target time by 3. I will always use this extra time, even if the fail low is resolved earlier. I figure it can't hurt to use a little extra time because the position is probably quite tricky. Ok, so I couldn't be bothered putting the extra logic in :-) > >Any thoughts, comments, suggestions or discussion would be appreciated. I use less time for obvious recaptures. Such a recapture must pass some static criteria, and it should also be the prefered move at the root at all depths during the search. > >William >wbryant@ix.netcom.com
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.