Author: Richard Pijl
Date: 01:08:34 01/24/05
Go up one level in this thread
On January 23, 2005 at 21:05:25, Daniel Mehrmannn wrote:
>Hi Richard,
>
>that is what we're talking about. Still very experimental and maybe not
>the fastest code but makes homer clear faster:
>Interessting idea and i thought thats new :)
I may have two additions:
- I try the hashmove before looping over all moves, as that move has the biggest
chance to fail high.
- Perhaps also look at the depth stored in the hashtable. If hashtable depth is
closer to the remaining searchdepth, the chance is larger that the move will
fail high too.
Richard.
>
>copy & paste:
> /* hash move */
> if ((tryHashMove || tryUpperBound)) {
> if (from == MainHashTable[slot].move.from &&
> to == MainHashTable[slot].move.to) {
> /* killer ? */
> if (MainHashTable[slot].score >= beta) {
> *oPtr += HASH_MOVE+1;
> tryHashMove = FALSE;
> tryUpperBound = FALSE;
> continue;
> }
> if (MainHashTable[slot].flag & hashfBest) {
> *oPtr += HASH_MOVE;
> tryHashMove = FALSE;
> continue;
> } else if (MainHashTable[slot].flag & hashfUpper) {
> *oPtr += UPPER_MOVE;
> tryUpperBound = FALSE;
> continue;
> }
> }
> }
>
>daniel
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.