HCha/Stockfish
1
0
Fork 0
mirror of https://github.com/HChaZZY/Stockfish.git synced 2025-12-20 09:06:45 +08:00
Code Issues Packages Projects Releases Wiki Activity

Extend intermediate LMR to root search

Browse Source 
Almost no change, but it is in sync with what we do in search
and in any case the ELO difference is very small (because the
events when the intermediate research triggers are very rare),
too small to be measured, we just verify we don't have any
unexpected regressions.

After 802 games at 1+0 full QUAD
Mod vs Orig +114 =581 -107 +3 ELO

Signed-off-by: Marco Costalba <mcostalba@gmail.com>
This commit is contained in:
Marco Costalba
2010-06-02 11:47:53 +01:00
parent 5f3c660d5d
commit 9337c6da46
1 changed files with 15 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

16
src/search.cpp
View File

@@ -892,17 +892,31 @@ namespace {
ss->reduction = reduction<PV>(depth, i - MultiPV + 2);
if (ss->reduction)
{
assert(newDepth-ss->reduction >= OnePly);
// Reduced depth non-pv search using alpha as upperbound
value = -search<NonPV>(pos, ss+1, -(alpha+1), -alpha, newDepth-ss->reduction);
doFullDepthSearch = (value > alpha);
}
// The move failed high, but if reduction is very big we could
// face a false positive, retry with a less aggressive reduction,
// if the move fails high again then go with full depth search.
if (doFullDepthSearch && ss->reduction > 2 * OnePly)
{
assert(newDepth - OnePly >= OnePly);
ss->reduction = OnePly;
value = -search<NonPV>(pos, ss+1, -(alpha+1), -alpha, newDepth-ss->reduction);
doFullDepthSearch = (value > alpha);
}
ss->reduction = Depth(0); // Restore original reduction
}
// Step 15. Full depth search
if (doFullDepthSearch)
{
// Full depth non-pv search using alpha as upperbound
ss->reduction = Depth(0);
value = -search<NonPV>(pos, ss+1, -(alpha+1), -alpha, newDepth);
// If we are above alpha then research at same depth but as PV