Dev Builds » 20260504-0631

You are viewing an old NCM Stockfish dev build test. You may find the most recent dev build tests using Stockfish 15 as the baseline here.

Use this dev build

NCM plays each Stockfish dev build 20,000 times against Stockfish 14. This yields an approximate Elo difference and establishes confidence in the strength of the dev builds.

Summary

Host Duration Avg Base NPS Games WLD Standard Elo Ptnml(0-2) Gamepair Elo

Test Detail

ID Host Base NPS Games WLD Standard Elo Ptnml(0-2) Gamepair Elo CLI PGN

Commit

Commit ID 3e3e976fdf54adf6fce636ff21f7c0f4ce66a70a
Author Robert Nurnberg @ elitebook
Date 2026-05-04 06:31:06 UTC
[bugfix] Respect bound flags for searched moves in multiPV analysis Master in `pv()` incorrectly assumes that any root move with `i != pvIdx` has an exact score. This can lead to the suppression of bound flags for (a) root moves that have been partially searched in an iteration but do not end up to be best move, or (b) when an aborted search for `pvIdx > 0` leads to a new best move. This latter case is the cause for issue #6783. Fixes #6783. The bug in its incarnation (a) was independently discovered by @ciekce a couple of weeks ago, and we have jointly discussed a patch that fixes it. He is therefore a co-author of the bugfix part of this patch. While at it, we also fix three minor oversights in the PV roll-back for aborted searches with a loss score: * We only roll back the PV if the aborted search would report an exact score. * Master inadvertently may have reported a bound for the rolled-back score+PV. * We ensure a final UCI info line with the correct search depth and searched nodes statistics etc. closes https://github.com/official-stockfish/Stockfish/pull/6784 No functional change Co-Authored-By: Ciekce <44617491+Ciekce@users.noreply.github.com>
Copyright 2011–2026 Next Chess Move LLC