TalkChess.com

i have no 7-men syzygy's DTZ installed, only some WDL (among these is KBPPvKBP)
[d]2b5/8/3k3p/P7/8/3K1BP1/8/8 b - - 0 1
here SF lost with 1...Bh3 ?? while WDL is giving 0.00 for 1...Kc5 (only move)...SF was running on multiPV=1.
if i look "into its brain" by using multiPV say =5 its showing 1...Kc5 as 3rd best move with a score of 0.00 while all other moves have a -149.xx or -mate score (after ms's of search). i used latest SF-dev.

im pretty much confused that SF is choosing loosing moves while knowing that 1...Kc5 is a draw...might this be a problem with the DTZ i have not installed or do only i have this problem...some confirmation about this strange behavior ?

little update:
i think i found out that this issue happens only after some patch released on exactely the 09.06.2019...sry for not beeing able to delimit this further (internet connection problems)...sry for the (probably) misleading headline as well because it seems pretty much clear now to me that this is not an issue with DTZ !!?

So with multiPV=5 Stockfish is showing 5 moves that lose, lose, draw, lose, lose in that order? That's pretty weird...

It is correct that Kc5 draws and all other moves lose:
https://syzygy-tables.info/?fen=2b5/8/3 ... _b_-_-_0_1

Of course it should not play a move that is known to lose if there is a move that is known to draw. But none of the June 9 patches seem to relate to the relevant code.

See also https://github.com/official-stockfish/S ... ssues/2196

@ syzygy: yes thats correct: with multiPV=5 last SF-dev's are showing: lose, lose, draw, lose, lose ...insane but true (here on my pc at least)

@Louis: thanks for the related link...seems like a bug indeed from my pov

syzygy wrote: ↑Tue Jun 18, 2019 9:45 pm But none of the June 9 patches seem to relate to the relevant code.

I overlooked that this patch makes changes to "tbRank" in search.cpp and search.h. So that must be the problem. Should be happening in 6-piece positions too, I guess.

i see that the bug was found and already fixed (as far as i understood)...im glad that i could help a tiny bit in terms of verifying when to search...this should close the thread, thx for the replys

Spliffjiffer wrote: ↑Wed Jun 19, 2019 2:58 am i see that the bug was found and already fixed (as far as i understood)...im glad that i could help a tiny bit in terms of verifying when to search...this should close the thread, thx for the replys

Thanks for reporting!

some time has passed by when i tested a new SF so i cant tell when a specific patch has been merged where this issue would appear the 1st time here on my PC
the problem is that eg current SF-dev will stop thinking after only a few seconds in almost every position !...i have installed some few 6-men TB's and the full set of 5-men TB's and use an AVX2.exe
if syzygy-ProbeLimit is set to 6 or 7 i will get the bug discribed above but if i set it to 5 or less then everything works as it should !
with eg cfish 25032021 everything works just perfectly and witheg SugarAi 2.00 i get the same bug as with SF
last time it was a cleanup-patch that was the culprit...is it possible that this or a similar bug has been reintroduced again ?
anybody else who faces this problem as well ?

Do you really have the same problem again? Lose, lose, draw, lose, lose with MultiPV = 5 in the same position?