a strange bkunder of rubichess 100M contempt50

[Uri Blass]

51...Nd5

I will test if I can reproduce the bug because rubi can see even after less than 0.1 second that 51...Nd5 is losing

[pgn][Event "Computer event"] [Site "Somewhere on Earth"] [Date "2024.02.28"] [Time "23:31:28"] [Round "2"] [Board "1"] [White "Stockfish 15 30M dup"] [Black "RubiChess_contempt50 100M 20240112 (avx2) "] [Result "1-0"] [Termination "mate"] [ECO "D37"] [Opening "QGD"] [TimeControl "nodes: 30000000;nodes: 100000000"] [PlyCount "149"] 1.d4 {+0.3/32 28474 30000538} Nf6 {+0.0/32 79123 100000246} 2.c4 {+0.3/32 27664 30001463} e6 {+0.0/35 79887 100000361} 3.Nf3 {+0.2/32 27630 30000231} d5 {+0.0/36 80519 90524658} 4.Nc3 {+0.3/32 28534 30000276} Be7 {+0.0/30 78576 92037414; D37: QGD, 4.Nf3} 5.Bf4 {+0.4/32 27360 30000567} O-O {+0.1/32 78114 98039345; D37: QGD, classical variation (5.Bf4)} 6.e3 {+0.5/33 26608 30001083} Nbd7 {+0.0/35 77241 100000522} 7.c5 {+0.3/31 26877 30000832} Nh5 {+0.0/40 78035 73617752} 8.Bd3 {+0.4/35 26144 30001144} c6 {+0.1/32 74867 100000453} 9.O-O {+0.5/34 25804 30000631} Nxf4 {+0.0/35 75821 100000406} 10.exf4 {+0.5/37 24910 30001511} b6 {+0.1/35 76612 54743198} 11.b4 {+0.4/38 25423 30001431} a5 {+0.1/32 73181 61385354} 12.a3 {+0.5/37 24660 30001174} Qc7 {+0.1/33 73655 90679703} 13.g3 {+0.4/32 25248 30000403} Ba6 {+0.1/32 70237 100000480} 14.Bxa6 {+0.6/34 24783 30000079} Rxa6 {+0.1/33 74307 97878452} 15.Qe2 {+0.4/36 24688 30000346} Ra7 {+0.0/35 73775 100000443} 16.f5 {+0.3/34 25273 30000207} e5 {+0.1/36 73069 98472835} 17.dxe5 {+0.5/34 25625 30000946} bxc5 {+0.0/39 75864 100000493} 18.b5 {+0.6/33 25967 30001327} d4 {+0.0/41 77472 100000373} 19.b6 {+0.3/34 25532 30000785} Nxb6 {+0.0/41 77695 40531549} 20.Ne4 {+0.2/35 25359 30000160} Nd7 {+0.0/39 79136 10389220} 21.Ned2 {+0.2/34 25820 30000585} Nb6 {+0.0/43 78811 100000337} 22.Rfe1 {+0.4/32 26200 30000173} Qd7 {+0.1/32 76032 96823661} 23.Rab1 {+0.4/34 26153 30000658} Nd5 {+0.0/37 78652 100000338} 24.Qd3 {+0.2/35 26190 30001097} Nc3 {+0.1/37 78099 98054770} 25.Rb2 {+0.6/36 25596 30001084} Qd5 {+0.0/42 78058 100000828} 26.Nc4 {+0.3/33 27041 30000382} Na4 {+0.0/43 77762 35706392} 27.Rb3 {+0.4/32 27303 30000908} Nc3 {+0.0/41 71832 98998685} 28.h4 {+0.3/34 25632 30001373} h6 {+0.0/43 75565 100000066} 29.Rb2 {+0.5/31 26113 30000175} Na4 {+0.0/47 78479 100000470} 30.Rb3 {+0.4/31 26745 30000747} Nc3 {+0.0/45 77208 100000713} 31.g4 {+0.2/29 26175 30000001} a4 {-0.0/38 81671 100000396} 32.Rb2 {+0.1/30 26878 30000331} Qd8 {+0.0/41 79307 100001028} 33.e6 {+0.0/34 27217 30000823} fxe6 {-0.0/41 80083 100000438} 34.Rxe6 {+0.0/35 26732 30000804} Bxh4 {+0.0/43 80006 100000331} 35.Nce5 {+0.2/36 26637 30000900} Bf6 {+0.0/41 78703 100000653} 36.Ng6 {+0.5/34 25801 30000345} Re8 {-0.0/32 71883 100000866} 37.g5 {+0.1/35 25651 30000222} Bxg5 {-0.1/31 71963 100000520} 38.Nfe5 {+0.0/37 25353 30001174} Ra8 {+0.0/37 74633 100000507} 39.Rb7 {+0.1/33 25153 30001415} Nd5 {+0.0/41 71029 100000847} 40.Qe4 {+0.0/34 24809 30001225} Qc8 {+0.0/42 70453 81064608} 41.Rxe8+ {+0.1/40 23783 30000728} Qxe8 {+0.0/42 72821 100000613} 42.Rf7 {+0.0/42 22766 30001483} Rd8 {+0.0/44 66794 100000614} 43.Rf8+ {+0.0/43 22382 30000097} Qxf8 {+0.0/45 60860 100000183} 44.Nxf8 {+0.0/47 21896 30001201} Rxf8 {+0.0/47 71371 100000487} 45.Nd7 {+0.0/47 23257 30001081} Rd8 {+0.0/47 68639 100000339} 46.Nxc5 {+0.0/48 21537 30000135} Nf4 {+0.0/46 71739 17027233} 47.Ne6 {+0.0/49 21399 30001067} d3 {+0.0/40 63212 100000472} 48.Nxd8 {+0.0/52 21106 30000734} d2 {+0.0/42 60945 100000573} 49.Qxa4 {+0.0/53 20471 30000001} Ne2+ {+0.0/41 55787 100000584} 50.Kg2 {+0.0/52 20045 30001192} Nc3 {+0.0/41 56755 100000195} 51.Qc4+ {+0.0/51 18847 30000691} Nd5 {+0.0/39 62260 100000494} 52.Qb3 {+6.9/30 22119 30001471} Bxd8 {+4.6/27 57564 67623100} 53.Kf1 {+7.1/30 21586 30001145} Kh7 {+4.7/27 54784 99050986} 54.Ke2 {+7.6/30 20338 30001121} Ba5 {+4.8/26 50096 96698539} 55.Qc4 {+8.6/27 19717 30000275} Bc3 {+5.0/30 50727 85677112} 56.Kd1 {+9.3/26 19005 30000007} Bf6 {+5.2/26 48047 97024071} 57.Kxd2 {+9.7/26 19200 30000262} Nc3 {+5.2/37 47501 100000725} 58.Ke1 {+10.9/24 17433 30001287} c5 {+5.2/37 37873 100000663} 59.Qxc5 {+11.2/24 17174 30000907} Na4 {+5.2/42 45892 81818254} 60.Qc6 {+13.5/22 15492 30000482} Nc3 {+5.5/30 40400 84192912} 61.a4 {+17.0/23 14710 30001258} Nxa4 {+5.7/34 41101 98012152} 62.Qxa4 {+152.7/34 13376 30000097} Kg8 {+5.8/32 39432 79972852} 63.Qe8+ {M+13/53 13231 30000194} Kh7 {M-12/31 35362 100000204} 64.Ke2 {M+12/55 12207 30000879} h5 {M-11/31 33024 100000220} 65.Qxh5+ {M+11/63 13068 30000611} Kg8 {M-10/31 29692 100000161} 66.Qe8+ {M+10/68 13967 30000249} Kh7 {M-14/32 27486 100000296} 67.Kf3 {M+9/72 13358 30000118} Bd4 {M-8/31 27929 100000148} 68.Kg4 {M+8/77 12967 30000267} Bf6 {M-7/32 27906 100000179} 69.Kh5 {M+7/92 12132 30001007} Bc3 {M-6/33 29506 100000215} 70.Qe4 {M+6/166 8326 30000336} Bf6 {M-5/36 31364 100000252} 71.f4 {M+5/245 1874 7825311} Bb2 {M-4/37 30760 100000170} 72.f6+ {M+4/245 179 1085933} Kh8 {M-3/42 24582 100000080} 73.Qe8+ {M+3/245 57 364811} Kh7 {M-2/45 25611 100000163} 74.Qg6+ {M+2/245 27 7257} Kg8 {M-1/255 24 7909} 75.Qxg7# {M+1/245 26 5148} 1-0 [/pgn]

[Uri Blass]

I can reproduce the bug
Rubichess already planned the right move one move earlier but it seems it read from hash a wrong move.

FEN: 3N2k1/6p1/2p4p/3n1Pb1/2Q5/P7/3p1PK1/8 w - - 5 52

RubiChess-20240112_x86-64-avx2:
38/1 00:00 0 1,840k 0.00 Nc3-d5
39/74 01:01 100,000k 1,619k 0.00 Nc3-d5