Houdini's trick

[lech]

The next one:
Chiron 1.5 64-bit 4CPU - Houdini 3 64-bit 4CPU [E00]
The Champions 2012 4CPU ChessGUI (7.7), 21.11.2012
[Graham Banks] 29 minutes per 40 moves
[d]1r2n3/1Pq2k2/3p2p1/2pP2p1/Q3N2p/4R2P/7K/8 b - - 0 50
50...Kg8 (Kg8) 0.47/22 0 51.Qxe8+ (Rb3) 0.00/28 55 51...Rxe8 (Rxe8) 0.25/20 16 52.Nf6+ (Nf6) 0.00/28 48 52...Kf7 (Kf7) 0.23/19 0 53.Nxe8 (Nxe8) 0.00/28 13 53...Qb6 (Qb6) 0.03/26 48 54.Re6 (Re6) 0.00/28

Houdini takes 0 second for move 50 and is blind.

[Master Om]

I don't find any bug!!

New game
[d]8/7R/P3k3/1B3p2/1p1b3P/1P2p1PK/5r2/8 w - - 0 1

Analysis by Houdini 3 Pro x64:

55.Rd7 Bc5 56.Rd8 Rb2 57.Bc4+ Ke7 58.Rd5 e2 59.Re5+ Kd6 60.Rxe2 Rxe2 61.Bxe2 Ke7 62.Kg2 Ba7 63.Kf3 Kf6 64.g4 fxg4+ 65.Kxg4 Bb8 66.h5 Ba7 67.Bd3 Bd4 68.h6 Be3 69.Kf3 Ba7 70.h7 Kg7 71.Kf4 Bb8+ 72.Kf3 Ba7 73.Kf4
+- (2.34) Depth: 25/61 00:01:11 853mN
(Prakash, Bhubaneswar 22.11.2012)

[Master Om]

What blind ?

New game
[d]1r2n3/1Pq2k2/3p2p1/2pP2p1/Q3N2p/4R2P/7K/8 b - - 0 1

Analysis by Houdini 3 Pro x64:

50...Kg8 51.Qxe8+ Rxe8 52.Nf6+ Kf7 53.Nxe8 Qb6 54.Re6 Qb2+ 55.Kg1 g4 56.Nxd6+ Kg7 57.hxg4 h3 58.Re7+ Kf6 59.Rf7+ Kg5 60.Ne4+ Kh4 61.Rh7+ Kxg4 62.Nf2+ Kg5 63.Nxh3+ Kf6 64.d6 Qb1+ 65.Kf2 Qf5+ 66.Ke2 Qe5+ 67.Kd3 Qxd6+ 68.Kc4 Qe5 69.Kb5 Qd5 70.Kb6 Qd8+ 71.Kxc5 g5 72.Kc6 Qe8+ 73.Rd7 g4 74.Nf4 Kf5 75.Nd3 Qe6+ 76.Rd6 Qe8+ 77.Kc7 Qf7+ 78.Kb6 Qf8 79.Kc7 Qf7+ 80.Kb6
= (0.00) Depth: 26/80 00:01:07 803mN
(Prakash, Bhubaneswar 22.11.2012)

[lech]

Om, You don't feel some difference play-analysis.

[Master Om]

So is Houdini 3

New game
[d]Q7/8/4k3/5p2/Bp5P/1Pb3PK/r3p3/8 w - - 0 1

Analysis by Houdini 3 Pro x64:

1.Qc6+ Ke5 2.Qc7+ Kf6 3.Qd8+ Ke5 4.Qb8+ Kd5 5.Qb5+ Ke6 6.Qe8+ Kd5 7.Qg8+ Ke5 8.Qg7+ Kd5 9.Qg8+ Ke5
= (0.00) Depth: 26/61 00:00:12 166mN
(Prakash, Bhubaneswar 22.11.2012)

[Master Om]

Ok Marek I will try that too

[lech]

Ok Marek I will try that too

It seems to be impossible to set the same conditions like in a game.
BTW. IMO, the tricks are unworthy of the ELO leader.

[Master Om]

Ok Marek I will try that too

It seems to be impossible to set the same conditions like in a game.
BTW. IMO, the tricks are unworthy of the ELO leader.

True. I am still not getting the buggy move.

[lech]

I don't think less pruning is the right answer.
Houdini should recognize that nothing can stop the black pawn, and that it is effectively worth a Queen. Houdini is rather conservative in assigning bonus points for advanced passers, ---

This example: http://talkchess.com/forum/viewtopic.ph ... highlight=
shows that you you lie.
It is a side effect of your devilish tricks.

Houdini 3 - Deep Rybka 4.1 (The Champions 2012)
[d]8/1k1r4/p2P4/1b6/2nQN1p1/6K1/8/8 w - - 0 60
White could play 60.Nc5+to win the quality (it leads to win - tablebases), but resigned for advanced passed pawn d6.

[carldaman]

You're saying White resigned ?


Regards,
CL