Deep Shredder 12 didn't see this coming

[zullil]

It was THE Chiron's wonderful move:
[d]r1b2rk1/pp3ppp/1bn1p3/3q4/3P4/3BBN1P/PP3PP1/R2Q1RK1 w - -
Rc1!!
The next Qxa2 it is only a wrong effect of some "easy move".
It means that if a move has a much better evoluation (by a few iterations) than others, is played quickly.
It happened.
In play mode, other engines can play Qxa2 too.

While I was sleeping, Critter switched from Rc1 to Re1:

Code: Select all

25/66  17:06   10679M  10398K  +0.28 Rc1 Bd7 Bc4 Qf5 d5 exd5 Qxd5 Qxd5 Bxd5 Bxe3 fxe3 Rad8 Rfd1 g6 Nd4 Kg7 Bf3 Rc8 Nxc6 Bxc6 Bxc6 Rxc6 Rxc6 bxc6 Kf2 Rb8 b3 Rb7 e4 Re7 Kf3 c5 Rd6 a5 Kf4
25/66+ 27:24   17179M  10446K  +0.31 Re1 Bd7 Ng5 h6 Bh7+ Kh8 Be4 Qd6 Nh7 Bc7 Nxf8 Qh2+ Kf1 Rxf8 f3 Qh1+ Bg1 Rd8 Rc1 Bb6 Rc5 Bxc5 dxc5 Be8 Qc1 Kg8
25/66  31:08   19474M  10425K  +0.24 Re1 Bd7 Ng5 f5 Rc1 Qd6 d5 exd5 Bc4 Ne7 Bxb6 axb6 Rxe7 Qxe7 Qxd5+ Kh8 Nf7+ Rxf7 Qxf7 Qxf7 Bxf7 Bc6 Rd1 g6 f3 Kg7 Bb3 Re8 Kf2 Kf6 f4 Kg7 Rd6 Re4 g3 Re8 Bc4
25/66  32:32   20375M  10438K  +0.28 Rc1 Bd7 Bc4 Qf5 d5 exd5 Qxd5 Qxd5 Bxd5 Bxe3 fxe3 Rae8 e4 g6 a3 Kg7 b4 a6 Rfd1 Re7 g4 f5 gxf5 gxf5 e5 Rfe8 Rc2 Nxe5 Nxe5 Rxe5 Rg2+ Kf6 Bxb7
26/66  45:22   28823M  10585K  +0.24 Rc1 Bd7 Bc4 Qf5 Qd2 h6 b4 Ne7 Bd3 Qf6 b5 Nd5 Ne5 Rfd8 Qb2 Bc7 Rfe1 Rac8 Bd2 Qh4 Nf3 Qe7 Be4 Qf6 Qb3 a6 bxa6
27/66  76:04   48870M  10706K  +0.27 Rc1 Bd7 Bc4 Qf5 Re1 Rad8 Bd3 Qd5 a3 h6 Bc4 Qh5 d5 Bxe3 Rxe3 exd5 Qxd5 Qxd5 Bxd5 Rfe8 Rb3 Rb8 Rd1 Be6 Bxe6 Rxe6 Rd7 b6 Rc3 Rd8 Rc7 Rd1+ Kh2 Rdd6 Rc1
28/66  108:46   70604M  10818K  +0.26 Rc1 Bd7 Bc4 Qf5 Re1 Rad8 Bd3 Qd5 a3 h6 Qc2 Qd6 Qe2 Rc8 Qd2 Qd5 Rc3 Rfd8 Bxh6 Nxd4 Ne5 f6 Nxd7 Qxd7 Rxc8 Rxc8 Bf4 e5 Be3 Qd5 Bg6 Kf8 b4 a5 Bf5 Rd8 Bd3 Kg8
28/70+ 236:42  154399M  10871K  +0.35 Re1 Bd7 Ng5 f5 Rc1 Qd6 d5 exd5 Bc4 Ne7 Bxb6 axb6 Rxe7 Qxe7 Qxd5+ Kh8 Nf7+ Rxf7 Qxf7 Qxf7 Bxf7 Bc6 Rd1 g6 f3 Kg7 Bb3 Re8 Kf2 Kf6 f4 Kg7 Rd6 Re4 g3 Re8 Bc4
28/70  272:58  177837M  10857K  +0.37 Re1 Bd7 Ng5 f5 Qe2 Qd6 Rad1 Nb4 Bc4 h6 Nf3 Rac8 Ne5 Nd5 Qd3 Ba4 b3 Be8 Bd2 Bh5 Rc1 Nc7 Bc3 Nd5 Bxd5 Qxd5 Bb4 Rfd8 Rxc8 Rxc8 g4 a5
29/70  307:33  200974M  10890K  +0.28 Re1 Bd7 Ng5 f5 Qe2 Qd6 Rad1 Nb4 Bc4 h6 Nf3 Rf6 Bd2 Nc2 Bf4 Qxf4 Qxc2 Rd8 d5 exd5 Bxd5+ Kh7 Bxb7 Rg6

[mclane]

deep shredder12 is a very old program and so far there is no shredder13.

[styx]

stockfish 2.2.2 JA SSE42 got it after 12 meganodes or 3 seconds on my rather slow system. faster than critter 1.6a and houdini 2.0. pretty good!

[gleperlier]

I think I posted here already some great moves by Chiron and the sack on opponent king structure seems to be something Chiron is well trained in.

See bellow :

http://www.talkchess.com/forum/viewtopi ... ght=chiron

http://www.talkchess.com/forum/viewtopi ... ght=chiron

Cheers,

Gab

[lech]

While I was sleeping, Critter switched from Rc1 to Re1:

Rc1!! is a wonderful ambush move in play mode. Probably Re1 can be better in analysis.
As Graham noticed, the score after Bxh7+ was very high. It is very important and means that Chiron is a very dangerous opponent.
Let it be a final wonderful play: Chiron v Houdini

[Dan Astrachan]

[d]r1b2rk1/pp3ppp/1bn1p3/8/3P4/3BBN1P/qP3PP1/2RQ1RK1 w

Chiron 1.1a 64-bit (8 cores) v Deep Shredder 12 (8 cores) from the "Revving Up The V8s" tournament.

Chiron plays Bxh7+ and goes on to win in impressive fashion.

Ivanhoe 9.46h, 4-Gb hash

Code: Select all

23	00:52	 535 449 184	10 329 000	 0,00	Nf3-g5 h7-h6 Qd1-h5 Qa2xb2 Bd3-h7+ Kg8-h8 Bh7-e4 f7-f5 Be4xc6 b7xc6 Ng5-f7+ Kh8-g8 Nf7xh6+ g7xh6 Qh5-g6+ Kg8-h8 Qg6xh6+ Kh8-g8 Qh6-g6+ Kg8-h8 Qg6-h6+
23+	01:41	1 050 041 293	10 490 000	+0,08	Bd3xh7+

(...)

  28+	2:05:00	81 622 587 833	10 883 000	+2,06	Bd3xh7+ Kg8xh7 Nf3-g5+ Kh7-g6 Qd1-g4 f7-f5 Qg4-g3 Kg6-f6 Qg3-h4 Qa2-d5 Ng5-h7+ Kf6-f7 Qh4-h5+ g7-g6 Nh7-g5+ Kf7-f6 Qh5-h4 f5-f4 Be3xf4 Kf6-e7 Ng5-h7+ Ke7-e8 Nh7xf8 Qd5-h5 Qh4xh5 g6xh5 Nf8-h7 Bb6xd4 b2-b3 e6-e5 Bf4-e3 Ke8-f7 Be3xd4 e5xd4 Nh7-g5+ Kf7-g6 Ng5-e4 Ra8-b8 Ne4-d6 Bc8-d7 Rf1-e1 b7-b6 Re1-d1 a7-a6 b3-b4 b6-b5 Nd6-e4 Rb8-e8 f2-f3 h5-h4 Kg1-f2 

Position reached after this variation, Black to play

[d]4r3/3b4/p1n3k1/1p6/1P1pN2p/5P1P/5KP1/2RR4 b - - 1 26

[S.Taylor]

That's a nice one. Some Bxh7 sacs are easy but this one isn't. Houdini 1.5a solved it in just under 2 minutes on my older quad system.

--Jon

Agreed. Definitely not an easy one.

But it doesn't take long to see that you atleast get a pawn and the exchange for the bishop. (I just had a glimpse, so i might be wrong!)

[or a very big battle for mate if Kh4?]

[Graham Banks]

That's a nice one. Some Bxh7 sacs are easy but this one isn't. Houdini 1.5a solved it in just under 2 minutes on my older quad system.

--Jon

Agreed. Definitely not an easy one.

But it doesn't take long to see that you atleast get a pawn and the exchange for the bishop. (I just had a glimpse, so i might be wrong!)

[or a very big battle for mate if Kh4?]

Here is what happened:

[Event "Revving Up The V8s (8 cores)"]
[Site "Auckland"]
[Date "2012.07.10"]
[Round "7"]
[White "Chiron 1.1a 64bit"]
[Black "Deep Shredder 12 x64"]
[Result "1-0"]
[ECO "E56"]
[Annotator "0.30;0.07"]
[PlyCount "53"]
[EventDate "2012.07.09"]
[Source "Banks"]
[TimeControl "40/2400:40/2400:40/2400"]

{Intel(R) Xeon(R) CPU E5430 @ 2.66GHz 2659 MHz W=24.1 plies; 2,
671kN/s; 10 TBAs; SuP2012LV.ctg B=16.6 plies; 3,658kN/s; SuP2012LV.ctg} 1. d4
{[%eval 35534,0] [%emt 0:00:00]} Nf6 {[%eval 35534,0] [%emt 0:00:00]} 2. c4 {
[%eval 35534,0] [%emt 0:00:00]} e6 {[%eval 35534,0] [%emt 0:00:00]} 3. Nc3 {
[%eval 35534,0] [%emt 0:00:00]} Bb4 {[%eval 35534,0] [%emt 0:00:00]} 4. e3 {
[%eval 35534,0] [%emt 0:00:00]} O-O {[%eval 35534,0] [%emt 0:00:00]} 5. Bd3 {
[%eval 35534,0] [%emt 0:00:00]} d5 {[%eval 35534,0] [%emt 0:00:00]} 6. Nf3 {
[%eval 35534,0] [%emt 0:00:00]} c5 {[%eval 35534,0] [%emt 0:00:00]} 7. O-O {
[%eval 35534,0] [%emt 0:00:00]} Nc6 {[%eval 35534,0] [%emt 0:00:00]} 8. h3 {
[%eval 35534,0] [%emt 0:00:00]} cxd4 {[%eval 35534,0] [%emt 0:00:00]} 9. cxd5 {
[%eval 30,23] [%emt 0:01:41]} Nxd5 {[%eval 7,19] [%emt 0:02:30]} 10. Nxd5 {
[%eval 27,23] [%emt 0:01:25]} Qxd5 {[%eval 3,16] [%emt 0:00:08]} 11. exd4 {
[%eval 34,23] [%emt 0:03:42]} Ba5 {(Bd7) [%eval 16,19] [%emt 0:03:44]} 12. Be3
{[%eval 46,23] [%emt 0:02:18]} Bb6 {[%eval 19,19] [%emt 0:01:47]} 13. Rc1 {
(Qc2) [%eval 39,23] [%emt 0:03:53]} Qxa2 {(h6) [%eval 1,17] [%emt 0:00:15]} 14.
Bxh7+ {(Ng5) [%eval 207,25] [%emt 0:00:55]} Kxh7 {[%eval 2,16] [%emt 0:01:03]}
15. Ng5+ {[%eval 230,24] [%emt 0:01:37]} Kg6 {[%eval 3,16] [%emt 0:01:19]} 16.
Qg4 {[%eval 230,26] [%emt 0:00:46]} f5 {[%eval 3,15] [%emt 0:02:10]} 17. Qg3 {
[%eval 230,24] [%emt 0:01:11]} Rf7 {(Re8) [%eval 91,15] [%emt 0:04:44]} 18. Qh4
{[%eval 264,24] [%emt 0:01:28]} Qd5 {(Rf8) [%eval 16,15] [%emt 0:00:42]} 19.
Qh7+ {(g4) [%eval 398,23] [%emt 0:02:00]} Kf6 {[%eval 1,10] [%emt 0:00:00]} 20.
Qg8 {(Nxf7) [%eval 445,25] [%emt 0:00:55]} Nd8 {
(Qd7) [%eval 277,18] [%emt 0:02:12]} 21. Nf3 {
(h4) [%eval 540,24] [%emt 0:00:40]} f4 {[%eval 451,17] [%emt 0:01:56]} 22. Bxf4
{[%eval 540,23] [%emt 0:00:04]} Qh5 {(Bd7) [%eval 460,16] [%emt 0:00:52]} 23.
Rfe1 {(Bg5+) [%eval 713,23] [%emt 0:01:00]} Qf5 {
(Bd7) [%eval 554,17] [%emt 0:02:43]} 24. Bd2 {
(Bg5+) [%eval 719,23] [%emt 0:01:02]} Ke7 {(Bd7) [%eval 752,17] [%emt 0:02:13]}
25. Bb4+ {[%eval 35534,24] [%emt 0:00:58]} Kf6 {[%eval 757,18] [%emt 0:00:55]}
26. Re5 {[%eval 941,26] [%emt 0:00:58]} Qf4 {[%eval 766,17] [%emt 0:00:45]} 27.
Bd2 {[%eval 35534,27] [%emt 0:00:58]} 1-0

[zullil]

14. Bxh7+ {(Ng5) [%eval 207,25] [%emt 0:00:55]}

Does this mean that Chiron took 55 seconds to play Bxh7, with an evaluation of 207 centipawns?

Here's what Critter-1.6a (8 threads, 2GB hash) gave on my machine after 58 seconds:

Code: Select all

17/74  00:58  519141K   8823K  +1.42 Bxh7+ Kxh7 Ng5+ Kg6 Qg4 f5 Qg3 Kf6 Qh4 Qd5 Nh7+ Kf7 Qh5+ g6 Ng5+ Ke7 Qh7+ Kd8 Qg7 Qd6 Nh7 Re8 Bg5+ Ne7 Nf6 Bd7 Nxe8 Bxe8 Rfe1 Ba5 Re5 Bb4 Rec5 Bxc5 dxc5 Qd7 Qf8 Kc7 Bxe7 Qd2 Bd6+ Kc6 Rb1 Kb5 Qf6 Qd5 b4 Bc6 f3 Qa2 Ra1 Qd2 Qe5 Kxb4 Rb1+ Ka3 Qxe6

The search is ongoing. Here's the most recent output:

Code: Select all

25/114+ 79:56   50193M  10465K  +2.45 Bxh7+ Kxh7 Ng5+ Kg6 Qg4 f5 Qg3 Kf6 Qh4 Qd5 Nh7+ Kf7 Qh5+ g6 Ng5+ Kf6 Qh4 f4 Bxf4 Ke7 Ne4+ Ke8 Nd6+ Kd7 Qh7+ Kd8 Be5 Ne7 Qg7 Rg8 Qf7 Bd7 Bf6 Qxd6

[fern]

I wonder if a line code could be programmed in order to put in suspicion an engine if he see a too much very good move in material terms. After all, those are the traps of sacrifices...

Fern