Houdini 3 reducing the depth feature

[Master Om]

It works! I'm using a similar method in the new Vitruvius (less sophisticated than the one implemented by Robert) and results are surprising. For example, in this position:

FEN: r3b1k1/1rq1bppp/2n1p3/ppPpP3/2pP1P2/P1P1BB2/3Q2PP/1R2NRK1 w - - 0 1

Vitruvius 1.16 (in "Deep-Analysis" mode) finds, surprisingly quickly, the best move: 1.f5!
This without going too deep... I believe that this move is very difficult to find, for any engine ... Except, of course, Vitruvius 1.16 in "Deep-Analysis" mode and Houdini-3 in Tactical Mode.

Another example:

FEN: r1bq1rk1/ppp3b1/3p1nn1/P1PPp1Bp/1P2P3/1QNN1Pp1/4B1P1/R4RK1 b - - 0 1

Vitruvius 1.16 "Deep-Analysis" finds the best move (1.Bh6!) in only 4 minuts, on my computer. Depth: 21 plyes. At the same time Stockfish is at depth=27 plyes... And has not yet found the solution!

I do not own Houdini-3 and then I could not do the tests, but I think between the various engines is the only one who can find, such as Vitruvius 1.16, the solution in reasonable time...

P.S. I'm sorry, but I don't know how you post pictures.

Roberto.

What???? 4 minutes ? what the hell ? Stockfish requires only 19 seconds to find it!!. (That means u don't know how to use stockfish)

New game
[D]r1bq1rk1/ppp3b1/3p1nn1/P1PPp1Bp/1P2P3/1QNN1Pp1/4B1P1/R4RK1 b - - 0 1

Analysis by Stockfish 2.3 JA 64bit SSE4.2:

1...Bh6 2.Bxf6 Rxf6 3.Nb5 Be3+ 4.Kh1 Rf8
-/+ (-1.15 ++) Depth: 22/40 00:00:19 143mN
(Prakash, Bhubaneswar 21.10.2012)


And who says Stock doesnot find ?

New game
[d]r3b1k1/1rq1bppp/2n1p3/ppPpP3/2pP1P2/P1P1BB2/3Q2PP/1R2NRK1 w - - 0 1

Analysis by Stockfish 2.3 JA 64bit SSE4.2:

1.f5 Qc8 2.Bg4 h5 3.Bh3 b4 4.Qd1
+/- (1.09 ++) Depth: 27/43 00:00:53 381mN
(Prakash, Bhubaneswar 21.10.2012)

[lkaufman]

Houdini 3 finds THE score and PV.

The technical idea behind this is the difference in pruning and extensions between PV nodes and non-PV nodes.
If you find a fail-high for a non-PV node at depth 30, it's very likely that when it becomes a PV node the fail-high can be found at reduced depth.

Houdini 3 demonstrates how remarkably effective this idea is.

I think I understand the idea based on all the comments here, we'll probably try it in a day or two. You say that the idea is "remarkably effective", implying that it accounts for a significant part of the 55-60 elo gain in Houdini 3. Would you mind estimating how much of that gain you attribute to this idea? Also, if the gain is significant, why are you willing to share it with the competition? I'm not complaining, just curious.

[Uri Blass]

Houdini 3 finds THE score and PV.

The technical idea behind this is the difference in pruning and extensions between PV nodes and non-PV nodes.
If you find a fail-high for a non-PV node at depth 30, it's very likely that when it becomes a PV node the fail-high can be found at reduced depth.

Houdini 3 demonstrates how remarkably effective this idea is.

I think I understand the idea based on all the comments here, we'll probably try it in a day or two. You say that the idea is "remarkably effective", implying that it accounts for a significant part of the 55-60 elo gain in Houdini 3. Would you mind estimating how much of that gain you attribute to this idea? Also, if the gain is significant, why are you willing to share it with the competition? I'm not complaining, just curious.

I will be surprised if this idea is responsible for a significant part of the 55-60 elo gain in houdini3(maybe it is responsible for 1 or 2 elo improvement and I will be very surprised if it is responsible for more than 5 elo improvement).

In most positions it is irrelevant(houdini does not need to reduce the depth) and even in positions when houdini reduce the depth the advantage is not a better move but using less time for the specific move or better information in the hash for the next move.

[Uri Blass]

It works! I'm using a similar method in the new Vitruvius (less sophisticated than the one implemented by Robert) and results are surprising. For example, in this position:

FEN: r3b1k1/1rq1bppp/2n1p3/ppPpP3/2pP1P2/P1P1BB2/3Q2PP/1R2NRK1 w - - 0 1

Vitruvius 1.16 (in "Deep-Analysis" mode) finds, surprisingly quickly, the best move: 1.f5!
This without going too deep... I believe that this move is very difficult to find, for any engine ... Except, of course, Vitruvius 1.16 in "Deep-Analysis" mode and Houdini-3 in Tactical Mode.

Another example:

FEN: r1bq1rk1/ppp3b1/3p1nn1/P1PPp1Bp/1P2P3/1QNN1Pp1/4B1P1/R4RK1 b - - 0 1

Vitruvius 1.16 "Deep-Analysis" finds the best move (1.Bh6!) in only 4 minuts, on my computer. Depth: 21 plyes. At the same time Stockfish is at depth=27 plyes... And has not yet found the solution!

I do not own Houdini-3 and then I could not do the tests, but I think between the various engines is the only one who can find, such as Vitruvius 1.16, the solution in reasonable time...

P.S. I'm sorry, but I don't know how you post pictures.

Roberto.

What???? 4 minutes ? what the hell ? Stockfish requires only 19 seconds to find it!!. (That means u don't know how to use stockfish)

New game
[D]r1bq1rk1/ppp3b1/3p1nn1/P1PPp1Bp/1P2P3/1QNN1Pp1/4B1P1/R4RK1 b - - 0 1

Analysis by Stockfish 2.3 JA 64bit SSE4.2:

1...Bh6 2.Bxf6 Rxf6 3.Nb5 Be3+ 4.Kh1 Rf8
-/+ (-1.15 ++) Depth: 22/40 00:00:19 143mN
(Prakash, Bhubaneswar 21.10.2012)


And who says Stock doesnot find ?

New game
[d]r3b1k1/1rq1bppp/2n1p3/ppPpP3/2pP1P2/P1P1BB2/3Q2PP/1R2NRK1 w - - 0 1

Analysis by Stockfish 2.3 JA 64bit SSE4.2:

1.f5 Qc8 2.Bg4 h5 3.Bh3 b4 4.Qd1
+/- (1.09 ++) Depth: 27/43 00:00:53 381mN
(Prakash, Bhubaneswar 21.10.2012)

The positions are clearly bad and I believe that it may be possible also to find Bh6 for the wrong reason.
I used stockfish2.2.2 that is supposed to have a similiar strength to 2.3
and I did not download 2.3 because I saw no strength improvement in the rating list and 2.2.2 always likes Qe8 in more than 20 minutes that I gave it to search.

The score for Qe8 is 0.00 at depth 29 after more than 20 minutes of search and more than 4,400,000,000 nodes

Here is the analysis under arena

FEN: r1bq1rk1/ppp3b1/3p1nn1/P1PPp1Bp/1P2P3/1QNN1Pp1/4B1P1/R4RK1 b - - 0 1

Stockfish-222-64-ja-intel:
1/1 00:00 133 133,000 -2.02 Bc8-d7
2/2 00:00 357 357,000 -2.18 Bc8-d7 c5-c6
3/3 00:00 1,973 493,250 -2.18 Rf8-f7 c5xd6 c7xd6
4/4 00:00 2,763 41,863 -2.18 Rf8-f7 c5-c6 b7xc6 d5xc6
5/7 00:00 18,764 131,216 -1.97 h5-h4 c5xd6 c7xd6 Rf1-c1 h4-h3
6/7 00:00 20,099 135,804 -1.93 h5-h4 c5xd6 c7xd6 f3-f4 e5xf4 Nd3xf4 Ng6xf4 Rf1xf4
7/11 00:00 32,310 198,220 -1.37 h5-h4 c5-c6 h4-h3 g2xh3 Ra8-b8 c6xb7 Bc8xb7 f3-f4 e5xf4 Nd3xf4 Ng6xf4 Bg5xf4
8/13 00:00 52,336 299,062 -1.37 h5-h4 c5-c6 h4-h3 g2xh3 Ra8-b8 c6xb7 Bc8xb7 f3-f4 e5xf4 Nd3xf4 Ng6xf4 Rf1xf4
9/13 00:00 58,301 327,533 -1.37 h5-h4 c5-c6 h4-h3 g2xh3 Ra8-b8 c6xb7 Bc8xb7 f3-f4 e5xf4 Nd3xf4 Ng6xf4 Rf1xf4
10/13 00:00 76,429 406,537 -1.45 h5-h4 c5-c6 h4-h3 g2xh3 Ra8-b8 c6xb7 Bc8xb7 f3-f4 e5xf4 Nd3xf4 Ng6xf4 Rf1xf4 a7-a6
11/19 00:00 304,983 1,125,398 -1.09 h5-h4 f3-f4 e5xf4 c5xd6 c7xd6 Nd3xf4 Ng6xf4 Rf1xf4 h4-h3 g2xh3 Qd8-e8 Bg5-h4 Qe8-g6 a5-a6 b7xa6
12/19 00:00 822,922 1,909,331 -1.09 h5-h4 f3-f4 e5xf4 Nd3xf4 Ng6xf4 Rf1xf4 h4-h3 g2xh3 Qd8-e8 c5xd6 c7xd6 Bg5-h4 Qe8-g6 a5-a6 b7xa6
13/19 00:00 1,455,533 2,314,042 -0.92 h5-h4 f3-f4 e5xf4 Nd3xf4 Ng6xf4 Rf1xf4 h4-h3 Ra1-f1 Qd8-e7 c5xd6 c7xd6 Nc3-b5 h3-h2+ Kg1-h1 Bc8-d7 Qb3xg3
14/26 00:00 1,920,018 2,449,002 -1.29 h5-h4 f3-f4 e5xf4 Nd3xf4 Ng6xf4 Rf1xf4 h4-h3 Ra1-f1 a7-a6 g2xh3 Qd8-e7 c5xd6 c7xd6 Rf1-f3 Qe7-e5 Rf3xg3 Nf6xe4 Rf4xf8+ Kg8xf8 Nc3xe4 Qe5xe4
15/26 00:01 2,588,624 2,588,624 -1.41 h5-h4 f3-f4 e5xf4 Nd3xf4 Ng6xf4 Rf1xf4 h4-h3 Ra1-f1 h3xg2 Rf1-f3 Qd8-e7 Rf3xg3 Qe7-e5 Kg1xg2 Nf6-g4 Rf4xf8+ Kg8xf8 Be2xg4 Qe5xg5 Bg4-e6
16/27+ 00:02 6,225,848 2,957,647 -1.15 Qd8-e8 a5-a6 b7xa6 c5-c6 Ra8-b8 Bg5-e3 h5-h4 Be3xa7 Rb8-a8
16/27+ 00:02 7,763,775 3,045,812 -0.87 Qd8-e8 a5-a6 b7xa6 c5-c6 Ra8-b8 Bg5-e3 h5-h4 Be3xa7 Rb8-a8
16/27 00:03 10,680,690 3,134,005 -0.88 Qd8-e8 Qb3-a4 Qe8xa4 Nc3xa4 h5-h4 Na4-c3 h4-h3 c5xd6 c7xd6 Nc3-b5 Nf6-h7 Bg5-e3 Ng6-h4 g2xh3 Bc8xh3 Nb5xd6 Bh3xf1 Ra1xf1
17/27+ 00:03 11,851,663 3,152,876 -0.80 Qd8-e8 Qb3-a4 Qe8xa4 Nc3xa4 h5-h4 Na4-c3 h4-h3 c5xd6 h3-h2+ Kg1-h1 c7xd6 Nc3-b5 Nf6-e8 Rf1-c1 Rf8-f7 Nb5xa7 Bc8-d7
17/27- 00:04 12,762,823 3,177,993 -0.96 Qd8-e8 Qb3-c4 Nf6-h7 Nc3-b5 Qe8-d7 c5-c6 b7xc6 d5xc6+ Qd7-f7 Qc4xf7+ Rf8xf7 Bg5-d8 Bc8-a6 Nb5xc7 Ba6xd3 Nc7xa8 Bd3xe2 c6-c7 Be2-a6 b4-b5
17/27- 00:04 12,866,282 3,169,034 -1.27 Qd8-e8 Qb3-c4 Nf6-h7 Nc3-b5 Qe8-d7 c5-c6 b7xc6 d5xc6+ Qd7-f7 Qc4xf7+ Rf8xf7 Bg5-d8 Bc8-a6 Nb5xc7 Ba6xd3 Nc7xa8 Bd3xe2 c6-c7 Be2-a6 b4-b5
17/29 00:04 13,663,611 3,168,006 -0.80 Qd8-e8 Qb3-c4 Nf6-h7 Nc3-b5 Qe8-f7 Qc4-c1 a7-a6 c5xd6 a6xb5 d6xc7 Ng6-f4 Bg5xf4 e5xf4 Qc1-c5 Bg7xa1 Rf1xa1 Qf7-f6 Ra1-b1 Qf6-h4 Nd3-e5 Qh4-h2+ Kg1-f1 Rf8-e8 Ne5-g6 Qh2-h1+ Qc5-g1
18/29 00:04 15,540,271 3,175,372 -0.80 Qd8-e8 Qb3-c4 Nf6-h7 Nc3-b5 Qe8-f7 Qc4-c1 a7-a6 c5xd6 a6xb5 d6xc7 Ng6-f4 Bg5xf4 e5xf4 Qc1-c5 Bg7xa1 Rf1xa1 Qf7-f6 Ra1-b1 Qf6-h4 Nd3-e5 Qh4-h2+ Kg1-f1 Rf8-e8 Ne5-g6 Qh2-h1+ Qc5-g1
19/30+ 00:05 19,185,671 3,218,532 -0.72 Qd8-e8 Qb3-c4 Nf6-h7 Nc3-b5 Qe8-f7 Qc4-c1 a7-a6 c5xd6 a6xb5 d6xc7 Ng6-f4 Bg5xf4 e5xf4 Qc1-c5 Bg7xa1 Rf1xa1 Qf7-f6 Ra1-b1 Qf6-h4 Nd3-e5 Rf8-e8 Ne5-g6 Qh4-g5
19/34+ 00:07 24,390,127 3,225,354 -0.64 Qd8-e8 Qb3-c4 Nf6-h7 Nc3-b5 Qe8-f7 Qc4-c1 a7-a6 c5xd6 a6xb5 d6xc7 h5-h4 Bg5-d8 h4-h3
19/34+ 00:08 29,153,498 3,260,652 -0.52 Qd8-e8 Qb3-c4 Nf6-h7 Nc3-b5 Qe8-f7 Qc4-c1 a7-a6 c5xd6 a6xb5 d6xc7 h5-h4 Bg5-d8 h4-h3
19/34 00:11 36,316,219 3,269,669 -0.52 Qd8-e8 Qb3-a4 Qe8xa4 Nc3xa4 h5-h4 Na4-c3 h4-h3 c5xd6 h3-h2+ Kg1-h1 c7xd6 Nc3-b5 Nf6-e8 Rf1-c1 Rf8-f7 Nb5xa7 Bc8-d7 Bg5-e3 Ng6-f4 Be2-f1 Bg7-f6 Ra1-a3 Rf7-h7 Nd3xf4 e5xf4
20/34- 00:12 42,192,034 3,285,728 -0.60 Qd8-e8 Qb3-a4 Qe8xa4 Nc3xa4 h5-h4 b4-b5 h4-h3 b5-b6 h3xg2 Kg1xg2 Nf6-h7 Bg5-d2
20/34- 00:12 42,510,618 3,282,166 -0.68 Qd8-e8 Qb3-a4 Qe8xa4 Nc3xa4 h5-h4 b4-b5 h4-h3 b5-b6 h3xg2 Kg1xg2 Nf6-h7 Bg5-d2
20/34- 00:14 46,401,389 3,299,537 -0.80 Qd8-e8 Qb3-a4 Qe8xa4 Nc3xa4 h5-h4 b4-b5 h4-h3 b5-b6 h3xg2 Kg1xg2 Nf6-h7 Bg5-d2
20/34 00:15 52,638,688 3,307,696 -0.80 Qd8-e8 Qb3-a4 h5-h4 Qa4xe8 Nf6xe8 Rf1-c1 h4-h3 Kg1-h1 Kg8-f7 Be2-f1 Rf8-h8 Nc3-e2 Rh8-h5 Bg5-e3 Rh5-h7 Ne2xg3 d6xc5 g2xh3 Bc8xh3 Kh1-g1 Bh3xf1 Kg1xf1 c5xb4 Nd3xb4
21/34+ 00:20 67,638,463 3,342,316 -0.72 Qd8-e8 Qb3-a4 Bc8-d7 Qa4-a2 Nf6-h7 Qa2-d2 h5-h4 a5-a6 b7xa6 c5-c6 Bd7-c8 Bg5-e3 h4-h3
21/34+ 00:22 75,769,734 3,345,389 -0.64 Qd8-e8 Qb3-a4 Bc8-d7 Qa4-a2 Nf6-h7 Qa2-d2 h5-h4 a5-a6 b7xa6 c5xd6 c7xd6 Ra1xa6 Nh7xg5 Qd2xg5 Rf8-f6 Rf1-c1 h4-h3 g2xh3 Qe8-c8 Ra6-a1 g3-g2 Qg5xg2 Bd7xh3
21/34+ 00:24 83,163,740 3,360,965 -0.52 Qd8-e8 Qb3-a4 Bc8-d7 Qa4-a2 Nf6-h7 c5xd6 Nh7xg5 d6xc7 Ng6-h4 Nd3-c5 Bd7-h3 d5-d6+ Kg8-h7 Qa2-d2 Qe8-g6 g2xh3 Ng5xh3+ Kg1-h1 g3-g2+
21/34+ 00:29 98,004,629 3,372,608 -0.34 Qd8-e8 b4-b5 Nf6-h7 Bg5-e3 Qe8-e7 c5xd6 c7xd6 f3-f4 e5xf4 Nd3xf4 Qe7-h4
21/34 00:32 108,037,290 3,367,850 -0.44 Qd8-e8 Qb3-a4 Bc8-d7 Qa4-a2 Nf6-h7 Qa2-d2 h5-h4 a5-a6 b7xa6 c5xd6 c7xd6 Ra1xa6 Nh7xg5 Qd2xg5 Rf8-f6 Rf1-c1 h4-h3 g2xh3 Ng6-f4 Nd3xf4 e5xf4 Be2-f1 Rf6-g6 Qg5xf4 Bg7-d4+ Kg1-h1
22/34+ 00:35 120,920,205 3,365,625 -0.36 Qd8-e8 Qb3-a4 Bc8-d7 Qa4-a2 Nf6-h7 Qa2-d2 h5-h4 a5-a6 b7xa6 c5xd6 c7xd6 Ra1xa6 Nh7xg5 Qd2xg5 Rf8-f6 Rf1-c1 h4-h3 g2xh3 Ng6-f4 Nd3xf4 e5xf4 Be2-f1 Rf6-g6 Qg5xf4 Bg7-h6
22/35- 00:40 136,284,507 3,372,794 -0.52 Qd8-e8 Nc3-b5 Qe8xb5 c5-c6 Nf6xe4 f3xe4 Bc8-g4 Rf1xf8+ Ra8xf8 Be2-f1 Rf8xf1+ Ra1xf1 Bg4-e2 Rf1-e1 Be2xd3 c6xb7 Ng6-f4 Bg5xf4 e5xf4 Qb3xd3 Qb5xd3 b7-b8Q+ Kg8-h7 Qb8-e8
22/35- 00:40 136,864,875 3,372,052 -0.64 Qd8-e8 Nc3-b5 Qe8xb5 c5-c6 Nf6xe4 f3xe4 Bc8-g4 Rf1xf8+ Ra8xf8 Be2-f1 Rf8xf1+ Ra1xf1 Bg4-e2 Rf1-e1 Be2xd3 c6xb7 Ng6-f4 Bg5xf4 e5xf4 Qb3xd3 Qb5xd3 b7-b8Q+ Kg8-h7 Qb8-e8
22/35 00:41 140,265,641 3,372,662 -0.68 Qd8-e8 Nc3-b5 Qe8xb5 c5-c6 Nf6xe4 f3xe4 Bc8-g4 Rf1xf8+ Ng6xf8 Be2xg4 h5xg4 c6xb7 Qb5xb7 Qb3-d1 Qb7-b5 Ra1-c1 Qb5-d7 Qd1-e2 Nf8-h7 Bg5-h4 Bg7-h6 Rc1-c6 Qd7-g7 Bh4xg3 Nh7-f6 Bg3-h4 Ra8-f8
23/39+ 00:44 150,124,201 3,378,588 -0.60 Qd8-e8 Nc3-b5 Qe8xb5 c5-c6 Nf6xe4 f3xe4 Bc8-g4 Rf1xf8+ Ng6xf8 Be2xg4 h5xg4 c6xb7 Qb5xb7 Qb3-d1 Nf8-h7 Bg5-h4 a7-a6 Qd1xg4 Qb7-a7+ Kg1-h1 Qa7-d4 Ra1-d1 Ra8-f8 Bh4xg3 Nh7-f6
23/39+ 00:47 161,153,350 3,383,797 -0.52 Qd8-e8 Nc3-b5 Qe8xb5 c5-c6 Nf6xe4 f3xe4 Bc8-g4 Rf1xf8+ Ng6xf8 Be2xg4 h5xg4 c6xb7 Qb5xb7 Qb3-d1 Nf8-h7 Bg5-h4 a7-a6 Qd1xg4 Qb7-a7+ Kg1-h1 Qa7-d4 Ra1-d1 Ra8-f8 Bh4xg3 Nh7-f6
23/39+ 00:51 172,746,264 3,387,248 -0.40 Qd8-e8 Nc3-b1 Nf6-h7 Bg5-e3 Qe8-e7 c5xd6 c7xd6 Rf1-c1 Qe7-h4 Rc1-c7 Qh4-h2+ Kg1-f1 Qh2-h1+ Be3-g1 Ng6-h4
23/46 01:05 222,542,931 3,400,613 -0.44 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 Nf6-h7 Bg5-d2 h5-h4 Rf1-c1 Rf8-f7 Bd2-e3 h4-h3 g2xh3 Ng6-h4 Nd3-e1 Rf7-f6 c5xd6 c7xd6 Be2xa6 Bc8xh3 Ba6-b7 Ra8-f8 Ra4xa7 Nh4xf3+ Ne1xf3 Rf6xf3
24/46+ 01:18 267,870,042 3,393,466 -0.36 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 Rf1-c1 h4-h3 g2xh3 Nf6-h5 c5xd6 c7xd6 h3-h4 Bg7-f6 Bg5xf6 Rf8xf6
24/46+ 01:28 300,248,887 3,392,566 -0.28 Qd8-e8 f3-f4 e5xf4 Qb3-c4 d6xc5 b4xc5 Nf6-g4 Be2xg4 Bc8xg4 a5-a6 b7xa6 Nd3xf4 Qe8-e5 Nf4xg6 Rf8xf1+ Ra1xf1 Qe5xg5 d5-d6+ Kg8-h7 Ng6-f4 Qg5-h4
24/46 01:37 328,036,975 3,379,002 -0.36 Qd8-e8 Nc3-b5 Qe8xb5 c5-c6 Nf6xe4 f3xe4 Bc8-g4 Rf1xf8+ Ng6xf8 c6xb7 Qb5xb7 Be2xg4 h5xg4 Qb3-d1 Nf8-h7 Qd1xg4 Nh7xg5 Qg4xg5 c7-c6 d5xc6 Qb7xc6 Qg5-e3 Ra8-f8 Ra1-c1 Qc6-b7 Rc1-e1 Qb7-b5 Qe3xg3 a7-a6
25/46+ 01:47 360,817,742 3,361,008 -0.28 Qd8-e8 Nc3-b5 Qe8xb5 c5-c6 Nf6xe4 f3xe4 Bc8-g4 Rf1xf8+ Ra8xf8 Be2-f1 b7xc6 Nd3-f4 e5xf4 d5xc6+ Kg8-h7 Bf1xb5 Bg7xa1 Bb5-a4 f4-f3 Qb3-e3 Ng6-e5 Qe3xa7 f3-f2+ Kg1-f1 Bg4-e2+
25/46+ 02:03 412,706,705 3,354,929 -0.20 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 Rf1-a1 Ra8-b8 Bg5-e3 h4-h3 g2xh3 Nf6-h5 c5xd6 c7xd6 Be3xa7 Rb8-a8 Ba7-e3 Ng6-h4 Nd3-e1 Nh5-f4
25/46 02:21 473,730,169 3,342,058 -0.40 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 Rf1-a1 Ra8-b8 Ra4xa6 Bc8xa6 Ra1xa6 Nf6-h5 c5xd6 c7xd6 Ra6xd6 Nh5-f4 Nd3xf4 Ng6xf4 Bg5xh4 Nf4xe2+ Nc3xe2 a7-a5 Bh4xg3 a5xb4 Rd6-g6 Kg8-h7 Rg6xg7+ Kh7xg7 Bg3xe5+ Kg7-f7 Be5xb8
26/48- 02:27 492,034,805 3,344,263 -0.48 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 Rf1-a1 Ra8-b8 Ra4xa6 Bc8xa6
26/48+ 02:48 564,386,221 3,340,571 -0.32 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 Rf1-a1 Ra8-b8 Ra4xa6 Bc8xa6 Ra1xa6 Nf6-h5 c5xd6 c7xd6 Ra6xd6 Nh5-f4 Nd3xf4 Ng6xf4 Bg5xh4 Nf4xe2+
26/48- 03:19 676,114,883 3,387,145 -0.56 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 Rf1-a1 Ra8-b8 Ra4xa6 Bc8xa6 Ra1xa6 Nf6-h5 c5xd6 c7xd6 Ra6xd6
26/48+ 03:42 752,539,268 3,382,092 -0.20 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 Rf1-a1 Ra8-b8 Ra4xa6 Bc8xa6 Ra1xa6 Rf8-c8 Ra6xa7 Nf6-h5 Be2-f1 h4-h3 g2xh3 Nh5-f4 Nd3xf4 Ng6xf4 Bg5-e7 d6xc5 b4xc5 Rb8-b3
26/48 04:01 817,478,144 3,390,731 -0.44 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 Rf1-a1 Ra8-b8 Bg5-e3 h4-h3 c5xd6 c7xd6 Be3xa7 h3-h2+ Kg1-h1 Rb8-a8 Ba7-b6 Bc8-d7 Ra4xa6 Ra8xa6 Ra1xa6 Rf8-c8 Ra6-a3 Nf6-h5 Bb6-e3
27/48+ 04:34 931,763,346 3,391,103 -0.36 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 Rf1-a1 Ra8-b8 Bg5-e3 h4-h3 c5xd6 c7xd6 Be3xa7 Rb8-a8 Ba7-e3 Ra8-b8 g2xh3 Nf6-h5 Nd3-e1 Nh5-f4 Be2xa6 Ng6-h4 Ba6xc8 Rf8xc8 Be3xf4 e5xf4
27/48+ 05:07 1,040,760,056 3,387,703 -0.28 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 Rf1-a1 Ra8-b8 Ra4-a5 Rf8-f7 b4-b5 Nf6-h7 Bg5-e3 h4-h3 g2xh3 Ng6-f4
27/48+ 06:14 1,268,564,962 3,389,419 -0.16 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 c5xd6 c7xd6 Rf1-c1 h4-h3 g2xh3 Nf6-h5 Kg1-g2 Bg7-f6 Bg5xf6 Rf8xf6
27/48 08:06 1,665,309,791 3,421,382 -0.12 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 Rf1-a1 Ra8-b8 Ra4xa6 Bc8xa6 Ra1xa6 Rf8-c8 b4-b5 Nf6-h7 Bg5-d2 Ng6-f4 Nd3xf4 e5xf4 Nc3-a4 Bg7-d4+ Kg1-f1 Bd4-e3 Bd2xe3 f4xe3 Ra6xa7 Rb8-a8 Ra7-a6 Kg8-f7 Na4-b2 d6xc5 Nb2-c4 Nh7-g5 Nc4xe3 Rc8-b8 Ra6-c6
28/50- 08:19 1,712,074,572 3,426,664 -0.20 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 Rf1-a1 Ra8-b8 Bg5-e3
28/50 09:03 1,857,112,946 3,415,260 -0.16 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 Rf1-a1 Ra8-b8 Bg5-e3 h4-h3 c5xd6 c7xd6 Be3xa7 Rb8-a8 Ba7-e3 Nf6-h5 b4-b5 Nh5-f4 Nd3xf4 Ng6xf4 Be2-f1 h3xg2 Bf1xg2 Bc8-h3 Be3xf4 e5xf4 Ra4xa6 Ra8xa6 Ra1xa6
29/50+ 13:03 2,672,317,645 3,410,025 -0.08 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Bc8-d7 c5-c6 Bd7-c8 Qa4-a5 Nf6-h7 Bg5-e3 Qe8-e7 Rf1-b1 Qe7-h4 Qa5xc7
29/50+ 16:05 3,252,662,444 3,369,300 0.00 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Bc8-d7 c5-c6 Bd7-c8 Bg5xf6 Bg7xf6 Nd3-e1 Bf6-g5 Ne1-c2 h5-h4 Qa4-a5 h4-h3 Qa5xc7 h3xg2 Rf1-b1 Ng6-f4 Qc7xd6 Nf4xe2+ Nc3xe2
29/50 22:15 4,455,995,314 3,337,464 0.00 Qd8-e8 Nc3-b5 Qe8xb5 c5-c6 Nf6xe4 Nd3-c1 Qb5xe2 Nc1xe2 Ne4xg5 Ne2xg3 Ng6-f4 Qb3-c4 b7xc6 Qc4xc6 Ra8-b8 Qc6xc7 Rb8xb4 Ra1-b1 Rb4-d4 Rb1-b8 Rd4-d2 Ng3xh5 Nf4xh5 Rb8xc8 Nh5-f4 Rc8xf8+ Bg7xf8 Qc7xa7 Rd2xg2+ Kg1-h1 Rg2-g3 Qa7-e3 Rg3-h3+ Kh1-g1 Rh3-g3+ Kg1-h1

[Master Om]

It works! I'm using a similar method in the new Vitruvius (less sophisticated than the one implemented by Robert) and results are surprising. For example, in this position:

FEN: r3b1k1/1rq1bppp/2n1p3/ppPpP3/2pP1P2/P1P1BB2/3Q2PP/1R2NRK1 w - - 0 1

Vitruvius 1.16 (in "Deep-Analysis" mode) finds, surprisingly quickly, the best move: 1.f5!
This without going too deep... I believe that this move is very difficult to find, for any engine ... Except, of course, Vitruvius 1.16 in "Deep-Analysis" mode and Houdini-3 in Tactical Mode.

Another example:

FEN: r1bq1rk1/ppp3b1/3p1nn1/P1PPp1Bp/1P2P3/1QNN1Pp1/4B1P1/R4RK1 b - - 0 1

Vitruvius 1.16 "Deep-Analysis" finds the best move (1.Bh6!) in only 4 minuts, on my computer. Depth: 21 plyes. At the same time Stockfish is at depth=27 plyes... And has not yet found the solution!

I do not own Houdini-3 and then I could not do the tests, but I think between the various engines is the only one who can find, such as Vitruvius 1.16, the solution in reasonable time...

P.S. I'm sorry, but I don't know how you post pictures.

Roberto.

What???? 4 minutes ? what the hell ? Stockfish requires only 19 seconds to find it!!. (That means u don't know how to use stockfish)

New game
[D]r1bq1rk1/ppp3b1/3p1nn1/P1PPp1Bp/1P2P3/1QNN1Pp1/4B1P1/R4RK1 b - - 0 1

Analysis by Stockfish 2.3 JA 64bit SSE4.2:

1...Bh6 2.Bxf6 Rxf6 3.Nb5 Be3+ 4.Kh1 Rf8
-/+ (-1.15 ++) Depth: 22/40 00:00:19 143mN
(Prakash, Bhubaneswar 21.10.2012)


And who says Stock doesnot find ?

New game
[d]r3b1k1/1rq1bppp/2n1p3/ppPpP3/2pP1P2/P1P1BB2/3Q2PP/1R2NRK1 w - - 0 1

Analysis by Stockfish 2.3 JA 64bit SSE4.2:

1.f5 Qc8 2.Bg4 h5 3.Bh3 b4 4.Qd1
+/- (1.09 ++) Depth: 27/43 00:00:53 381mN
(Prakash, Bhubaneswar 21.10.2012)

The positions are clearly bad and I believe that it may be possible also to find Bh6 for the wrong reason.
I used stockfish2.2.2 that is supposed to have a similiar strength to 2.3
and I did not download 2.3 because I saw no strength improvement in the rating list and 2.2.2 always likes Qe8 in more than 20 minutes that I gave it to search.

The score for Qe8 is 0.00 at depth 29 after more than 20 minutes of search and more than 4,400,000,000 nodes

Here is the analysis under arena

FEN: r1bq1rk1/ppp3b1/3p1nn1/P1PPp1Bp/1P2P3/1QNN1Pp1/4B1P1/R4RK1 b - - 0 1

Stockfish-222-64-ja-intel:
1/1 00:00 133 133,000 -2.02 Bc8-d7
2/2 00:00 357 357,000 -2.18 Bc8-d7 c5-c6
3/3 00:00 1,973 493,250 -2.18 Rf8-f7 c5xd6 c7xd6
4/4 00:00 2,763 41,863 -2.18 Rf8-f7 c5-c6 b7xc6 d5xc6
5/7 00:00 18,764 131,216 -1.97 h5-h4 c5xd6 c7xd6 Rf1-c1 h4-h3
6/7 00:00 20,099 135,804 -1.93 h5-h4 c5xd6 c7xd6 f3-f4 e5xf4 Nd3xf4 Ng6xf4 Rf1xf4
7/11 00:00 32,310 198,220 -1.37 h5-h4 c5-c6 h4-h3 g2xh3 Ra8-b8 c6xb7 Bc8xb7 f3-f4 e5xf4 Nd3xf4 Ng6xf4 Bg5xf4
8/13 00:00 52,336 299,062 -1.37 h5-h4 c5-c6 h4-h3 g2xh3 Ra8-b8 c6xb7 Bc8xb7 f3-f4 e5xf4 Nd3xf4 Ng6xf4 Rf1xf4
9/13 00:00 58,301 327,533 -1.37 h5-h4 c5-c6 h4-h3 g2xh3 Ra8-b8 c6xb7 Bc8xb7 f3-f4 e5xf4 Nd3xf4 Ng6xf4 Rf1xf4
10/13 00:00 76,429 406,537 -1.45 h5-h4 c5-c6 h4-h3 g2xh3 Ra8-b8 c6xb7 Bc8xb7 f3-f4 e5xf4 Nd3xf4 Ng6xf4 Rf1xf4 a7-a6
11/19 00:00 304,983 1,125,398 -1.09 h5-h4 f3-f4 e5xf4 c5xd6 c7xd6 Nd3xf4 Ng6xf4 Rf1xf4 h4-h3 g2xh3 Qd8-e8 Bg5-h4 Qe8-g6 a5-a6 b7xa6
12/19 00:00 822,922 1,909,331 -1.09 h5-h4 f3-f4 e5xf4 Nd3xf4 Ng6xf4 Rf1xf4 h4-h3 g2xh3 Qd8-e8 c5xd6 c7xd6 Bg5-h4 Qe8-g6 a5-a6 b7xa6
13/19 00:00 1,455,533 2,314,042 -0.92 h5-h4 f3-f4 e5xf4 Nd3xf4 Ng6xf4 Rf1xf4 h4-h3 Ra1-f1 Qd8-e7 c5xd6 c7xd6 Nc3-b5 h3-h2+ Kg1-h1 Bc8-d7 Qb3xg3
14/26 00:00 1,920,018 2,449,002 -1.29 h5-h4 f3-f4 e5xf4 Nd3xf4 Ng6xf4 Rf1xf4 h4-h3 Ra1-f1 a7-a6 g2xh3 Qd8-e7 c5xd6 c7xd6 Rf1-f3 Qe7-e5 Rf3xg3 Nf6xe4 Rf4xf8+ Kg8xf8 Nc3xe4 Qe5xe4
15/26 00:01 2,588,624 2,588,624 -1.41 h5-h4 f3-f4 e5xf4 Nd3xf4 Ng6xf4 Rf1xf4 h4-h3 Ra1-f1 h3xg2 Rf1-f3 Qd8-e7 Rf3xg3 Qe7-e5 Kg1xg2 Nf6-g4 Rf4xf8+ Kg8xf8 Be2xg4 Qe5xg5 Bg4-e6
16/27+ 00:02 6,225,848 2,957,647 -1.15 Qd8-e8 a5-a6 b7xa6 c5-c6 Ra8-b8 Bg5-e3 h5-h4 Be3xa7 Rb8-a8
16/27+ 00:02 7,763,775 3,045,812 -0.87 Qd8-e8 a5-a6 b7xa6 c5-c6 Ra8-b8 Bg5-e3 h5-h4 Be3xa7 Rb8-a8
16/27 00:03 10,680,690 3,134,005 -0.88 Qd8-e8 Qb3-a4 Qe8xa4 Nc3xa4 h5-h4 Na4-c3 h4-h3 c5xd6 c7xd6 Nc3-b5 Nf6-h7 Bg5-e3 Ng6-h4 g2xh3 Bc8xh3 Nb5xd6 Bh3xf1 Ra1xf1
17/27+ 00:03 11,851,663 3,152,876 -0.80 Qd8-e8 Qb3-a4 Qe8xa4 Nc3xa4 h5-h4 Na4-c3 h4-h3 c5xd6 h3-h2+ Kg1-h1 c7xd6 Nc3-b5 Nf6-e8 Rf1-c1 Rf8-f7 Nb5xa7 Bc8-d7
17/27- 00:04 12,762,823 3,177,993 -0.96 Qd8-e8 Qb3-c4 Nf6-h7 Nc3-b5 Qe8-d7 c5-c6 b7xc6 d5xc6+ Qd7-f7 Qc4xf7+ Rf8xf7 Bg5-d8 Bc8-a6 Nb5xc7 Ba6xd3 Nc7xa8 Bd3xe2 c6-c7 Be2-a6 b4-b5
17/27- 00:04 12,866,282 3,169,034 -1.27 Qd8-e8 Qb3-c4 Nf6-h7 Nc3-b5 Qe8-d7 c5-c6 b7xc6 d5xc6+ Qd7-f7 Qc4xf7+ Rf8xf7 Bg5-d8 Bc8-a6 Nb5xc7 Ba6xd3 Nc7xa8 Bd3xe2 c6-c7 Be2-a6 b4-b5
17/29 00:04 13,663,611 3,168,006 -0.80 Qd8-e8 Qb3-c4 Nf6-h7 Nc3-b5 Qe8-f7 Qc4-c1 a7-a6 c5xd6 a6xb5 d6xc7 Ng6-f4 Bg5xf4 e5xf4 Qc1-c5 Bg7xa1 Rf1xa1 Qf7-f6 Ra1-b1 Qf6-h4 Nd3-e5 Qh4-h2+ Kg1-f1 Rf8-e8 Ne5-g6 Qh2-h1+ Qc5-g1
18/29 00:04 15,540,271 3,175,372 -0.80 Qd8-e8 Qb3-c4 Nf6-h7 Nc3-b5 Qe8-f7 Qc4-c1 a7-a6 c5xd6 a6xb5 d6xc7 Ng6-f4 Bg5xf4 e5xf4 Qc1-c5 Bg7xa1 Rf1xa1 Qf7-f6 Ra1-b1 Qf6-h4 Nd3-e5 Qh4-h2+ Kg1-f1 Rf8-e8 Ne5-g6 Qh2-h1+ Qc5-g1
19/30+ 00:05 19,185,671 3,218,532 -0.72 Qd8-e8 Qb3-c4 Nf6-h7 Nc3-b5 Qe8-f7 Qc4-c1 a7-a6 c5xd6 a6xb5 d6xc7 Ng6-f4 Bg5xf4 e5xf4 Qc1-c5 Bg7xa1 Rf1xa1 Qf7-f6 Ra1-b1 Qf6-h4 Nd3-e5 Rf8-e8 Ne5-g6 Qh4-g5
19/34+ 00:07 24,390,127 3,225,354 -0.64 Qd8-e8 Qb3-c4 Nf6-h7 Nc3-b5 Qe8-f7 Qc4-c1 a7-a6 c5xd6 a6xb5 d6xc7 h5-h4 Bg5-d8 h4-h3
19/34+ 00:08 29,153,498 3,260,652 -0.52 Qd8-e8 Qb3-c4 Nf6-h7 Nc3-b5 Qe8-f7 Qc4-c1 a7-a6 c5xd6 a6xb5 d6xc7 h5-h4 Bg5-d8 h4-h3
19/34 00:11 36,316,219 3,269,669 -0.52 Qd8-e8 Qb3-a4 Qe8xa4 Nc3xa4 h5-h4 Na4-c3 h4-h3 c5xd6 h3-h2+ Kg1-h1 c7xd6 Nc3-b5 Nf6-e8 Rf1-c1 Rf8-f7 Nb5xa7 Bc8-d7 Bg5-e3 Ng6-f4 Be2-f1 Bg7-f6 Ra1-a3 Rf7-h7 Nd3xf4 e5xf4
20/34- 00:12 42,192,034 3,285,728 -0.60 Qd8-e8 Qb3-a4 Qe8xa4 Nc3xa4 h5-h4 b4-b5 h4-h3 b5-b6 h3xg2 Kg1xg2 Nf6-h7 Bg5-d2
20/34- 00:12 42,510,618 3,282,166 -0.68 Qd8-e8 Qb3-a4 Qe8xa4 Nc3xa4 h5-h4 b4-b5 h4-h3 b5-b6 h3xg2 Kg1xg2 Nf6-h7 Bg5-d2
20/34- 00:14 46,401,389 3,299,537 -0.80 Qd8-e8 Qb3-a4 Qe8xa4 Nc3xa4 h5-h4 b4-b5 h4-h3 b5-b6 h3xg2 Kg1xg2 Nf6-h7 Bg5-d2
20/34 00:15 52,638,688 3,307,696 -0.80 Qd8-e8 Qb3-a4 h5-h4 Qa4xe8 Nf6xe8 Rf1-c1 h4-h3 Kg1-h1 Kg8-f7 Be2-f1 Rf8-h8 Nc3-e2 Rh8-h5 Bg5-e3 Rh5-h7 Ne2xg3 d6xc5 g2xh3 Bc8xh3 Kh1-g1 Bh3xf1 Kg1xf1 c5xb4 Nd3xb4
21/34+ 00:20 67,638,463 3,342,316 -0.72 Qd8-e8 Qb3-a4 Bc8-d7 Qa4-a2 Nf6-h7 Qa2-d2 h5-h4 a5-a6 b7xa6 c5-c6 Bd7-c8 Bg5-e3 h4-h3
21/34+ 00:22 75,769,734 3,345,389 -0.64 Qd8-e8 Qb3-a4 Bc8-d7 Qa4-a2 Nf6-h7 Qa2-d2 h5-h4 a5-a6 b7xa6 c5xd6 c7xd6 Ra1xa6 Nh7xg5 Qd2xg5 Rf8-f6 Rf1-c1 h4-h3 g2xh3 Qe8-c8 Ra6-a1 g3-g2 Qg5xg2 Bd7xh3
21/34+ 00:24 83,163,740 3,360,965 -0.52 Qd8-e8 Qb3-a4 Bc8-d7 Qa4-a2 Nf6-h7 c5xd6 Nh7xg5 d6xc7 Ng6-h4 Nd3-c5 Bd7-h3 d5-d6+ Kg8-h7 Qa2-d2 Qe8-g6 g2xh3 Ng5xh3+ Kg1-h1 g3-g2+
21/34+ 00:29 98,004,629 3,372,608 -0.34 Qd8-e8 b4-b5 Nf6-h7 Bg5-e3 Qe8-e7 c5xd6 c7xd6 f3-f4 e5xf4 Nd3xf4 Qe7-h4
21/34 00:32 108,037,290 3,367,850 -0.44 Qd8-e8 Qb3-a4 Bc8-d7 Qa4-a2 Nf6-h7 Qa2-d2 h5-h4 a5-a6 b7xa6 c5xd6 c7xd6 Ra1xa6 Nh7xg5 Qd2xg5 Rf8-f6 Rf1-c1 h4-h3 g2xh3 Ng6-f4 Nd3xf4 e5xf4 Be2-f1 Rf6-g6 Qg5xf4 Bg7-d4+ Kg1-h1
22/34+ 00:35 120,920,205 3,365,625 -0.36 Qd8-e8 Qb3-a4 Bc8-d7 Qa4-a2 Nf6-h7 Qa2-d2 h5-h4 a5-a6 b7xa6 c5xd6 c7xd6 Ra1xa6 Nh7xg5 Qd2xg5 Rf8-f6 Rf1-c1 h4-h3 g2xh3 Ng6-f4 Nd3xf4 e5xf4 Be2-f1 Rf6-g6 Qg5xf4 Bg7-h6
22/35- 00:40 136,284,507 3,372,794 -0.52 Qd8-e8 Nc3-b5 Qe8xb5 c5-c6 Nf6xe4 f3xe4 Bc8-g4 Rf1xf8+ Ra8xf8 Be2-f1 Rf8xf1+ Ra1xf1 Bg4-e2 Rf1-e1 Be2xd3 c6xb7 Ng6-f4 Bg5xf4 e5xf4 Qb3xd3 Qb5xd3 b7-b8Q+ Kg8-h7 Qb8-e8
22/35- 00:40 136,864,875 3,372,052 -0.64 Qd8-e8 Nc3-b5 Qe8xb5 c5-c6 Nf6xe4 f3xe4 Bc8-g4 Rf1xf8+ Ra8xf8 Be2-f1 Rf8xf1+ Ra1xf1 Bg4-e2 Rf1-e1 Be2xd3 c6xb7 Ng6-f4 Bg5xf4 e5xf4 Qb3xd3 Qb5xd3 b7-b8Q+ Kg8-h7 Qb8-e8
22/35 00:41 140,265,641 3,372,662 -0.68 Qd8-e8 Nc3-b5 Qe8xb5 c5-c6 Nf6xe4 f3xe4 Bc8-g4 Rf1xf8+ Ng6xf8 Be2xg4 h5xg4 c6xb7 Qb5xb7 Qb3-d1 Qb7-b5 Ra1-c1 Qb5-d7 Qd1-e2 Nf8-h7 Bg5-h4 Bg7-h6 Rc1-c6 Qd7-g7 Bh4xg3 Nh7-f6 Bg3-h4 Ra8-f8
23/39+ 00:44 150,124,201 3,378,588 -0.60 Qd8-e8 Nc3-b5 Qe8xb5 c5-c6 Nf6xe4 f3xe4 Bc8-g4 Rf1xf8+ Ng6xf8 Be2xg4 h5xg4 c6xb7 Qb5xb7 Qb3-d1 Nf8-h7 Bg5-h4 a7-a6 Qd1xg4 Qb7-a7+ Kg1-h1 Qa7-d4 Ra1-d1 Ra8-f8 Bh4xg3 Nh7-f6
23/39+ 00:47 161,153,350 3,383,797 -0.52 Qd8-e8 Nc3-b5 Qe8xb5 c5-c6 Nf6xe4 f3xe4 Bc8-g4 Rf1xf8+ Ng6xf8 Be2xg4 h5xg4 c6xb7 Qb5xb7 Qb3-d1 Nf8-h7 Bg5-h4 a7-a6 Qd1xg4 Qb7-a7+ Kg1-h1 Qa7-d4 Ra1-d1 Ra8-f8 Bh4xg3 Nh7-f6
23/39+ 00:51 172,746,264 3,387,248 -0.40 Qd8-e8 Nc3-b1 Nf6-h7 Bg5-e3 Qe8-e7 c5xd6 c7xd6 Rf1-c1 Qe7-h4 Rc1-c7 Qh4-h2+ Kg1-f1 Qh2-h1+ Be3-g1 Ng6-h4
23/46 01:05 222,542,931 3,400,613 -0.44 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 Nf6-h7 Bg5-d2 h5-h4 Rf1-c1 Rf8-f7 Bd2-e3 h4-h3 g2xh3 Ng6-h4 Nd3-e1 Rf7-f6 c5xd6 c7xd6 Be2xa6 Bc8xh3 Ba6-b7 Ra8-f8 Ra4xa7 Nh4xf3+ Ne1xf3 Rf6xf3
24/46+ 01:18 267,870,042 3,393,466 -0.36 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 Rf1-c1 h4-h3 g2xh3 Nf6-h5 c5xd6 c7xd6 h3-h4 Bg7-f6 Bg5xf6 Rf8xf6
24/46+ 01:28 300,248,887 3,392,566 -0.28 Qd8-e8 f3-f4 e5xf4 Qb3-c4 d6xc5 b4xc5 Nf6-g4 Be2xg4 Bc8xg4 a5-a6 b7xa6 Nd3xf4 Qe8-e5 Nf4xg6 Rf8xf1+ Ra1xf1 Qe5xg5 d5-d6+ Kg8-h7 Ng6-f4 Qg5-h4
24/46 01:37 328,036,975 3,379,002 -0.36 Qd8-e8 Nc3-b5 Qe8xb5 c5-c6 Nf6xe4 f3xe4 Bc8-g4 Rf1xf8+ Ng6xf8 c6xb7 Qb5xb7 Be2xg4 h5xg4 Qb3-d1 Nf8-h7 Qd1xg4 Nh7xg5 Qg4xg5 c7-c6 d5xc6 Qb7xc6 Qg5-e3 Ra8-f8 Ra1-c1 Qc6-b7 Rc1-e1 Qb7-b5 Qe3xg3 a7-a6
25/46+ 01:47 360,817,742 3,361,008 -0.28 Qd8-e8 Nc3-b5 Qe8xb5 c5-c6 Nf6xe4 f3xe4 Bc8-g4 Rf1xf8+ Ra8xf8 Be2-f1 b7xc6 Nd3-f4 e5xf4 d5xc6+ Kg8-h7 Bf1xb5 Bg7xa1 Bb5-a4 f4-f3 Qb3-e3 Ng6-e5 Qe3xa7 f3-f2+ Kg1-f1 Bg4-e2+
25/46+ 02:03 412,706,705 3,354,929 -0.20 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 Rf1-a1 Ra8-b8 Bg5-e3 h4-h3 g2xh3 Nf6-h5 c5xd6 c7xd6 Be3xa7 Rb8-a8 Ba7-e3 Ng6-h4 Nd3-e1 Nh5-f4
25/46 02:21 473,730,169 3,342,058 -0.40 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 Rf1-a1 Ra8-b8 Ra4xa6 Bc8xa6 Ra1xa6 Nf6-h5 c5xd6 c7xd6 Ra6xd6 Nh5-f4 Nd3xf4 Ng6xf4 Bg5xh4 Nf4xe2+ Nc3xe2 a7-a5 Bh4xg3 a5xb4 Rd6-g6 Kg8-h7 Rg6xg7+ Kh7xg7 Bg3xe5+ Kg7-f7 Be5xb8
26/48- 02:27 492,034,805 3,344,263 -0.48 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 Rf1-a1 Ra8-b8 Ra4xa6 Bc8xa6
26/48+ 02:48 564,386,221 3,340,571 -0.32 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 Rf1-a1 Ra8-b8 Ra4xa6 Bc8xa6 Ra1xa6 Nf6-h5 c5xd6 c7xd6 Ra6xd6 Nh5-f4 Nd3xf4 Ng6xf4 Bg5xh4 Nf4xe2+
26/48- 03:19 676,114,883 3,387,145 -0.56 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 Rf1-a1 Ra8-b8 Ra4xa6 Bc8xa6 Ra1xa6 Nf6-h5 c5xd6 c7xd6 Ra6xd6
26/48+ 03:42 752,539,268 3,382,092 -0.20 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 Rf1-a1 Ra8-b8 Ra4xa6 Bc8xa6 Ra1xa6 Rf8-c8 Ra6xa7 Nf6-h5 Be2-f1 h4-h3 g2xh3 Nh5-f4 Nd3xf4 Ng6xf4 Bg5-e7 d6xc5 b4xc5 Rb8-b3
26/48 04:01 817,478,144 3,390,731 -0.44 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 Rf1-a1 Ra8-b8 Bg5-e3 h4-h3 c5xd6 c7xd6 Be3xa7 h3-h2+ Kg1-h1 Rb8-a8 Ba7-b6 Bc8-d7 Ra4xa6 Ra8xa6 Ra1xa6 Rf8-c8 Ra6-a3 Nf6-h5 Bb6-e3
27/48+ 04:34 931,763,346 3,391,103 -0.36 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 Rf1-a1 Ra8-b8 Bg5-e3 h4-h3 c5xd6 c7xd6 Be3xa7 Rb8-a8 Ba7-e3 Ra8-b8 g2xh3 Nf6-h5 Nd3-e1 Nh5-f4 Be2xa6 Ng6-h4 Ba6xc8 Rf8xc8 Be3xf4 e5xf4
27/48+ 05:07 1,040,760,056 3,387,703 -0.28 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 Rf1-a1 Ra8-b8 Ra4-a5 Rf8-f7 b4-b5 Nf6-h7 Bg5-e3 h4-h3 g2xh3 Ng6-f4
27/48+ 06:14 1,268,564,962 3,389,419 -0.16 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 c5xd6 c7xd6 Rf1-c1 h4-h3 g2xh3 Nf6-h5 Kg1-g2 Bg7-f6 Bg5xf6 Rf8xf6
27/48 08:06 1,665,309,791 3,421,382 -0.12 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 Rf1-a1 Ra8-b8 Ra4xa6 Bc8xa6 Ra1xa6 Rf8-c8 b4-b5 Nf6-h7 Bg5-d2 Ng6-f4 Nd3xf4 e5xf4 Nc3-a4 Bg7-d4+ Kg1-f1 Bd4-e3 Bd2xe3 f4xe3 Ra6xa7 Rb8-a8 Ra7-a6 Kg8-f7 Na4-b2 d6xc5 Nb2-c4 Nh7-g5 Nc4xe3 Rc8-b8 Ra6-c6
28/50- 08:19 1,712,074,572 3,426,664 -0.20 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 Rf1-a1 Ra8-b8 Bg5-e3
28/50 09:03 1,857,112,946 3,415,260 -0.16 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Qe8xa4 Ra1xa4 h5-h4 Rf1-a1 Ra8-b8 Bg5-e3 h4-h3 c5xd6 c7xd6 Be3xa7 Rb8-a8 Ba7-e3 Nf6-h5 b4-b5 Nh5-f4 Nd3xf4 Ng6xf4 Be2-f1 h3xg2 Bf1xg2 Bc8-h3 Be3xf4 e5xf4 Ra4xa6 Ra8xa6 Ra1xa6
29/50+ 13:03 2,672,317,645 3,410,025 -0.08 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Bc8-d7 c5-c6 Bd7-c8 Qa4-a5 Nf6-h7 Bg5-e3 Qe8-e7 Rf1-b1 Qe7-h4 Qa5xc7
29/50+ 16:05 3,252,662,444 3,369,300 0.00 Qd8-e8 a5-a6 b7xa6 Qb3-a4 Bc8-d7 c5-c6 Bd7-c8 Bg5xf6 Bg7xf6 Nd3-e1 Bf6-g5 Ne1-c2 h5-h4 Qa4-a5 h4-h3 Qa5xc7 h3xg2 Rf1-b1 Ng6-f4 Qc7xd6 Nf4xe2+ Nc3xe2
29/50 22:15 4,455,995,314 3,337,464 0.00 Qd8-e8 Nc3-b5 Qe8xb5 c5-c6 Nf6xe4 Nd3-c1 Qb5xe2 Nc1xe2 Ne4xg5 Ne2xg3 Ng6-f4 Qb3-c4 b7xc6 Qc4xc6 Ra8-b8 Qc6xc7 Rb8xb4 Ra1-b1 Rb4-d4 Rb1-b8 Rd4-d2 Ng3xh5 Nf4xh5 Rb8xc8 Nh5-f4 Rc8xf8+ Bg7xf8 Qc7xa7 Rd2xg2+ Kg1-h1 Rg2-g3 Qa7-e3 Rg3-h3+ Kh1-g1 Rh3-g3+ Kg1-h1

And I forgot to state that Stockfish is not running in default parameters. Even Stockfish 2.2.2 can find it. Provided you have to tweak the parameters as per the position to get it right. I have always done that to take advantage of Houdini most of the time. see My blog.

[Vinvin]

Houdini 3 finds THE score and PV.

The technical idea behind this is the difference in pruning and extensions between PV nodes and non-PV nodes.
If you find a fail-high for a non-PV node at depth 30, it's very likely that when it becomes a PV node the fail-high can be found at reduced depth.

Houdini 3 demonstrates how remarkably effective this idea is.

I think I understand the idea based on all the comments here, we'll probably try it in a day or two. You say that the idea is "remarkably effective", implying that it accounts for a significant part of the 55-60 elo gain in Houdini 3. Would you mind estimating how much of that gain you attribute to this idea? Also, if the gain is significant, why are you willing to share it with the competition? I'm not complaining, just curious.

As Uri wrote, the concept give probably very few rating points. But it's very efficient for study-like position when the key move is very deep and the engine have to resolve a fail-high for an unexplore part of the tree-search.

[mcostalba]

In most positions it is irrelevant(houdini does not need to reduce the depth) and even in positions when houdini reduce the depth the advantage is not a better move but using less time for the specific move or better information in the hash for the next move.

In game play this idea could bring something on the table in case of multiple best move changes, because if the best move changes only once it will be played as soon as allotted time expires, regardless if the fail-high has been resolved or not.

What makes me really sad is to see the developers of what is supposed to be the second strongest engine in the world shameless whining for some clue the man they have multiple times attacked. If they really believe that this idea brings big part of ELO gain of Houdini they are clearly completely blind to computer chess and their engine improvement is all due to a very good test framework feed by throwing at it more or less random /monkey style stuff until something good is detected.

On the other side we have an old glory, that keeps proudly insisting on not understanding and blames people willing to help him.

...I see I am more and more Vincent-izing myself, perhaps it is better to leave before to became a complete arrogant fool.

[Carotino]

It works! I'm using a similar method in the new Vitruvius (less sophisticated than the one implemented by Robert) and results are surprising. For example, in this position:

FEN: r3b1k1/1rq1bppp/2n1p3/ppPpP3/2pP1P2/P1P1BB2/3Q2PP/1R2NRK1 w - - 0 1

Vitruvius 1.16 (in "Deep-Analysis" mode) finds, surprisingly quickly, the best move: 1.f5!
This without going too deep... I believe that this move is very difficult to find, for any engine ... Except, of course, Vitruvius 1.16 in "Deep-Analysis" mode and Houdini-3 in Tactical Mode.

Another example:

FEN: r1bq1rk1/ppp3b1/3p1nn1/P1PPp1Bp/1P2P3/1QNN1Pp1/4B1P1/R4RK1 b - - 0 1

Vitruvius 1.16 "Deep-Analysis" finds the best move (1.Bh6!) in only 4 minuts, on my computer. Depth: 21 plyes. At the same time Stockfish is at depth=27 plyes... And has not yet found the solution!

I do not own Houdini-3 and then I could not do the tests, but I think between the various engines is the only one who can find, such as Vitruvius 1.16, the solution in reasonable time...

P.S. I'm sorry, but I don't know how you post pictures.

Roberto.

houdini3 tactical does not see f5 in the first diagram(note that I did not check that f5 is the best move)

[D]r3b1k1/1rq1bppp/2n1p3/ppPpP3/2pP1P2/P1P1BB2/3Q2PP/1R2NRK1 w
- - 0 1

The position of the diagram, is derived from a key game for the 24th ICCF World Championship:

[Event "WCCC27CT02(WS)"]
[Site "ICCF"]
[Date "2007.05.10"]
[Round "-"]
[White "Semrl, Marjan"]
[Black "Turkov, Vladimir Sergeevich"]
[Result "1-0"]
[WhiteElo "2539"]
[BlackElo "2521"]
[ECO "C11i"]

1.e4 e6 2.d4 d5 3.Nc3 Nf6 4.e5 Nfd7 5.f4 c5 6.Nf3 Nc6 7.Be3 Qb6 8.Na4 Qa5+ 9.c3
c4 10.b4 Qc7 11.Be2 Be7 12.a3 b5 13.Nc5 Nxc5 14.bxc5 a5 15.O-O Bd7 16.Qd2 O-O
17.Rab1 Rfb8 18.Ne1 Rb7 19.Bf3 Be8 20.f5 exf5 21.g3 Rd8 22.Ng2 Bf8 23.h4 h6 24.
h5 Ne7 25.Bf4 Qc8 26.Ne3 Kh8 27.Qg2 Ra7 28.g4 Bd7 29.g5 hxg5 30.Bxg5 f4 31.Bxe7
Bh3 32.Bg4 Bxg4 33.Qxg4 fxe3 34.Qxc8 Rxc8 35.Bxf8 Rxf8 36.Rxb5 f5 37.Kg2 f4 38.
Kf3 Ra6 39.Rb6 Rxb6 40.cxb6 Kh7 41.b7 g5 42.e6 1-0

GM Semrl told me that despite many hours of analysis, none of its engines found this move (20.f5!). Here is the output from Scid:

1 14 +0.26 1.Cc2 b4 2.axb4 axb4 3.Cxb4 Cxb4 4.Txb4 Txb4 5.cxb4 g6 6.Tb1 Db7 7.Db2 Db5 8.Af2 Ac6 9.Dc2 Ta3 (0.83)
15 +0.31 1.Cc2 h6 2.f5 exf5 3.Df2 Td8 4.Dg3 g5 5.Ah5 b4 6.axb4 axb4 7.Cxb4 Cxb4 8.Txb4 Txb4 9.cxb4 Db7 10.Ad2 Ta8 11.Txf5 Ta2 12.Dc3 (1.57)
16 +0.23 1.Cc2 h6 2.Ta1 Tab8 3.Tfb1 Dd7 4.De1 f5 5.Tb2 Af7 6.Af2 g5 7.De3 g4 8.Ae2 Af8 (3.42)
16 +0.21 1.Cc2 Dd7 2.Tb2 f6 3.exf6 Axf6 4.f5 exf5 5.Ag5 Axg5 6.Dxg5 h6 7.Df4 b4 8.axb4 axb4 9.Ce3 bxc3 10.Txb7 Dxb7 11.Axd5+ Af7 12.Dxf5 Axd5 13.Dxd5+ Rh8 14.Cxc4 c2 15.Cd6 (3.65)
17 +0.24 1.Cc2 Dd7 2.Tb2 f6 3.exf6 Axf6 4.f5 exf5 5.Ag5 Axg5 6.Dxg5 h6 7.Df4 b4 8.Ce3 Ce7 9.axb4 axb4 10.Txb4 Txb4 11.cxb4 Ta2 12.Te1 c3 13.Cxd5 Cxd5 14.Axd5+ Dxd5 15.Txe8+ Rf7 (4.15)
18 +0.26 1.Cc2 Dd7 2.De2 f6 3.exf6 Axf6 4.Af2 Ag6 5.Tfe1 Te8 6.Ah5 Axh5 7.Dxh5 Df7 8.Dxf7+ Rxf7 9.Te2 Tf8 10.Tbe1 Te7 11.Ce3 b4 (6.95)
19 +0.29 1.Cc2 Dd7 2.De2 f6 3.exf6 Axf6 4.Af2 Ag6 5.Tfe1 Te8 6.Ah5 Axh5 7.Dxh5 Df7 8.De2 Ad8 9.Ce3 Ce7 10.Cg4 Ac7 11.Dxe6 Axf4 (10.85)
20 +0.22 1.Cc2 Dd7 2.Af2 b4 3.axb4 axb4 4.cxb4 Tab8 5.Ag3 Dd8 6.Ae2 h6 7.f5 exf5 8.Txf5 Ag5 9.Dc3 Ad7 10.Tff1 Ae6 11.Ta1 Cxb4 12.Cxb4 Txb4 13.Ta7 Tb2 (14.94)
21 +0.24 1.Cc2 Dd7 2.Af2 f6 3.exf6 Axf6 4.De2 Ag6 5.Tfe1 Td8 6.Tb2 Te8 7.Ah5 Axh5 8.Dxh5 Df7 9.Dxf7+ Rxf7 10.Teb1 Teb8 11.h3 h5 12.Rh2 g6 13.Ce3 Ag7 14.Ah4 (37.19)
22 +0.24 1.Cc2 Dd7 2.Af2 f6 3.exf6 Axf6 4.De2 Ag6 5.Tfe1 Td8 6.Tb2 Te8 7.Ah5 Axh5 8.Dxh5 Df7 9.Dxf7+ Rxf7 10.Teb1 Teb8 11.h3 h5 12.Rh2 g6 13.Ce3 Ag7 14.Ah4 (46.93)
23 +0.32 1.Cc2 Dd7 2.Af2 f6 3.exf6 Axf6 4.De2 Ag6 5.Tfe1 Td8 6.Tb2 Te8 7.Ah5 Axh5 8.Dxh5 Df7 9.Dxf7+ Rxf7 10.Teb1 Teb8 11.h3 h5 12.Rh2 g6 13.Ce3 Ag7 14.Ah4 (166.64)
23 +0.32 1.Cc2 Dd7 2.g3 f6 3.exf6 Axf6 4.f5 exf5 5.Ag5 Axg5 6.Dxg5 Af7 7.Ag2 Ae6 8.Ce3 h6 9.Dh5 Ce7 10.Tfe1 Af7 11.De2 Te8 12.Dc2 Tc7 13.Df2 Tcc8 (169.25)
24 +0.26 1.Cc2 Dd7 2.g4 f6 3.exf6 Axf6 4.Dg2 Af7 5.g5 Ae7 6.Ag4 Ag6 7.Dh3 Af5 8.Axf5 exf5 9.Tfe1 b4 10.axb4 axb4 11.Cxb4 Cxb4 12.Txb4 Txb4 13.cxb4 Af8 14.b5 Dxb5 15.Dxf5 Dc6 16.Dc2 g6 17.f5 (219.23)
24 +0.31 1.f5 exf5 2.g3 Dd7 3.Cg2 Af8 4.Cf4 Ce7 5.Dg2 Td8 6.Tfe1 Dc8 7.Ac1 Dc6 8.h4 b4 9.axb4 axb4 10.Txb4 Txb4 11.cxb4 Da4 12.Cxd5 Cxd5 13.Axd5 Dxb4 (439.23)
25 +0.32 1.f5 exf5 2.g3 Dd7 3.Cg2 Af8 4.Cf4 Ce7 5.Dg2 Td8 6.g4 fxg4 7.Axg4 Dc6 8.Af3 h6 9.Tb2 Dc8 10.h4 Dc6 11.Af2 b4 12.axb4 axb4 (652.02)

[Master Om]

Stockfish is a class engine. No matter what other think i am a bit addicted to it. It has helped me a lot in my games. By tweaking I have many personalities that help me in my wild games. So thumbs up to you and your team for this great engine. And there is an old saying that When society/people start saying against you then become sure that you are in the right direction.


" An Intelligent monkey is that monkey that doesn't monkey around others monkey's monkeys."

[Master Om]

It works! I'm using a similar method in the new Vitruvius (less sophisticated than the one implemented by Robert) and results are surprising. For example, in this position:

FEN: r3b1k1/1rq1bppp/2n1p3/ppPpP3/2pP1P2/P1P1BB2/3Q2PP/1R2NRK1 w - - 0 1

Vitruvius 1.16 (in "Deep-Analysis" mode) finds, surprisingly quickly, the best move: 1.f5!
This without going too deep... I believe that this move is very difficult to find, for any engine ... Except, of course, Vitruvius 1.16 in "Deep-Analysis" mode and Houdini-3 in Tactical Mode.

Another example:

FEN: r1bq1rk1/ppp3b1/3p1nn1/P1PPp1Bp/1P2P3/1QNN1Pp1/4B1P1/R4RK1 b - - 0 1

Vitruvius 1.16 "Deep-Analysis" finds the best move (1.Bh6!) in only 4 minuts, on my computer. Depth: 21 plyes. At the same time Stockfish is at depth=27 plyes... And has not yet found the solution!

I do not own Houdini-3 and then I could not do the tests, but I think between the various engines is the only one who can find, such as Vitruvius 1.16, the solution in reasonable time...

P.S. I'm sorry, but I don't know how you post pictures.

Roberto.

houdini3 tactical does not see f5 in the first diagram(note that I did not check that f5 is the best move)

[D]r3b1k1/1rq1bppp/2n1p3/ppPpP3/2pP1P2/P1P1BB2/3Q2PP/1R2NRK1 w
- - 0 1

The position of the diagram, is derived from a key game for the 24th ICCF World Championship:

[Event "WCCC27CT02(WS)"]
[Site "ICCF"]
[Date "2007.05.10"]
[Round "-"]
[White "Semrl, Marjan"]
[Black "Turkov, Vladimir Sergeevich"]
[Result "1-0"]
[WhiteElo "2539"]
[BlackElo "2521"]
[ECO "C11i"]

1.e4 e6 2.d4 d5 3.Nc3 Nf6 4.e5 Nfd7 5.f4 c5 6.Nf3 Nc6 7.Be3 Qb6 8.Na4 Qa5+ 9.c3
c4 10.b4 Qc7 11.Be2 Be7 12.a3 b5 13.Nc5 Nxc5 14.bxc5 a5 15.O-O Bd7 16.Qd2 O-O
17.Rab1 Rfb8 18.Ne1 Rb7 19.Bf3 Be8 20.f5 exf5 21.g3 Rd8 22.Ng2 Bf8 23.h4 h6 24.
h5 Ne7 25.Bf4 Qc8 26.Ne3 Kh8 27.Qg2 Ra7 28.g4 Bd7 29.g5 hxg5 30.Bxg5 f4 31.Bxe7
Bh3 32.Bg4 Bxg4 33.Qxg4 fxe3 34.Qxc8 Rxc8 35.Bxf8 Rxf8 36.Rxb5 f5 37.Kg2 f4 38.
Kf3 Ra6 39.Rb6 Rxb6 40.cxb6 Kh7 41.b7 g5 42.e6 1-0

GM Semrl told me that despite many hours of analysis, none of its engines found this move (20.f5!). Here is the output from Scid:

1 14 +0.26 1.Cc2 b4 2.axb4 axb4 3.Cxb4 Cxb4 4.Txb4 Txb4 5.cxb4 g6 6.Tb1 Db7 7.Db2 Db5 8.Af2 Ac6 9.Dc2 Ta3 (0.83)
15 +0.31 1.Cc2 h6 2.f5 exf5 3.Df2 Td8 4.Dg3 g5 5.Ah5 b4 6.axb4 axb4 7.Cxb4 Cxb4 8.Txb4 Txb4 9.cxb4 Db7 10.Ad2 Ta8 11.Txf5 Ta2 12.Dc3 (1.57)
16 +0.23 1.Cc2 h6 2.Ta1 Tab8 3.Tfb1 Dd7 4.De1 f5 5.Tb2 Af7 6.Af2 g5 7.De3 g4 8.Ae2 Af8 (3.42)
16 +0.21 1.Cc2 Dd7 2.Tb2 f6 3.exf6 Axf6 4.f5 exf5 5.Ag5 Axg5 6.Dxg5 h6 7.Df4 b4 8.axb4 axb4 9.Ce3 bxc3 10.Txb7 Dxb7 11.Axd5+ Af7 12.Dxf5 Axd5 13.Dxd5+ Rh8 14.Cxc4 c2 15.Cd6 (3.65)
17 +0.24 1.Cc2 Dd7 2.Tb2 f6 3.exf6 Axf6 4.f5 exf5 5.Ag5 Axg5 6.Dxg5 h6 7.Df4 b4 8.Ce3 Ce7 9.axb4 axb4 10.Txb4 Txb4 11.cxb4 Ta2 12.Te1 c3 13.Cxd5 Cxd5 14.Axd5+ Dxd5 15.Txe8+ Rf7 (4.15)
18 +0.26 1.Cc2 Dd7 2.De2 f6 3.exf6 Axf6 4.Af2 Ag6 5.Tfe1 Te8 6.Ah5 Axh5 7.Dxh5 Df7 8.Dxf7+ Rxf7 9.Te2 Tf8 10.Tbe1 Te7 11.Ce3 b4 (6.95)
19 +0.29 1.Cc2 Dd7 2.De2 f6 3.exf6 Axf6 4.Af2 Ag6 5.Tfe1 Te8 6.Ah5 Axh5 7.Dxh5 Df7 8.De2 Ad8 9.Ce3 Ce7 10.Cg4 Ac7 11.Dxe6 Axf4 (10.85)
20 +0.22 1.Cc2 Dd7 2.Af2 b4 3.axb4 axb4 4.cxb4 Tab8 5.Ag3 Dd8 6.Ae2 h6 7.f5 exf5 8.Txf5 Ag5 9.Dc3 Ad7 10.Tff1 Ae6 11.Ta1 Cxb4 12.Cxb4 Txb4 13.Ta7 Tb2 (14.94)
21 +0.24 1.Cc2 Dd7 2.Af2 f6 3.exf6 Axf6 4.De2 Ag6 5.Tfe1 Td8 6.Tb2 Te8 7.Ah5 Axh5 8.Dxh5 Df7 9.Dxf7+ Rxf7 10.Teb1 Teb8 11.h3 h5 12.Rh2 g6 13.Ce3 Ag7 14.Ah4 (37.19)
22 +0.24 1.Cc2 Dd7 2.Af2 f6 3.exf6 Axf6 4.De2 Ag6 5.Tfe1 Td8 6.Tb2 Te8 7.Ah5 Axh5 8.Dxh5 Df7 9.Dxf7+ Rxf7 10.Teb1 Teb8 11.h3 h5 12.Rh2 g6 13.Ce3 Ag7 14.Ah4 (46.93)
23 +0.32 1.Cc2 Dd7 2.Af2 f6 3.exf6 Axf6 4.De2 Ag6 5.Tfe1 Td8 6.Tb2 Te8 7.Ah5 Axh5 8.Dxh5 Df7 9.Dxf7+ Rxf7 10.Teb1 Teb8 11.h3 h5 12.Rh2 g6 13.Ce3 Ag7 14.Ah4 (166.64)
23 +0.32 1.Cc2 Dd7 2.g3 f6 3.exf6 Axf6 4.f5 exf5 5.Ag5 Axg5 6.Dxg5 Af7 7.Ag2 Ae6 8.Ce3 h6 9.Dh5 Ce7 10.Tfe1 Af7 11.De2 Te8 12.Dc2 Tc7 13.Df2 Tcc8 (169.25)
24 +0.26 1.Cc2 Dd7 2.g4 f6 3.exf6 Axf6 4.Dg2 Af7 5.g5 Ae7 6.Ag4 Ag6 7.Dh3 Af5 8.Axf5 exf5 9.Tfe1 b4 10.axb4 axb4 11.Cxb4 Cxb4 12.Txb4 Txb4 13.cxb4 Af8 14.b5 Dxb5 15.Dxf5 Dc6 16.Dc2 g6 17.f5 (219.23)
24 +0.31 1.f5 exf5 2.g3 Dd7 3.Cg2 Af8 4.Cf4 Ce7 5.Dg2 Td8 6.Tfe1 Dc8 7.Ac1 Dc6 8.h4 b4 9.axb4 axb4 10.Txb4 Txb4 11.cxb4 Da4 12.Cxd5 Cxd5 13.Axd5 Dxb4 (439.23)
25 +0.32 1.f5 exf5 2.g3 Dd7 3.Cg2 Af8 4.Cf4 Ce7 5.Dg2 Td8 6.g4 fxg4 7.Axg4 Dc6 8.Af3 h6 9.Tb2 Dc8 10.h4 Dc6 11.Af2 b4 12.axb4 axb4 (652.02)

In that case he has no Idea what to use where!!