Gata Kamsky wins the fourth game!

[Eelco de Groot]

The match is an open battle again with four more games to go. Tomorrow is a restday I presume?


[Event "World Chess Challenge"]
[Site "Sofia BUL"]
[Date "2009.02.21"]
[Round "4"]
[White "Kamsky,G"]
[Black "Topalov,V"]
[Result "1-0"]
[WhiteElo "2725"]
[BlackElo "2796"]
[EventDate "2009.02.17"]
[ECO "C92"]

1. e4 e5 2. Nf3 Nc6 3. Bb5 a6 4. Ba4 Nf6 5. O-O Be7 6. Re1 b5 7. Bb3 d6 8.
c3 O-O 9. h3 Bb7 10. d4 Re8 11. Nbd2 Bf8 12. Ng5 Re7 13. d5 Nb8 14. Nf1
Nbd7 15. Ng3 g6 16. Bc2 h6 17. Nf3 Nb6 18. h4 Qd7 19. Nh2 Bg7 20. h5 Rf8
21. Nhf1 c6 22. dxc6 Qxc6 23. Ne3 Kh7 24. Qf3 Bc8 25. Rd1 Be6 26. b3 Qxc3
27. Bd2 Qc7 28. Ba5 Qb8 29. Rd2 Nc8 30. Rad1 b4 31. Qe2 Kh8 32. Bd3 Na7 33.
Rc1 Nb5 34. Bxb5 axb5 35. Bxb4 Rd7 36. Rc6 Rfd8 37. Qd1 Bf8 38. Qc2 Kh7 39.
Ba5 Re8 40. hxg6+ fxg6 41. Bc7 Qb7 42. Bxd6 Bf7 43. Bb4 Bxb4 44. Rxd7 Qxd7
45. Rxf6 Re6 46. Nd5 Bf8 47. Rf3 Kg7 48. Rc3 Ra6 49. Rc7 Qd6 50. Qe2 Kg8
51. Qxb5 Rxa2 52. Qb7 Ra1+ 53. Kh2 Bxd5 54. exd5 Qf6 55. Qc8 Qh4+ 56. Qh3
Qxh3+ 57. Kxh3 Rd1 58. Ne4 Ba3 59. Ra7 Bb4 60. Rb7 Ba3 61. f3 Kf8 62. Rb5
h5 63. Kg3 Rc1 64. Rb8+ Kf7 65. Rb7+ Kf8 66. Kf2 Rc2+ 67. Kf1 Rc1+ 68. Ke2
Rc2+ 69. Kd3 Rxg2 70. Ra7 Be7 71. d6 Bd8 72. Nc5 Ke8 73. Rh7 1-0

My computer is busy at the moment doing extended analysis of EET position No. 14 from Walter Eigenmann, and I have not replayed the game so no analysis from me. I have a gut feeling it is an impressive win by Kamsky though! Mihail Marin will analyze the game on Chessbase tomorrow I think.

Eelco

[Marc Lacrosse]

The match is an open battle again with four more games to go.

Indeed.

Tomorrow is a restday I presume?

Yes.

My computer is busy at the moment doing extended analysis of EET position No. 14 from Walter Eigenmann, and I have not replayed the game so no analysis from me. I have a gut feeling it is an impressive win by Kamsky.

It is (although quite a few more rapid wins through devishly complicated tactical paths were wisely avoided by Kamsky).

But let me know :
I just do not see how and where your post is in any aspect relevant for a computer chess forum.

Marc

[Matthias Gemuh]

But let me know :
I just do not see how and where your post is in any aspect relevant for a computer chess forum.

Marc

It is such games that we seek to analyze with computers and chess engines, hoping to find good positions for engine testsuites.

Matthias.

[Eelco de Groot]

The match is an open battle again with four more games to go.

Indeed.

Tomorrow is a restday I presume?

Yes.

My computer is busy at the moment doing extended analysis of EET position No. 14 from Walter Eigenmann, and I have not replayed the game so no analysis from me. I have a gut feeling it is an impressive win by Kamsky.

It is (although quite a few more rapid wins through devishly complicated tactical paths were wisely avoided by Kamsky).

But let me know :
I just do not see how and where your post is in any aspect relevant for a computer chess forum.

Marc

Thanks for the game comments Marc! Excellent, and I am looking forward to replaying the game or reading the GM analysis from Mihail Marin or others tomorrow.

I think these high level games are very relevant for computer chess. To suggest they are not, it would be an affront to the dedication and preparation shown by Gata Kamsky and Veselin Topalov. With all the computer preparation by top players chess and computer chess only get more interwoven.

But I understand your complaint that I did not add much computer oriented stuff to my post. I explained that the computer was busy with a Toga test of EET No. 14.

The position No. 14 is more or less chosen at random but to illustrate it can be tricky how long exactly you let the program continue calculating. If the variations differ too much in length at any depth, you cannot compare the leafs anymore. But if you stop too soon there are problems with quiescence.

Added problem are extensions because in the most basic form these are unchecked, every extension grafts a subtree that is really one ply too deep and because the extensions can be invoked recursively, you quickly can run into exploding trees.

First version of analysis by build 356.

[D]4qr2/4p2k/1p2P1pP/5P2/1P3P2/6Q1/8/3B1K2 w - - bm Ba4; id "E_E_T 014 - D&L vs D&T";

4qr2/4p2k/1p2P1pP/5P2/1P3P2/6Q1/8/3B1K2 w - -

Engine: Blueberry Beta 4 DM70 Build 356 (128 MB)
by F. Letouzey, T. Gaksch, E. de Groot


1/27 0:00 +1.62 1.fxg6+ Qxg6 2.Bc2 Rxf4+ 3.Qxf4 Qxc2
4.Qf7+ Kxh6 5.Qxe7 Qd1+ 6.Kf2 Qd4+
7.Kf3 Qd3+ 8.Kf4 Qd4+ 9.Kf5 Qd5+
10.Kg4 Qe4+ 11.Kg3 Qd3+ 12.Kg2 Qe4+
13.Kf2 (326.048)

2/27 0:00 +1.65 1.fxg6+ Qxg6 2.Bc2 Rxf4+ 3.Qxf4 Qxc2
4.Qf7+ Kxh6 5.Qxe7 Qd1+ 6.Kf2 Qc2+
7.Ke3 Qc3+ 8.Ke4 Qc4+ 9.Kf5 Qd3+
10.Kf4 Qd4+ 11.Kf3 (672.930)

2/32 0:00 +2.21 1.Bc2 Kxh6 2.fxg6 Qb5+ 3.Kf2 Kg7
4.Qh4 Qxb4 5.Qh7+ Kf6 6.Qh6 Qd4+
7.Kf3 (830.358)

2/33 0:01 +2.21 1.Bc2 Kxh6 2.fxg6 Qb5+ 3.Kf2 Qd5
4.g7 Qd4+ 5.Ke1 Qxb4+ 6.Kd1 Qd4+ (920.407) 920

2/44 0:01 +2.21 1.Bc2 Kxh6 2.Kg1 gxf5 3.Qh3+ Kg7 (1.459.487) 920

2/44 0:01 +2.51 1.Ba4 b5 2.Bc2 Kxh6 3.fxg6 Kg7 4.Be4 (1.517.672) 920

2/44 0:01 +2.50 1.Ba4 b5 2.Bc2 Kxh6 3.fxg6 Kg7 4.Kg1 Rf6 (1.614.274) 920

2/44 0:02 +4.70 1.Ba4 b5 2.fxg6+ Kg8 3.Bc2 Qc8 4.Bd3 Qc1+
5.Kg2 Qd2+ 6.Kh1 Qc1+ 7.Kh2 Qd2+
8.Kh3 Rxf4 9.Bxb5 Qxb4 (1.664.569) 810

3/44 0:02 +1.16 1.Ba4 b5 2.Bc2 (1.670.564) 798

3/52 0:08 +1.32 1.Bc2 Kxh6 2.Bd3 gxf5 3.Qg5+ Kh7
4.Bxf5+ Rxf5 5.Qxf5+ Kg7 6.Kf2 Kh6
7.Qf7 (6.848.375) 842

3/52 0:09 +1.90 1.Bc2 Kxh6 2.fxg6 Kg7 3.Qc3+ Rf6
4.Kf2 Qh8 5.f5 Qh4+ 6.Kg2 Qg4+ 7.Kh2 Qh4+
8.Kg1 Qg4+ 9.Kf1 (7.819.146) 831

3/52 0:11 +1.90 1.Bc2 Kxh6 2.fxg6 (9.418.362) 807

3/52 0:12 +2.74 1.fxg6+ Qxg6 2.Bc2 Rxf4+ 3.Kg1 Rf6
4.Bd3 Rxe6 5.Kg2 Re2+ 6.Bxe2 Qxg3+
7.Kxg3 Kxh6 8.Kf4 Kg6 9.Bc4 Kf6 (9.973.408) 802

3/52 0:12 +2.70 1.fxg6+ Qxg6 2.Bc2 Rxf4+ 3.Qxf4 Qxc2
4.Qg5 Qc4+ 5.Ke1 Qxb4+ 6.Ke2 Qd4
7.Qxe7+ Kxh6 8.Qf8+ Kg5 9.Qg8+ Kf4 (10.249.849) 802

3/52 0:14 +2.70 1.fxg6+ Qxg6 2.Bc2 Rxf4+ 3.Qxf4 Qxc2
4.Qg5 Qc4+ 5.Ke1 Qxb4+ 6.Ke2 Qd4
7.Qxe7+ Kxh6 8.Qf8+ Kg5 9.Qg8+ Kf4 (11.639.848) 796

4/52 0:16 +4.71 1.fxg6+ Qxg6 2.Bc2 Rxf4+ 3.Kg1 Rf6
4.Bd3 Rxe6 5.Kh2 Rd6 6.Bxg6+ Rxg6
7.Qd3 Kxh6 8.Qe3+ Kg7 9.Qxe7+ Kh6
10.Qh4+ Kg7 11.Qd4+ Rf6 12.Qg4+ Kf7
13.Qd7+ Kg6 14.Qe8+ (13.249.169) 813

5/52 0:18 +4.17 1.fxg6+ Qxg6 2.Bc2 Rxf4+ 3.Kg1 Rf6
4.Bd3 Rxe6 5.Kh2 Re2+ 6.Kh3 Re6
7.Bxg6+ Rxg6 8.Qd3 Kxh6 9.Qd4 Kh7
10.Qh4+ Rh6 (15.053.031) 807

6/52 0:19 +1.62 1.fxg6+ Qxg6 2.Bc2 Rxf4+ 3.Qxf4 Qxc2
4.Qf7+ Kxh6 5.Qxe7 Qd1+ 6.Kf2 Qd2+
7.Kf3 Qd3+ 8.Kf4 Qd4+ 9.Kf5 Qd5+
10.Kg4 Qe4+ 11.Kg3 Qd3+ 12.Kg2 Qe4+
13.Kf2 (16.300.847) 817

6/53 0:48 +1.98 1.Ba4 b5 2.Bc2 Kxh6 3.fxg6 Kg7
4.Qc3+ Rf6 5.Kg1 Qb8 6.f5 Qf4 7.Kg2 Qg4+
8.Kh1 Qh4+ 9.Kg1 (38.710.658) 796

6/53 0:50 +1.98 1.Ba4 b5 2.Bc2 Kxh6 3.fxg6 Kg7 4.Bd3 Qc6
5.f5 (39.784.285) 794

6/53 0:51 +1.98 1.Ba4 b5 2.Bc2 Kxh6 3.fxg6 (40.878.618) 793

7/53 0:53 +2.10 1.Ba4 b5 2.Bc2 Kxh6 3.fxg6 Kg7
4.Qc3+ Rf6 5.Kg2 Qa8+ 6.Kg3 (41.985.312) 791

8/53 1:11 +2.52 1.Ba4 b5 2.Bc2 Kxh6 3.fxg6 Kg7 4.Qg5 Rf6
5.Qh4 Qg8 (57.390.767) 799

9/57 4:23 +1.68 1.Ba4 b5 2.Bc2 Kxh6 3.fxg6 Kg7 4.Qg5 Qd8
5.Qe5+ Rf6 6.f5 Qd2 7.Qe4 Qh6 8.Kg2 Qg5+
9.Kh3 Qh5+ 10.Kg3 Qg5+ 11.Kh2 Qh5+
12.Kg1 Qg5+ 13.Kf2 Qd2+ 14.Kf3 (198.335.743) 751

10/57 9:19 +1.70 1.Ba4 b5 2.Bc2 Kxh6 3.fxg6 Kg7 4.Qg5 Qd8
5.Qe5+ Rf6 6.f5 Qd2 7.Qe4 Qh6 8.Kg2 Kh8
9.Bd3 (449.128.616) 803

11/62 12:31 +1.70 1.Ba4 b5 2.Bc2 Kxh6 3.fxg6 Kg7 4.Qg5 Qd8
5.Qe5+ Rf6 6.f5 Qd2 7.Qe4 Qh6 8.Kg2 Kh8
9.Bd3 (606.826.636) 806

best move: Bd1-a4 time: 15:31.078 min n/s: 853.183 CPU 100.0% n/s(1CPU): 853.183 nodes: 794.300.000

Ba4 is found, and then held after 48 seconds which is not so bad but the goal was to find it within at most twenty seconds. The problem is generated really by the variation in red; it is a three ply search but because the PV is actually only three plies deep, the leaf-evaluation (+ quiescence search which is not shown I think) is comparatively too low and the stability is lost. There are no "only moves" in the three plies deep PV so no singular extensions, no checks or captures which would have generated other extensions. The other moves in the list are searched differently and because of greater depth the move 1.Bc2 takes over as best move. At three plies this normally would not be such a problem but because of IID and 1.fxg6 taking over after that with again three consecututive IID searches, I am already down 14 seconds after a three ply search

First try of build 357a did not generate any output because of runaway extensions, not even a one ply searchresult. Deleted that build.

So I tried something else and it helped a bit, although the evals for 1.Ba4 now suddenly get suspiciously high. Also I was not thrilled with the late finding of second move 2.Bc2 but I now see that at a much greater depth 2.fxg6 is chosen again. That is a bit reassuring.

The five hour fifteen ply analysis now runs:

[D]4qr2/4p2k/1p2P1pP/5P2/1P3P2/6Q1/8/3B1K2 w - -

Engine: Blueberry Beta 4 DM70 Build 357 (Athlon 2009 MHz, 128 MB)
by F. Letouzey, T. Gaksch, E. de Groot

1/27 0:00 +1.64 1.fxg6+ Qxg6 2.Bc2 Rxf4+ 3.Qxf4 Qxc2
4.Qf7+ Kxh6 5.Qxe7 Qb1+ 6.Ke2 Qe4+
7.Kd2 Qd4+ 8.Kc2 Qe4+ 9.Kc3 Qe5+
10.Kd3 (186.802)

2/27 0:00 +1.64 1.fxg6+ Qxg6 2.Bc2 Rxf4+ 3.Qxf4 Qxc2
4.Qf7+ Kxh6 5.Qxe7 Qb1+ 6.Ke2 Qe4+
7.Kd2 Qd4+ 8.Kc2 Qe4+ 9.Kc3 Qe5+
10.Kd3 (357.792)

2/50 0:02 +2.14 1.Bc2 Kxh6 2.fxg6 Qb5+ 3.Kf2 Kg7
4.Qc3+ Rf6 5.f5 Qd5 6.Kg1 Kg8 7.Qg3 Qd4+
8.Kh1 Qxb4 9.Qh3 (2.162.732) 765

2/50 0:03 +2.14 1.Bc2 Kxh6 2.Qh4+ Kg7 3.fxg6 Qb5+
4.Kf2 Qxb4 5.Qh7+ Kf6 6.g7 (2.251.264) 740

2/50 0:03 +2.14 1.Bc2 Kxh6 2.fxg6 Qb5+ 3.Kf2 Kg7
4.Qc3+ Kh6 5.Qe1 Rxf4+ 6.Kg1 Rxb4
7.Qd2+ Kh5 8.Qh2+ Kg4 (2.303.979) 740

2/52 0:07 +2.33 1.Ba4 b5 2.Bc2 Kxh6 3.Qg5+ Kg7
4.fxg6 Rf6 5.f5 Qc6 6.Qh4 (6.269.396) 802

2/52 0:11 +4.87 1.Ba4 b5 2.fxg6+ Kg8 3.Bc2 Qb8 4.Kg2 Rxf4
5.Qh3 Qa8+ 6.Kg1 Rf6 7.Bd3 Qa1+
8.Bf1 Rxf1+ 9.Qxf1 Qd4+ 10.Kh2 (9.632.976) 818

2/54 0:16 +4.86 1.Ba4 b5 2.fxg6+ Kg8 3.Bd1 Rf6 4.Qh3 Rxf4+
5.Bf3 (13.343.662) 818

3/54 0:17 +5.46 1.Ba4 b5 2.fxg6+ Kxh6 3.Qh4+ Kg7
4.Qh7+ Kf6 5.g7 Kxe6 6.gxf8Q Qxf8
7.Qe4+ Kf7 8.Bxb5 Qh8 (14.160.833) 812

4/55 0:54 +3.74 1.Ba4 b5 2.fxg6+ Kg8 3.Bd1 Qc6 4.Bf3 Qc4+
5.Kg2 Rxf4 6.Qg5 Qc2+ 7.Kg3 Rxf3+
8.Kxf3 Qc6+ 9.Ke2 Qc4+ 10.Ke1 (41.768.139) 773

5/55 1:23 +4.31 1.Ba4 b5 2.fxg6+ Kg8 3.Bd1 Qc6 4.Bf3 Qxe6
5.Qg5 Rd8 6.Qg4 (63.797.305) 763

6/55 1:27 +4.31 1.Ba4 b5 2.fxg6+ Kg8 3.Bd1 Qc6 4.Bf3 Qxe6
5.Qg5 Qc4+ 6.Kg2 (67.094.897) 763

7/55 2:01 +4.39 1.Ba4 b5 2.fxg6+ Kg8 3.Bd1 Qc6 4.Qg5 Qc4+
5.Kg2 Qxe6 6.Kh2 Rf6 7.Qc5 Rxf4
8.Qxb5 Qxg6 9.Qb8+ Rf8 10.Bb3+ Kh7
11.Qxf8 Qxh6+ 12.Qxh6+ Kxh6 (93.869.320) 773

8/55 2:51 +4.61 1.Ba4 b5 2.fxg6+ Qxg6 3.Bc2 Rxf4+
4.Kg1 Rf6 5.Qg2 Rxe6 6.Kh2 Qxc2
7.Qxc2+ Kxh6 8.Qd2+ Kg7 9.Qg5+ Kf8
10.Qxb5 Rh6+ 11.Kg3 Rg6+ (130.471.488) 762

9/55 2:51 +4.61 1.Ba4 b5 2.fxg6+ Qxg6 3.Bc2 Rxf4+
4.Kg1 Rf6 5.Qg2 Rxe6 6.Kh2 Qxc2
7.Qxc2+ Kxh6 8.Qd2+ Kg7 9.Qg5+ Kf7
10.Qxb5 Rh6+ 11.Kg3 Rg6+ (131.169.188) 762

10/56 4:29 +1.70 1.Ba4 b5 2.Bc2 Kxh6 3.Qg5+ Kg7
4.fxg6 Qd8 5.Qe5+ Kg8 6.g7 Qd2
7.gxf8Q+ Kxf8 8.Qxb5 Qxf4+ 9.Ke2 Qg4+
10.Kf2 Qh4+ 11.Ke3 Qg3+ 12.Kd2 Qg2+
13.Ke1 Qxc2 14.Qf1+ (204.719.035) 760

11/61 16:13 +1.71 1.Ba4 b5 2.Bc2 Kxh6 3.Qg5+ Kg7
4.fxg6 Qd8 5.Qe5+ Kg8 6.g7 Re8
7.Qxb5 Kxg7 8.Ke2 Kf6 9.Ke3 Kxe6
10.Bb3+ Kd6 11.Qe5+ Kc6 12.Ba4+ Kb6
13.Bxe8 (776.214.703) 797

12/61 36:49 +1.71 1.Ba4 b5 2.Bc2 Kxh6 3.Qg5+ Kg7
4.fxg6 Qd8 5.Qe5+ Kg8 6.g7 Re8 (1.971.860.490) 892

13/63 84:59 +1.63 1.Ba4 b5 2.Bc2 Kxh6 3.Qg5+ Kg7
4.fxg6 Qd8 5.Qe5+ Rf6 6.f5 Qd2 7.Qe4 Kg8
8.Kg1 Qg5+ 9.Kh2 Qh6+ 10.Kg3 Qg5+
11.Kh3 Qh5+ 12.Qh4 Qxh4+ 13.Kxh4 Rf8
14.Bd3 (4.450.863.370) 872

14/65 285:51 +1.68 1.Ba4 b5 2.Bc2 Kxh6 3.fxg6 Kg7
4.Qc3+ Rf6 5.Ke1 Qa8 6.f5 Qg2 7.Bd3 Qa8
8.Qe5 Kh8 9.Bxb5 Qh1+ 10.Ke2 Qa8
11.Kd3 Qa3+ 12.Ke2 Qa2+ 13.Kd1 Qb1+
14.Kd2 (16.289.428.877) 949

15/67 379:10 +1.71 1.Ba4 b5 2.fxg6+ Kg8 3.g7 bxa4
4.gxf8Q+ Kxf8 5.Qc3 Qh5 6.Qg7+ Ke8
7.Kf2 Qh2+ 8.Ke3 Qh3+ 9.Kd2 Qh5
10.h7 Qd5+ 11.Kc2 Qc6+ 12.Kd3 Qf3+
13.Kd4 Qxf4+ 14.Kd5 (21.755.979.550) 956


Also I like the 19 ply deep PV generated by a 1 ply search Ba4 is stable although the searches for 1.Bc2 and 1.fxg6 presumably get evalutions that are very near 1.63 but I have not tested that.

Regards,
Eelco

[Norm Pollock]

...

But let me know :
I just do not see how and where your post is in any aspect relevant for a computer chess forum.

Marc

Marc,

Are you kidding?

Human chess and Computer chess are still part of the superset of chess, with the only differences being the nature of the players and the way they search for the best move. The goal in both cases is to make the best move each time, and therefore it is interesting to compare Human with Computer.

Maybe we should change Computer chess to make it "Different" and "Unique" from human chess. One thing we could do differently is number the moves in hexadecimal instead of decimal! [/sarcasm]

-Norm

[Eelco de Groot]

Sorry for the off topic content, but to complete analysis of EET 14 the computer was kept running so it can add two more plydepths after more than a twelvehundred minutes (with only 128 Mb transposition tables, this slows things down because of overwrites), 2.fxg6 is still the second move and the evaluation has gone up.


15/67 379:10 +1.71 1.Ba4 b5 2.fxg6+ Kg8 3.g7 bxa4
4.gxf8Q+ Kxf8 5.Qc3 Qh5 6.Qg7+ Ke8
7.Kf2 Qh2+ 8.Ke3 Qh3+ 9.Kd2 Qh5
10.h7 Qd5+ 11.Kc2 Qc6+ 12.Kd3 Qf3+
13.Kd4 Qxf4+ 14.Kd5 (21.755.979.550) 956

16/72 1023:20+7.14 1.Ba4 b5 2.fxg6+ Kg8 3.g7 Rf5 4.h7+ Kxh7
5.g8Q+ Qxg8 6.Qh4+ Kg6 7.Qg4+ Kf6
8.Qxg8 Rxf4+ 9.Ke1 bxa4 10.Qf8+ Ke5
11.Qxe7 Re4+ 12.Kf1 Rf4+ 13.Ke2 Re4+
14.Kf2 (58.497.901.011) 952

17/74 1239:14+7.10 1.Ba4 b5 2.fxg6+ Kg8 3.g7 Rf5 (71.338.254.867) 959


I think only after the very deep search with 2.fxg6 16 ply onwards we follow Walter Eigenmann's main line from the database, his .pgn is reproduced below, so clearly the search is not perfect if it takes us 1023 minutes to get there. Hmm, why does this happen? Blueberry likes very much its passed pawns on the sixth or seventh rank and will do anything just to keep them, that is one problem I think.. But at least the line is still improving and the search does not crash, and 2.fxg6 was also played already at plydepth = 2.

[Event "Pervakov 1997"]
[Site "?"]
[Date "????.??.??"]
[Round "?"]
[White "E_E_T 014"]
[Black "D&L vs D&T"]
[Result "*"]
[Annotator "W.Eigenmann"]
[SetUp "1"]
[FEN "4qr2/4p2k/1p2P1pP/5P2/1P3P2/6Q1/8/3B1K2 w - - 0 1"]
[PlyCount "9"]
[Source "W.Eigenmann"]
[SourceDate "2006.12.12"]

1. Ba4 (1. fxg6+ $2 Qxg6 2. Bc2 Rxf4+ 3. Kg1 Rf6 $11) (1. Bc2 $2 Kxh6 2. Ke1 (
2. fxg6 Qb5+ $11) 2... Kg7 $11) 1... b5 2. fxg6+ Kg8 (2... Qxg6 3. Bc2 Rxf4+ 4.
Kg1 Rf6 5. Qg5 Rxe6 (5... Kh8 6. Qxg6 Rxg6+ 7. Bxg6 $18) 6. Kf1 Rf6+ 7. Ke1
Re6+ 8. Kd1 Rd6+ 9. Kc1 Rc6 10. Qc5 $18) 3. g7 Rf5 4. h7+ Kxh7 5. Bc2 $18 *

Eelco

[Marc Lacrosse]

So ANY topic that is in any way chess-related has its place here?
But then why is this place the "Computer Chess Club" and not "The Chess Club"?
I have myself many different interests in chess : correspondence play, opening analysis, history, old pieces and boards, bibliophily and so on.
Should I freely post here on all these topics?
This is nonsense.
I followed online and greatly appreciated the game that Elco reported later.
If all members here create a new thread anytime they see, follow or read a nice game of chess that has _not_ been played by a computer or program we will have a lively forum for sure but we could well drop "computer" from our forum's name.

Marc