Too Eager for a Draw?

[Scott Tortorice]

One of the things I love about Chess is that even after the guns fall silent, there is still work to be done. I enjoy feeding my completed games into a chess engine or two for some interesting analysis, something unique to the wonderful, AI-heavy world of Chess.

The following was a tough game I had against a guy named “roo62.” When we started playing, his rating was roughly the same as mine; however during the course of the game, it eventually shot to over 2200 at one point! After fighting for my life and refuting his attack on my queen side of the board, I offered a draw as, even though I was up by a pawn, I had no confidence that I continue to fight for a win. I hate offering a draw as I am always then plagued by “what if” thoughts afterwards. Would I have won? Would I have lost? Now I’ll never know for sure! Nonetheless, considering his rating, our tough fight, and the fact that I was drowning in games due to the Gamesquad – Not Gamesquad Tourney, I offered a ceasefire and he accepted.

The following is the PGN:

[Site "www.ChessWorld.net"]
[Date "2008.1.28"]
[White "Crusader Scott"]
[Black "roo62"]
[ECO “C41 Philidor's Defense / Exchange Variation 3.d4 exd4 4.Nxd4 Nf6”]
[Result "1/2-1/2"]
[WhiteELO "1862"]
[BlackELO "2063"]

1.e4 e5 2.Nf3 d6 3.d4 exd4 4.Nxd4 Nf6 5.Nc3 c5 6.Nf3 Nc6 7.Bc4 h6 8.Be3 g5 9.Qd2 Ng4 10.O-O-O Nxe3 11.Qxe3 Bg7 12.Rhe1 O-O 13.Kb1 Qf6 14.a3 a6 15.e5 Nxe5 16.Nxe5 dxe5 17.Ne4 Qe7 18.Rd2 b5 19.Bd5 Rb8 20.Nxc5 a5 21.Be4 a4 22.Bd3 b4 23.Qe4 f5 24.Qd5+ Kh8 25.axb4 Rxb4 26.c3 Rb8 27.Nxa4 Be6 28.Qc5 {Fritz remarks "allows the opponent back into the game 28.Rxe5 and White hangs on 28...Bxe5 29.Qxe5+ Kh7 30.Bc4. I'd reached the same conclusion one turn later. "} Qb7 29.Rxe5 Bb3 30.Re7 Qc8 {CMXI comments “Leads to 31.Qa7 Ra8 32.Qb6 Rf7 33.Rxf7 Bxf7 34.Nc5 Bf8 35.b4 Kg8 36.Kc1 Bg7 37.Ba6, which wins a rook for a rook. Better is Qxe7, leading to 31.Qxe7 Rfe8 32.Qxe8+ Rxe8 33.Ka1 Bxa4 34.Bxf5 Bb3 35.Bc2 Be6 36.Be4 Bc4 37.Bf5, which wins a queen, a rook, and a knight for a queen, a rook, and a pawn. This was black's only significant mistake, but black was still able to manage a draw.} 31.Qxc8 Rfxc8 32.Bc2 Bxa4 33.Bxa4 Bxc3 34.Rde2 Bg7 35.b3 Bf8 36.R7e5 f4 37.Rb5 1/2-1/2

This is the final position:


What do you think? Was I correct to draw?

I couldn’t get this intriguing position out of my mind. My b3 pawn looked like it could, conceivably, be promoted at some point if I had persisted. I needed to know! So I fed the position into CMXI and this is what it came up with:

37…Rxb5 38.Bxb5 Kg7 39.Re6 Bc5 40.Bd3 Rd8 41.Kc2 Rc8 42.Kd2 Rb8 43.Kc3 Rc8 44.f3 Bg1+ 45.Kb4 Rb8+ 46.Ka4 Ra8+ 47.Ra6 Rd8 48.Be4 Bxh2 49.Rg6+ Kf7 50.Rxh6 Bg1 51.Rh7+ Kf6 52.Rc7 Rd2 53.Rc2 Rxc2 54.Bxc2 Ke5 55.Ka5 Kd6 56.b4 Kc7 57.Be4 Bf2 58.Kb5 Be3 59.Kc4 Kb6 60.b5 Kc7 61.Bc6 Kd6 62.Kd3 Bb6 63.Ke4 Ba5 64.Kf5 Bd8 65.Kg6 Kc7 66.Kg7 Kd6 67.Kh6 Kc7 68.Kg6 Kd6 69.Kf5 Kc7 70.Kg4 Kd6 71.Kh3 Kc5 72.g3 Kd4 73.Kg2 Ke3 74.g4 Bb6 75.Kf1 Kd4 76.Ke2 Kc3 77.Kd1 Kd3 78.Kc1 Kd4 79.Kb2 Kc4 80.Ka3 Bc7 81.Ka4 Bb6 82.Bd7 Bc7 83.Be8 Bb6 84.Ka3 Bd8 85.Ka2 Bb6 86.Kb2 Bd4+ 87.Kc2 Bc5 88.Bc6 Bd4 89.Kd1 Bb6 90.Kd2 Be3+ 91.Ke2 Bd4 92.Be8 Be3 93.Bd7 Ba7 94.Kd1 Kd5 95.Bc8 Kc5 96.Ba6 Kb6 97.Ke2 Bb8 98.Kd3 Kc5 99.Ke4 Bc7 100.Kf5 Bd8 101.Ke5 Ba5 102.Ke6 Bc7 103.Kd7 Be5 104.Ke8 Bc7 105.Kf7 Bd8 106.Kg6 Be7 107.Kh7 Bf8 108.Kg8 Be7 109.Kg7 Kb6 110.Kf7 Bd8 111.Ke6 Kc5 112.Kd7 Ba5 113.Ke7 Bb6 114.Kf6 Bd8+ 115.Kg7 Kb6 116.Kg6 Be7 117.Kf5 Bd8 118.Ke4 Be7 119.Ke5 Bd8 120.Kd5 Be7 121.Bc8 Bc5 122.Bb7 Be7 123.Ba6 Bf8 124.Bb7 1/2-1/2

With perfect play, another draw after an additional 87 moves! I guess I made the right choice after all!

However, move 121.Bc8 intrigued me. Why didn’t Black take advantage of the unprotected pawn on b5?



I restarted the analysis with the assumption that Black takes the pawn with 21…Kxb5.

Result:

1.Bd7+ Kb4 2.Be6 Bf6 3.Bf7 Bd8 4.Bg6 Kc3 5.Ke4 Kc4 6.Be8 Bf6 7.Bf7+ Kc5 8.Bg8 Bd8 9.Be6 Bb6 10.Bf7 Ba5 11.Kf5 Bd8 12.Ke5 Bc7+ 13.Kf6 Bd8+ 14.Ke6 Kd4 15.Be8 Ba5 16.Kf5 Bd8 17.Bc6 Be7 18.Be4 Ke3 19.Ke5 Bd8 20.Bc6 Bc7+ 21.Kf6 Bd8+ 22.Ke6 Kd4 23.Bb7 Ba5 24.Kf6 Bd8+ 25.Kf5 Be7 26.Bc6 Bd8 27.Bb7 Be7 28.Bc6 Bd8 29.Bb7 1/2-1/2

Another draw! I guess there’s no getting around it.

Of course, the above analysis presupposes perfect play on both our parts…which is super-extremely unlikely. But all things considered, I think I can now rest easier knowing I didn’t throw a victory away needlessly. Aren't we Chess players fortunate to have such wonderful analysis engines? Go ahead, brag a little to the non-chess players.

Opinions?

 

[Lurker]

Interesting game Scott. A quick glance at your final position would make one think of queening that passed pawn. However, bishops of opposite color endings are tough ones to win in such situations; particularly with his king side pawns all fixed on black squares, so I think your decision was good.

As a note, after 5...c5 I think 6.Bb5+ followed by Nf5 is better and gives white a slightly bigger edge.

I ran the game by Rybka and after 11...Bg7 it gives 12.e5, Nxe5 13.Bb5+, Ke7 with a decisive edge +- (1.84) for white at a depth search of 17. 12.Nb5 also gives a nice edge (+ 1.02). Rhe1 gave =.

 

[Scott Tortorice]

Interesting game Scott. A quick glance at your final position would make one think of queening that passed pawn. However, bishops of opposite color endings are tough ones to win in such situations; particularly with his king side pawns all fixed on black squares, so I think your decision was good.

As a note, after 5...c5 I think 6.Bb5+ followed by Nf5 is better and gives white a slightly bigger edge.

I ran the game by Rybka and after 11...Bg7 it gives 12.e5, Nxe5 13.Bb5+, Ke7 with a decisive edge +- (1.84) for white at a depth search of 17. 12.Nb5 also gives a nice edge (+ 1.02). Rhe1 gave =.

CMXI and Fritz X agrees with your analysis of 6.Bb5+. I reacted defensively when I should have been thinking offensively.

As for 12.e5, Fritz agrees, but sees the following continuation: 12.e5!? 0–0 13.Rxd6

I was playing it safe again.

How is Rybka?

Thanks for the comments!

 

[Lurker]

As for 12.e5, Fritz agrees, but sees the following continuation: 12.e5!? 0–0 13.Rxd6

How is Rybka?

Interesting differences in engine analysis. After 12.e5 0-0 Rybka gives it as a decisive advantage for white at depth 17 with +1.96:
13.Rxd6 Qe8 14.Nd5 Bxe5 15.Rxh6 Bf4 16.Nxf4 Qxe3+ 17.fxe3 gxf4 18.exf4.

At a depth of 20 or more it might reevaluate that but it probably wouldn't be by much.

Rybka has been the strongest commercial program in the world for awhile now, but I'm not sure of the very latest standings.

It's analysis interface isn't as good as CM apparently as it doesn't have the verbal analysis to go with the code analysis.

 

[Scott Tortorice]

It's analysis interface isn't as good as CM apparently as it doesn't have the verbal analysis to go with the code analysis.

I find CM's verbal analysis to be helpful in a friendly sort of way. If you're a beginner, having a program ditch the dry analysis for some coach-like verbal analysis is less scary. I guess Chessbase agrees because Fritz's verbal annotations are a step in a similar direction.

Rybka has been the strongest commercial program in the world for awhile now, but I'm not sure of the very latest standings.

I just read an article where GM Dzindzichashvili took on Rybka in a four-game match (the GM had a one pawn advantage in each game) and tied the machine 2-2. You can read it here.


Post some screenshots of Rybka!

 

[Lurker]

Ryba has nailed a few GMs in matches so far; all with a starting pawn advantage to the GM. So far it's beaten Joel Benjamin, Elvhest and Fontaine.
Dzindzi was the first draw.

Not too bad for a commercial program!