Chigorin vs Gunsberg, 1889 (28...?)

Black to play and win.

Material: Up a P. The White Kh2 is stalemated and vulnerable to checks from Qe1. The White Qc4 and particularly Na4 cannot aid Kh2 rapidly, suggesting that Black can build a lethal local superiority around Kh2. White threatens to capture Pc5, to achieve material equality. The Black Kg8 is secured from check, and moreover from perpetual check as well, because the Black Qe1 prevents the repetitive sequence Qd8-h4-d8.

Candidates (28...): Nc1

28…Nc1 (threatening 29…Ne2 30…Qg1#)

Only the infeasible defense of sacrificing Qc4 prevents mate in the following. <[According to Toga, 29.Nc3 staves off mate at a value worse than -7 P.] >

29.h4 [g3 or g4 Qf2+ 30.Kh1 Ne2 (threatening 31…Qg1#)]

[29.Qxc5 Ne2 30.Qc8+ Kh2, and White runs out of checks]

29…Ne2 (threatening 30…Qxh4#)

This game was played after the final 38th round, as a required replay of their 34th round drawn game. After 38 rounds, Chigorin had scored 29 points and had only his drawn game with Gunsberg to replay.

After 38 rounds, Weiss had 26 1/2 points, and 4 drawn games to replay. Weiss scored 1 win and 3 draws in these additional games, and tied Chigorin for first place with 29 points. Gunsberg was third with 28 1/2 points.

As Chigorin and Weiss were tied for first place, the tournament rules required an additional contest between these players to determine the tournament winner.

28... Nc1 and now:

A) 29.Qxc5 Ne2 30.Qc8+ Kh7 31.h4 (otherwise Qg1#) Qg1+ 31.Kh3 Qh1+ 32.Kg4 Qxg2#

B) 29.Nxc5 Ne2 30.Qxe2 appears to delay checkmate the longest.

29.Qb5 Ne2 30. Qb8+ Kh2 31.h4

Black is threatening mate starting with 27...Qe1+. Therefore white's candidates are restricted to stopping this. We've already seen what happens on 27.h3, but what about 27.Kg1, 27.Qf1, 27.g3, or 27.g4?

On 27.Kg1, I would consider first 27...Qe1+ 28.Qf1 Qxf1+ 29.Kxf1 and try to get a feel for what's going on. There is more in terms of forcing moves, such as 29...Nb4 30.c3 dxc3 31.Nxc3. Black does have a pawn advantage so I would say black is at least a little better given the forcing line above.

With 27.Qf1, this move just looks bad because it uses the queen (a powerful piece) in pure defense. I think there is hardly any reason to look at this further and I seriously doubt a master would spend much (or any) time on this.

After 27.g3, black still has a pawn advantage but c5 remains very weak. Steinitz is probably right about this leading to a draw. Even if not true, so far this appears to be the best move.

27.g4. It doesn't appear much worse than 27.g3 except it looks weaker. I would feel much safer with 27.g3, which looks more solid and that's good enough for me. 27...Qf4 becomes an option in this line.

To me, this is what most of chess is about--decisions. Not analyzing deep or finding a tactical shot, but trying to compare candidates and evaluating positions in an efficient manner and deciding which candidate is best given the time constraints. Of course tactics are of the utmost importance and analyzing deep may be required in some positions.

I admit that my first impression was that this was going to be some sort of promotion tactic, since the knight seemed too far away to join the queen in a king attack.

But the more I looked at it, the idea of 28...Nc1 & ...Ne2 & ...Qg1# seemed very difficult for black to stop effectively.

For example, 28...Nc1 29.h4 Ne2 30.Kh3 Qh1+ 31.Kg4 Qxg2#

No better is 28...Nc1 29.Qxc5 Ne2 since the knight resourcefully guards Pd4, preventing the white Q from guarding g1.

I really misunderstod the power of those connected pawns. Because, if play had continued 29 Qxc5 Ne2, as others have indicated, the pawn at d4 becomes a mightly blocker, protected by the knight, not allowing the white queen access to g1.

click for larger view

If that d pawn could be magically removed after 29 Qxc5 Ne2, the position is a draw.

click for larger view

1: Mikhail Chigorin - Isidor Gunsberg, USA-06.Congress New York (34-2) 1889

click for larger view

Analysis by Rybka 3 :