We return again to Alexander Alekhine's book of the New York 1927 international tournament. The fourth-round game between J.R. Capablanca and Milan Vidmar quickly reached a "hedghog" type position that would become familiar territory in the later stages of the 20th century. Alekhine's notes are interesting as always, but there is an improvement, of which more later.Capablanca – Vidmar, New York 1927
Queen's Indian Defence
Commentary by Alexander Alekhine
1.d4 Nf6 2.Nf3 e6 3.c4 b6 4.g3 Bb7 5.Bg2 c5
I have always considered this move to be unfavourable in this position because of the possible reply 6.d5. I prefer both 5...Bb4+ and 5...Be7.
Though quite playable, this reply is rather inoffensive in nature and allows Black to equalize the game in a number of ways.
6...cxd4 7.Nxd4 Bxg2 8.Kxg2 Be7
But this is inconsistent, because Black should be striving to rid himself of the weakness along the d-file as soon as possible. Although this aim would not be served by the immediate 8...d5 in view of 9.Qa4+ (as in the third match game Capablanca-Alekhine, Buenos Aires 1927), Black could make an effective preparatory move here: 8...Qc8!, for example: 9.b3 (9.Qd3 Nc6!) 9...Be7 10.Bb2 Qb7+ 11.f3 d5 12.cxd5 Nxd5 13.Qd2 0-0 14.e4 Nf6, etc.
Here too 9...Qc8 very much deserved attention, and if 10.b3 then 10...d5!
One move too late!
11.b3 Qb7 12.f3 Nc6 13.Bb2 Rfd8 14.Re1 Nxd4
Vidmar is playing the entire first part of the game inaccurately. Why, for example, does he develop the White queen in this position? Simpler was the immediate 14...d6.
Another obvious tempo loss!
16.Qd3 Be7 17.Rad1 d6
In spite of the multitude of inaccuracies committed by Black, his position is quite sturdy since in this varation the point d6 can, as is well-known, be defended without difficulty in the middlegame.
In the sequel Capablanca tries to make use of his last serious chance -- a flank attack. And indeed, this will demand exceptional accuracy from his opponent, who is very restricted in his freedom of action.
18.Re2 Rd7 19.Red2 Rad8 20.Ne2
Conclusively preventing – in the simplest way – the move ...d6-d5 (because of Bb2xf6 etc.)
In order to retreat the knight to e8, which at the moment would be met by 21.e5!
21.Qe3 h6 22.h4 Qb7 23.a4
In order to prevent the freeing move b6-b5 once and for all.
23...Ne8 24.Nf4 Bf6 25.Bxf6 Nxf6 26.g4 Nh7
After this White's attack has very little chance of success because the knight on f4 is restricted by the necessity of keeping the square d5 under observation. As soon as the White knight gives up control over this square (for example, after Nf4-h5) then the move e6-e5 would immediately follow, creating interesting attacking possibilities for Black's knight.
27.Qc3 Nf8 28.g5 hxg5 29.hxg5 Ng6!
Completely correct. If now 30.Nh5, then 30...e5 31.Kg3 Qc7 followed by Rd8-f8, Qc7-d8, etc. Thus White has nothing better than to exchange.
30.Nxg6 fxg6 31.Qd4 Qc6
The square c5 is now a sufficient counterargument for Black.
32.Kg3 Qc5 33.f4 Kf7 34.Kg4 a5 35.Rh2 Qxd4?
Because it not to White's advantage to exchange queens (over the last few moves because of ...b6xc5 with pressure on the b-pawn, and at this moment because of ...d6xc5 followed by an invasion along the d-file), and retreating the queen (to c3 or b2) is also not good for him on account of ...d6-d5!, etc, then the simplest way for Black to draw here was to adhere to waiting tactics. Correct therefore was ...Kf7-g8-f7-g8, etc., leaving it to White to make a decision about transforming the position. That Vidmar willingly goes into an endgame that is very dubious (to say the least) can only be explained by time pressure.
36.Rxd4 Re7 37.Rhd2 Red7 38.f5
In my opinion this tempting advance should have been delayed for one move: 38.R4d3! Ke7 (or Ke8) 39.f5 gxf5+ (if 39...Kf7 then 40.fxe6+ followed by 41.Rd5 and then Rd5-b5, or e4-e5 when convenient, with advantage to White) 40.exf5 exf5+ [stronger is 40...e5!?, when Black should hold – tr.] 41.Kxf5 Kf7 42.g6+ with a winning position. And in other variations too it would be easier than after the move in the game to transform White's positional advantage into material gain.
38...gxf5+ 39.exf5 exf5+ 40.Kxf5 g6+ 41.Kg4 Kf8
The only defence – and a temporary one at that – against the threat of Rd2-f2-f6 with a winning position. If 41...Ke8 then 42.Rh2!, etc. (see below).
This leads only to a draw. The proper way to realize the advantage that has cost White so much effort was – again – the win of a tempo: 42.R4d3!; then 42...Ke8 is met by 43.Rh2!, and after the exchange of a pair of rooks the remaining White rook penetrates to f6 or b8: 43...Kf8 (or 44...Rf7 44.Rh6, etc.) 44.Rf3+ Kg7 45.Rf6 (threatening 46.Rh6) 45...Rf8 46.Rxf8 Kxf8 47.Rh8+ followed by Rb8, etc. However, Capablanca forces the exchange of rooks under a different – and less advantageous – pawn configuration, after which Vidmar is able to save the game without difficulty.
If 43.Rfd2, then 43...Ke7 with a draw.
43...Rxf6 44.gxf6 Kf7 45.Kg5 Re8 46.Rxd6 Re5+ 47.Kf4 Re6
Much simpler than 47...Rf5+ and 48...Rxf6.
Here it was possible to quietly draw the curtain.
49.Rb5 Ke7 50.Kg5 Rc6 51.Kh6 Kf8 52.Rg5 Kf7 53.Rg3 Re6 54.Rd3 Re5 55.Rd7+ Kf6 56.Rd6+ Kf7 1/2-1/2
With this we conclude our extract from Alexander Alekhine's book of the New York 1927 international tournament. But wait - there's an improvement!
After Black's move 29...Ng6, Russian GM Sergei Shipov demonstrated the following imaginative continuation for White:
30.Nh5! e5 31.Rh1 a6 32.Rd5!! Rc8
If 32...b5 33.axb5 Ne7 34.f4 Nxd5 35.exd5 f6 36.g6 Kf8 37.Qh3, etc.
33.f4 b5 34.axb5 axb5 35.f5 bxc4 36.b4!
Thus far Shipov's analysis in his book The Complete Hedgehog, Volume 1. Here is a likely conclusion:
36...Ne7 37.Nf6+ gxf6 38.Qh3 Kf8 39.Qh6+ Ke8 40.Qh8+ Ng8 41.Qxg8+ Ke7 42.gxf6+ Kxf6 43.Rh6+ Ke7 44.f6+ Ke6 45.Qg4#!
Posting a rook on d5 in order to stabilize White's central position was unusual but very strong in this particular case.