Planning: Capablanca - Dus-Chotimirsky, St. Petersburg 1913

Continuing with Capablanca's games 'to be studied', the third World Champion considered that the diagrammed position 'should be carefully studied'. It was annotated by Garry Kasparov as game 78 of 'My Great Predecessors, Part I'.

The position is an excellent example of how logical thinking translates into a series of moves. Capablanca wrote,

This position should be carefully studied. It is evident to White that Black wants to play 20...Nd7 followed by 21...Ne5 or 21...Nc5 and 22...Na4, forcing the advance of the b-Pawn in some cases, and then through the combined action of the Bishop at f6, the Pawn at d4, and the Knight ultimately at c3, cramp White's game so as to make it impossible for him to hold out.

It is against this plan that White must evolve another that will meet it at every point. If this can be done, then White must come out on top, as he will be able in the long run to concentrate sufficient forces against the Pawn at d4 or the Pawn at b5, and take either one or the other. The text will show how this is done.

St. Petersburg 1913
Dus-Chotimirsky, Fyodor

Capablanca, Jose Raul
(After 19...a6-b5(xP))
[FEN "2r2rk1/4bppp/1q1p1n2/1p1P4/3pP1b1/3B1N2/1P3PPP/R2QRNK1 w - - 0 20"]

The Cuban continued 20.h3! (all '!'s are from his notes), and after 20...Bxf3 21.Qxf3 Nd7 22.Rec1! Nc5, noted,

White has not only led on Black to this maneuver, but what is more he will now induce him to go on with 23...Na4.

He did this with 23.b4! Na4. Capablanca:

While this game was being played there were present, besides masters of lesser rank, two of the leading players of the world, and they thought I had allowed my opponent to obtain a winning position. They had not seen my 25th move, which was to turn the tide of the battle. Had now Black played 23...Nxd3 then 24.Qxd3 Rc3? 25.Rxc3 dxc3 26.Ne3 Bf6 27.Nc2 followed by 28.Ra5 and White has the better game. Probably the best line of play for Black would be: 23...Nxd3 24.Qxd3 Bf6.

Kasparov agreed with this last line and continued, 'the sacrifice of the b-Pawn would have led to an unclear game: 25.Rxc8 Rxc8 26.Ra5 Qc7 27.Qxb5 g6.

After 24.Rxc8 Rxc8, Capablanca played the powerful 24th move mentioned in the last note: 25.e5!. Kasparov also gave the move a '!', but wrote, 'I don't understand why it was thought that Black had a 'winning position': even after the dull 25.Ng3 nothing bad for White is apparent.' One wonders what Capablanca's colleagues were thinking.

Black tried to defend with 25...g6, which Capablanca considered inevitable:

White threatened 26.Qf5. Had Black played 25...Rf8, he would later on be forced to play 26...g6.

The game continued 26.e6 Rf8 27.Ng3! Qb7. Capablanca:

If 27...fxe6 28.Qg4 threatening both 29.Bxg6 and also 29.Qxe6+.

Now White finished powerfully with 28.Nf5! fxe6. Capablanca:

28...Kh8 was best, but then 29.Qe4 should win. Black has been wanting to take this Pawn all the time, and thinks the time has come, but it only hastens the result.

After 29.dxe6 Qc7 30.Qc6!, Black resigned a few moves later. To play through the complete game see...

Jose Raul Capablanca vs Fyodor Ivanovich Dus Chotimirsky, St Petersburg exhibition 1913
http://www.chessgames.com/perl/chessgame?gid=1294504

...on Chessgames.com.

Endgame: Capablanca - Janowsky, San Sebastian 1911

This game, the third in a series on Capablanca's games 'to be studied', was also annotated by Garry Kasparov as game 77 of 'My Great Predecessors, Part I'. Capablanca called the endgame 'a masterpiece, which should be very carefully studied'.

As Capablanca pointed out, Janowsky had played very well up to the diagrammed position, and the Cuban was dead lost. After the next move 53.Bxe5, Black should have played 53...Qh1+ 54.Ka2 Nxe5, with an easy win. Instead he played 53...Qe1+ 54.Ka2 Nxe5. Capablanca offered a long explanation as to why Janowsky suffered this sudden attack of chess blindness, which goes beyond the aim of this post.

San Sebastian 1911
Janowsky, David

Capablanca, Jose Raul
(After 52...Qg6-e4)
[FEN "6k1/2B3p1/1P6/p3p3/N1p1q3/P2n3p/2Q5/1K6 w - - 0 53"]

Commenting on Black's last move, Capablanca wrote,

Janowsky, like the other masters watching the game, never thought that it would be possible to obtain more than a draw out of the game, hence his not taking the last chance afforded him to draw by perpetual check with 54...Nc1+. Before continuing I should add that the end-game coming is the finest of its kind ever played over the board, and that for some reason it has not been properly appreciated. It is a masterpiece, one of which I am very proud, and which should be very carefully studied. As I said, nobody thought at the time that the ending could be won.

The game continued 55.b7 Nd7 (both moves are forced) 56.Nc5. Capablanca:

An all-important move. At first sight it looks as though 56.Nb6 would be better. A profound study will show that this is not the case.

According to Kasparov's analysis, which appears to be correct, Capablanca's move shouldn't have won either. After a further 56...Nb8 57.Qxc4+ Kh8 58.Ne4. Capablanca noted,

The beauty of White's 56th move is now seen. Black has no check with his Queen, nor can the Pawn advance, because through a combination of checks I am threatening to win the Queen or obtain a similar position to the one I finally obtained in the game.

Now Black played 58...Kh7. Kasparov gave this move a '?', and explained, 'Only this is the decisive mistake! Two Queen moves would have left Janowski with every chance of saving the game.' First he analyzed 58...Qh4; his analysis is not entirely convincing, because there are several unexplored branch points and because the main line steers into a position with four Queens on the board.

His analysis of 58...Qe3, where the straightforward main line ends in perpetual check for Black, is more convincing. Janowsky's 58...Kh7 lost quickly to 59.Qd3. Kasparov wrote, 'For Janowsky the undeserved defeat became the drama of his entire life'. To play through the complete game see...

Jose Raul Capablanca vs David Janowski, San Sebastian 1911
http://www.chessgames.com/perl/chessgame?gid=1064741

...on Chessgames.com.

Capablanca is 'outplayed'

What's the worst way to lose a chess game? A blunder? A terrible blunder, like hanging your Queen?? In my opinion, the worst way to lose a game is to be outplayed.

In his notes to the game with Janowsky from San Sebastian 1911, Capablanca commented after his 13th move,

In this game I was for the first time in my life to have the feeling of being completely outplayed by my opponent; time after time up to my 23rd move, I would figure out some reply of my adversary only to find out immediately that I was wrong, and that some other move that he had made was superior to the one I had thought best.

That is an excellent definition of 'outplayed' : to find out time after time that some other move of your adversary is superior to the one you thought best.

Why is it the worst way to lose? Because it means that there is something about chess you don't understand, something that your opponent understood, but which you didn't. You don't even know what it is that you don't understand.

The Queen's Indian in the West

While preparing an article on the Queen's Indian Defense (QID), I consulted my usual references on the opening, among them the first five volumes of Kasparov's 'My Great Predecessors'. Tying together Kasparov's notes on a specific opening provides an excellent history of the evolution of that opening.

There were several good references in each volume to the QID. Vol.1 included Saemisch - Nimzowitsch, sometimes called the 'Immortal Zugzwang Game', which used the Rubinstein Variation. Another game was from the 1843 Staunton - St.Amant match, where Staunton transposed from a Queen's Gambit Tarrasch Defense into a 4.e3 variation of the QID. This is one example why Staunton is often regarded as an excellent opening theoretician, years ahead of his time.

The only volume which had no QIDs was the fourth. That is the 'Stars of the West' volume, featuring Reshevsky, Najdorf, Larsen, and Fischer. This struck me as curious. Have there been any great Western proponents of the QID or is it an opening which was the exclusive property of the Soviet School? Vol.2 had one game where Euwe played the White side against Keres, but no games with Euwe as Black.

Against 1.d4 Nf6 2.c4 e6 3.Nf3, Fischer played only 3...d5 or 3...c5. On Chesslab.com I found 33 games where Reshevsky played 3...b6, 30 games by Najdorf, and 39 by Larsen. It must have been an accident that none of these games was chosen for Predecessors Vol.4. They were all played in the era when the QID was considered stodgy and drawish.

Combination: Capablanca - Bernstein, San Sebastian 1911

Continuing with Capablanca's games 'to be studied', the diagram shows a position where Capablanca played a combination that he considered 'one of the longest and most difficult ever played over the board'. The same position was chosen by Kasparov as the starting point for game 76 of 'My Great Predecessors, Part I'.

Capablanca played 22.Ne2, sacrificing a Pawn to 22...Qxa2. He mentioned,

22...Qb6 instead of the text move, would have simplified matters, but at any rate White would have had the superior position.

As we will will see below, this is the first of several points where Kasparov disagreed with the Cuban.

San Sebastian 1911
Bernstein, Ossip

Capablanca, Jose Raul
(After 21...Nc5-e6)
[FEN "4rr2/p1pb1ppk/2ppn2p/q4N2/3NP1P1/1P2QP2/P1P4P/3RR1K1 w - - 0 22"]

After 23.Neg3, White's plan was clearer: the Knight is headed for h5. Bernstein grabbed a second Pawn with 23...Qxc2. Here Capablanca wrote,

This second capture is disastrous; as Lasker has pointed out, it was necessary to play 23...f6, to be followed by 24...Rf7 in case White continues the attack with Nh5.

After 24.Rc1 Qb2 25.Nh5, he remarked,

The march of this Knight is most remarkable. Even now it looks inoffensive, and yet it is the Knight that is going to decide the game.

Now Black played 25...Rh8. Capablanca:

There was nothing better. If 25...g5 26.e5 f6 27.Qd3 and with proper play White will win. I do not give the variations because they are very long and complicated.

This remark is why I decided this game deserved 'to be studied'. The third World Champion has left the analysis as an exercise. He also explained why 25...g6 is not satisfactory:

26.Qxh6+ Kg8 27.e5 gxh5 28.gxh5 and White wins because there is no way to stop one of the Rooks from checking in the open g-file.

After 25...g5 26.e5 f6 27.Qd3, it appears that Black's best is 27...Kh8. This could be followed by 28.Rc2 Qa3 (28...Nc5 29.Rxc5 also needs to be investigated) 29.Nh4 gxh4 30.Qg6.

The game continued 26.Re2 Qe5 (forced) 27.f4 Qb5 28.Nfxg7. Capablanca:

Now at last the result of the moves of the Queen Knight are clearly seen. This move marks the turning point of the long combination initiated on the 22nd move.

Now Bernstein blundered with 28...Nc5. Capablanca:

Weak. I expected 28...Nxg7, when would have come: 29.Nf6+ Kg6 30.Nxd7 f6 31.e5 Kf7 32.Nxf6 Re7 33.Ne4, and Black's position is untenable. A careful analysis and proper comparison will show that this combination is one of the longest and most difficult ever played over the board.

After 29.Nxe8, Black was lost but played on until the checkmate became unavoidable a few moves later.

Kasparov disagreed with Capablanca at several key points, starting with the Cuban's assessment of 22...Qb6. Of Capablanca's remark after move 25, the 13th World Champion wrote,

A typical comment: the third world champion did not greatly favor long and complicated variations! Yet no one has shown how after 25...g5 26.e5 f6 27.Qd3 Kh8!, White wins. But what is more important -- no one has shown what he is to do after 25...g5 26.e5 Nf4!, for example 27.Nxf4 Rxe5 28.Nd3 Rxe3 29.Nxb2 Rxb3, with an excellent game for Black.

He also analyzed 25...Rg8 and concluded that 'the only question is whether White can save the game.'

A few moves later, Kasparov gave a long analysis of 'the cool defense' 28...Rd8!, showing that, even after the previous inaccuracies, Bernstein had chances to save the game. This variation is worthy of a more detailed look, which will have to wait until another time.

To play through the complete game see...

Jose Raul Capablanca vs Ossip Bernstein, San Sebastian 1911
http://www.chessgames.com/perl/chessgame?gid=1264025

...on Chessgames.com.

Endgame: Marshall - Capablanca, match (game 5) 1909

Continuing with Capablanca's games 'to be studied', Capablanca wrote of the diagrammed position, 'It is from now on that it can be said that I played well. The ending is worth studying.' He gave no further notes.

The first question that comes to mind is, 'What is worth studying?'. Black is ahead in material a Bishop for a Pawn. Isn't this a simple problem of technique? Yes, it might be a technical win, but there is nothing simple about it. Black has three problems to tackle simultaneously:-

1. Stop the a-Pawn. It is not easily subject to capture, but stopping it from promoting isn't too difficult. The Queen and Bishop can form a battery on a diagonal to stop it from going to a6 or a8.
   
   
2. Stop the White Queen from harrassing the Black King, which has little Pawn cover. This job will go to the Black Queen, although the Black Bishop performs the important job of protecting Black's f-Pawn. The Black Queen can either block checks on the lines or prevent them by covering the checking squares.
   
   
3. Develop an attack against the White King. For this, the Black Bishop would appear to best placed on the a8-h1 diagonal.

It's also important to note that White can't allow a Queen exchange. If Queens are exchanged, the Black King moves to the Queen side and captures tha a-Pawn. The Black Bishop keeps the a-Pawn from promoting and prevents the White King from forcing the exchange of all Kingside Pawns. After eliminating all other Pawns, the h-Pawn promotion square is the right color for the Bishop.

Capablanca played 40...Bb1, protecting against 41.Qxd3. It is useless to attack immediately with 40...Qc1+ 41.Kh2 Qd2+ (41...Qc4 42.Qd7+ Kg6 43.Qe8+ Kf6 44.a4) 42.Kg1 Qe3+ 43.Kg2 Qe2+ 44.Kh3 Qf1+ 45.Kh2 Qf2+ 46.Kh1, when Black has gained nothing.

Match 1909
Capablanca, Jose Raul

Marshall, Frank
(After 40.Kh2-g1)
[FEN "8/7k/7p/3Q1p2/8/3b1PP1/Pq6/6K1 b - - 0 40"]

The game continued 41.a4 Qa1 42.Qb7+ Kg6 43.Qb6+ Kh5 44.Kh2 Ba2. The Bishop aims for d5. 45.Qb5 Kg6 46.a5 Qd4 47.Qc6+ Qf6 48.Qe8+ Qf7 49.Qa4 Qe6. Black can play 49...Bd5, but first positions the Queen for action against the White King. This prevents a continuation like 50.Kg2 Qe6 51.Kf2. Note how the Black Queen covers the White Queen's checking squares.

White played 50.a6. This lost immediately to 50...Qe2+ 51.Kh3 Bd5 52.a7 Bxf3 0-1, so it's a blunder. A better continuation would be 50.Qb5 Bd5 51.Kg2 Bc4 52.Qc5 (52.Qb2 Qe2+) 52...Qe2+ (With a battery on the a6-f1 diagonal.) 53.Kg1 Qe1+ 54.Kh2 Qd2+ 55.Kg1 Bd5 56.Qd6+ Kh5 (No more checks.) 57.Qe5 Qd1+ 58.Kh2 Qxf3 and wins anyway.

To play through the complete game see...

Frank James Marshall vs Jose Raul Capablanca, m 1909
http://www.chessgames.com/perl/chessgame?gid=1094814

...on Chessgames.com.

The Marshall IQP

The August Chess Life had an interesting article by GM Andy Soltis, 'The Marshall Variations of the Sicilian and French Defenses'. Soltis' articles are always interesting, but this was about two similar variations that I had bever seen before.

The diagram shows the position after 1.e4 c5 2.Nf3 e6 3.d4 d5. This is ECO B40.01, which Soltis calls 'The Marshall Variation of the Sicilian'. ECO also attributes 3...d5 to Frank Marshall.

'The Marshall Variation of the French' is 1.e4 e6 2.d4 d5 3.Nc3 c5, ECO C10.01. The authoritative ECO agrees again that the third move is Marshall's. The position is similar to the digram, except a Knight has been developed to c3 instead of f3.

My old copies of ECO only give a few lines in each variation, all of them ending with advantage to White (plus over minus). The problem is that Black ends up with an isolated d-Pawn. This leads to the question: If the IQP is playable in the French Tarrasch (3.Nd2 c5), why isn't it playable here?

Both variations should be worth a try in non-serious games.

Marshall - Capablanca, match (game 5) 1909

This is the first post in a series on Capablanca's games 'to be studied'. The diagrammed position occurred earlier in the game than the point where Capablanca suggested that the position was worthy of study. It is a good example of Capablanca's minimalist approach to annotation. Capablanca played 26...Rxd4 and remarked,

Not the best, 26...Qf6 was the right move. Incidentally it would have saved me a great deal of trouble which I had to win the game. Here I will call attention to the poor notes sometimes written by analysts. Games are often annotated by unknown players who have not sufficient knowledge of the game. As a matter of fact, the games of the great masters, at least, can only be properly annotated by a few players. Of course even the best are not exempt from mistakes, but while they make them few and far between the others do so continuously.

Since he gave no further analysis, I looked closer at the suggested move.

Match 1909
Capablanca, Jose Raul

Marshall, Frank
(After 26.Qg3-c7)
[FEN "3r2k1/pbQ2pq1/1p2p2p/8/3PP1nP/8/P4PP1/1B1R2K1 b - - 0 26"]

Capablanca's recommended move 26...Qf6 threatens 27...Qxf2+ 28.Kh1 Rxd4 with mate in a few moves. The Knight on g4 plays a key role by confining the White King to the back rank.

If White plays 27.f3, then the same move played in the game, 27...Rxd4, is even more powerful. If White plays 27.Qg3, withdrawing the Queen from its aggressive position, then 27...h5 protects the Knight and keeps it active in the Kingside attack. It also provides an escape square on h6 should the Knight be forced to retreat by f2-f3. If there was any doubt, Capablanca was right; 26...Qf6 was a better move.

To play through the complete game see...

Frank James Marshall vs Jose Raul Capablanca, m 1909
http://www.chessgames.com/perl/chessgame?gid=1094814

...on Chessgames.com.

Capablanca's Games 'To be studied very carefully'

Capablanca did not leave a large body of annotated games. When he did annotate a game, he often left more out than he put in. His book 'My Chess Career' is full of comments like 'each move should now be studied with care' with little or no further analysis. This is particularly true of his endgame analysis.

While flipping through the book I noted ten games where the Cuban leaves it to the reader to analyze the play...

p.032 1909 USA m (New York), Marshall - Capablanca
p.053 1911 San Sebastian, Capablanca - Bernstein
p.061 1911 San Sebastian, Capablanca - Janowski
p.091 1913 St.Petersburg exhibition, Capablanca - Dus Chotimirsky
p.096 1913 St.Petersburg exhibition, Capablanca - Alekhine
p.107 1914 Riga, Nimzowitsch - Capablanca
p.115 1914 Vienna, Kaufmann & Fahndrich - Capablanca
p.123 1914 St.Petersburg prel, Capablanca - Bernstein
p.129 1914 St.Petersburg prel, Nimzowitsch - Capablanca
p.133 1914 Buenos Aires, Capablanca - Molina & Ruiz

...Most of these concern endgames, but several refer to middlegame combinations. In future posts I'm going to tackle the list with the help of computer analysis. Since Capablanca's games and ideas were extremely profound, it may take more than one post to cover a single game.

Fortress • Keres - Portisch, Moscow 1967

A potential fortress position was pointed out by Garry Kasparov in 'My Great Predecessors, Part 3', in the chapter on Petrosian (p.124). The position was analyzed in the subsection on Portisch, which is why the ninth World Champion doesn't figure as one of the players.

The position is bad for White. Both the b- and e-Pawns are weak and subject to capture.

Moscow 1967
Portisch, Lajos

Keres, Paul
(After 44...Bf8-h6)
[FEN "8/8/3k2pb/1pn1p2p/pNp1P3/P1P1NK1P/1P4P1/8 w - - 0 45"]

Kasparov didn't say at which point the game had been adjourned, but Keres found the next move during his adjournment analysis. It is entirely possible that Portisch also looked at the sequence during adjournment. 45.h4! Nd3 46.Nd1. Here Kasparov noted that 46.Nxd3? loses after 46...cxd3 and 47...Kc5. 46...Bc1 47.Ke2!. The punctuation is all Kasparov's.

Kasparov: 'It transpires that after 47...Nxb2? 48.Nxb2 Bxb2 49.Kd2 Bxa3 50.Kc2 White has a fortress: 50...Bxb4 51.cxb4 and the remaining White Pawns create an impenetrable barrier in front of the Black King, or 50...g5 51.g3! Kc5 52.Na6+ Kb6 53.Nb4 with a draw.'

Portisch avoided the fortress with 47...Nc5. After 48.Kf3 g5!, he eventually won on the 90th move. To play through the complete game see...

Paul Keres vs Lajos Portisch, Moscow 1967
http://www.chessgames.com/perl/chessgame?gid=1073169

...on Chessgames.com.

Sveshnikov or Chelyabinsk?

While browsing Part 3 of Kasparov's 'Great Predecessors', which is the volume on Petrosian and Spassky, I noticed that Kasparov called the moves 1.e4 c5 2.Nf3 Nc6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 e5 6.Ndb5 d6 7.Bg5 a6 8.Na3 b5 9.Bxf6 gxf6 10.Nd5, the Chelyabinsk Variation (p.54). I had always thought that the opening was a line in the Sveshnikov Variation and decided to check further.

Google counted 30-35K pages for each of the two phrases, with 12K pages for all three key words 'Sveshnikov', 'Chelyabinsk', and 'Variation'. Looking at the first few results for each phrase, I discovered that there is no consistency on the naming. The divegence starts early. A Chessville.com page said, 'The Sveshnikov Variation refers to 1.e4 c5 2.Nf3 Nc6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 e5. This move order can also refer to the Lasker, the Pelikan, or Pilnik.' Although I was familiar with the other names, Pilnik Variation was new.

A JeremySilman.com page called the move 5...e5 the Sveshnikov-Pelikan and said, '9.Nd5 Be7; An interesting alternative for Black here is Timoshchenko’s 9...Qa5+. This whole variation we are looking at is mostly known as the Chelyabinsk variation, because it was mainly developed by two strong Masters from that city, Sveshnikov and the same Timoshchenko.' This is at odds with Kasparov's 9.Bxf6. Perhaps the author was referring to 8...b5 when he mentioned the 'whole variation we are looking at'.

There are many other online examples which I won't repeat here. The 'Oxford Companion to Chess' (p.468) calls 5...e5 the Pelikan Variation, 8...b5 the Chelyabinsk Variation, and 9.Bxf6 gxf6 10.Nd5 f5 the Sveshnikov Variation. About the only name on which everyone agrees is 8...Be6, the Bird Variation. Calling 10...f5 the 'Sveshnikov Variation' is at odds with all other authorities.

Who is the definitive authority on opening names?

Blog Fishing

A few days ago, when I made my monthly round of the chess blogs (see Elsewhere on the Web Blog Tripping in July), I counted 71 blogs with posts in July. Of these, I counted 51 hosted on Blogspot.com. This makes it convenient to search for blog posts using the following search string...

Results from blogspot.com for chess.
http://www.google.com/search?hl=en&q=chess+site%3Ablogspot.com&btnG=Search&meta=

...A small enhancement narrows that search to posts made during the month of July...

Results from blogspot.com for chess inurl:2006 inurl:07.
http://www.google.com/search?hl=en&lr=&q=chess+site%3Ablogspot.com+inurl%3A2006+inurl%3A07&btnG=Search

...Another small enhancement allows searching for posts on a specific topic. This search brings up July posts about Jessie Gilbert ...

Results 1 - 10 of about 39 from blogspot.com for chess gilbert inurl:2006 inurl:07.
http://www.google.com/search?hl=en&lr=&q=chess+gilbert+site%3Ablogspot.com+inurl%3A2006+inurl%3A07&btnG=Search

A couple of other tricks for Blogspot.com are...

July 2006 archive for this blog
http://chessforallages.blogspot.com/2006_07_01_...

Atom feed for this blog
http://chessforallages.blogspot.com/atom.xml

...Both URLs can be built automatically knowing only the subdomain (chessforallages in this case) of the blog.

***

Note: I had to truncate the link text for the July 2006 archive URL. When published, it failed to wrap around, broke the container, and pushed the right column down the page to start at the end of the left column. Longer URLs in this post behaved properly. It appears that certain special characters allow wrap-around.

***

Note: I discovered afterward that the inurl trick ('inurl:/2006/07/') didn't work as I thought. I corrected it to the format shown ('inurl:2006 inurl:07').

Index to Alekhine's Annotated Brilliancies

• Introduction
  
• Alekhine - Zubarev, Moscow 1916
  
• Alekhine - Bogoljubov, Triberg 1921
  
• Alekhine - Sterk, Triberg 1921
  
• Alekhine - Rubinstein, The Hague 1921
  
• Tarrasch - Alekhine, Pistyan 1922
  
• Alekhine - Selezniev, Pistyan 1922
  
• Alekhine - Wolf, Pistyan 1922
  
• Alekhine - Rubinstein, Karlsbad 1923
  
• Gruenfeld - Alekhine, Karlsbad 1923
  
• Davidson - Alekhine, Semmering 1926
  
• Rubinstein - Alekhine, Semmering 1926
  
• Rubinstein - Alekhine, Dresden 1926
  
• Alekhine - Marshall, New York 1927
  
• Alekhine - Asztalos, Kecskemet 1927
  
• Alekhine - Steiner, Bradley Beach 1929
  
• Stahlberg - Alekhine, Hamburg Olympiad 1930
  
• Alekhine - Koltanowski, London 1932
  
• Alekhine - Alexander, Nottingham 1936
  
• Tartakower - Alekhine, Nottingham 1936
  

Alekhine - Koltanowski, London 1932

This is the last game in this series on Alekhine's annotated brilliancies, and it's a game I've admired many times. The preparation leading to the diagrammed position is just as interesting as the continuation, although for different reasons.

The position I've chosen as a start point was also used by Alexander Kotov in both 'Think Like a Grandmaster' (no.53) and 'Play Like a Grandmaster' (no.190). In both books Kotov copied Alekhine's analysis verbatim.

London 1932
Koltanowski, Georges

Alekhine, Alexander
(After 21...Qb7-b5)
[FEN "1r6/2p2rkp/p1npbpp1/1qpN4/4P3/PNQ1R2P/1PP2PP1/3R2K1 w - - 0 22"]

Alekhine played 22.Nxc7!, and commented,

As a rule, so-called "positional" sacrifices are considered more difficult, and therefore more praiseworthy, than those which are based exclusively on an exact calculation of tactical possibilities. The present position offers, I believe, an exception, as the multitude and complexity of the variations following the Knight's sacrifice demanded much more intensive mental work than any general evaluation of mutual possibilities.

The game continued 22...Rxc7 23.Rxd6 Bc4. Alekhine:

Black had several answers, but all of them would finally lose, as shown below.

The following analysis is all Alekhine's. Kotov used it as an example of what he called a 'coppice', a position which has many variations, but where the variations follow a sequence of forced moves.

• 23...Bxb3? 24.Qxf6+ followed by 25.Rxb3, etc.
  
• 23...Nd4? 24.Nxd4 etc.
  
• 23...Qc4 24.Nxc5! etc.
  
• 23...Nd8 24.Rf3 Rf7 25.Nxc5 etc.
  
• 23...Bf7 24.Rxf6! Nd4 25.Nxd4 cxd4 26.Qxc7 Kxf6 27.Rf3+ etc.
  
• 23...Re8 24.Nxc5 Nd8 25.b4 Nf7 26.Rxe6 etc.
  
• 23...Kf7 24.Rf3 Ke7 25.a4 Qb6 26.Rxe6+ Kxe6 27.Nxc5+ Kd6 (27...Kf7 28.Qxf6+ Kg8 29.Ne6! etc.) 28.Qxf6+ Kxc5 29.Rc3+ Kb4 30.Qd6+ and wins.
  
• 23...Re8 24.Nxc5 Nd4 25.Qxd4 Rxc5 26.Rf3.

24.a4! Qxa4 25.Nxc5 Qb5 26.Qxf6+ Kg8 27.Nd7! Rd8 Alekhine: 'Or 27...Re8 28.Qc3 and wins.' 28.Rf3 Qb4 29.c3 Qb5 30.Ne5! and Black resigned a few moves later.

To play through the complete game see...

Alexander Alekhine vs Georges Koltanowski, London 1932
http://www.chessgames.com/perl/chessgame?gid=1012843

...on Chessgames.com.

***

Later: With the help of an engine I discovered that Alekhine's combination is neutralized by 23...Re8 24.Nxc5 Nd4 (instead of 24...Nd8), with good play for Black. Then I discovered that this possibility was documented on Chessgames.com, and already known at the time of my original post.

Stahlberg - Alekhine, Hamburg Olympiad 1930

I'm nearing the end of Alekhine's annotated brilliancies. The present game was played in the 3rd Olympiad, where Alekhine represented France. Stahlberg was first board for Sweden. France finished 12th out of the 18 teams, even though Alekhine had a 100% score. None of his teammates, including the famous artist Marcel Duchamp on fourth board, scored better than 50%.

I had a small decision choosing the start position. The most difficult moves in the game, where Alekhine went from equality to the decisive advantage shown in the diagram, were less interesting than the following sequence.

Hamburg Olympiad 1930
Alekhine, Alexander

Stahlberg, Gideon
(After 26...Rd7-f7)
[FEN "4qrk1/1b3rpp/1p1p4/4p3/1PPn4/4Q3/R3BPPP/3RN1K1 w - - 0 27"]

Of the next move, 27.f3, Alekhine wrote,

One would suppose that this Pawn, besides being protected by its neighbor, and easily supported by 3-4 pieces, cannot possibly form a welcome object for Black's attack. And yet White's f3 will be captured, almost inevitably. It was certainly the unusualness of Black's winning strategem which induced the judges to award to this game the brilliancy prize.

27...Rf4 28.Bd3 Qh5. Alekhine: 'Threatening 29...e4 etc.' If 30.Qxd4, then exf3. 29.Bf1 Qg5!. Alekhine: 'With the main threat 30...Rxf3 forcing the win of the Queen.' 30.Rf2 h6!. Alekhine:

A terrible move in its simplicity. Black threatens 31...Rxf3 32.Qxg5 Rxf2, and in case of 31.Qd2 (comparatively the best), he would play 31...Bxf3 32.Nxf3 Nxf3+ 33.Rxf3 Rxf3 34.Qxg5 Rxf1+ 35.Rxf1 Rxf1+ 36.Kxf1 36...hxg5 with a won Pawn end-game. White's next move practically does not change anything.

31.Kh1 Rxf3! 0-1 Alekhine: 'With the same point as mentioned above.' To play through the complete game see...

Gideon Stahlberg vs Alexander Alekhine, Hamburg ol (03) 1930
http://www.chessgames.com/perl/chessgame?gid=1012664

...on Chessgames.com.