It is a little-known fact that the winner of the first Soviet Championship in 1920 was none other than Alexander Alekhine. The contemporary title of the event was the First All-Russian Chess Olympiad, and it was only after 1922 – when the Soviet Union was formally established – that it was subsumed as part of the continuing series. Alekhine, of course, left his homeland in 1921, never to return. After his match victory over Capablanca in 1927, he reportedly made some anti-Soviet statements to a gathering of emigre Russians in Paris. His former compatriots immediately declared him persona non grata, and he was not rehabilitated until after the death of Stalin.
Along the way to winning the title in 1920, Alekhine had to survive a very difficult encounter with Grigory Levenfish (who won the title himself in 1937). Today we present an extract from Sergei Voronkov's 2007 book Masterpieces and Dramas from the Soviet Championships, 1920-1937. We pick up the action after Levenfish's 31st move ...Kg8-h7. From now on the commentary is by Voronkov.
“The only way to draw,” exclaims Alekhine [in his 1921 book Schachleben in Sowjet-Russland - tr]. “If first 32.Qd3+ g6 and now 33.Kg4, then 33...f5+ 34.exf6 h5+ and wins.”
Alexander Kotov in his book Alekhine's Chess Heritage (1982) shows exactly how: “After the only move 35.Kh4! (35.Kg5 Rc5+) the win is achieved as follows: 35...Qxf6+ 36.Kh3 Qb2! 37.Kh4! Kh6! 38.Rh8+! Kg7!”
Go ahead and laugh, but this is just not true! You don't believe me?
If we extend the variation with the five exclamation marks for one more move – 39.Qd8! – the draw becomes obvious: 39...Rc8 40.Rg8+ Kh7 41.Qxc8 Qxh2+ 42.Kg5 Qxg3+ 43.Kf6 Qxf4+ 44.Ke7 Qe4+. Or 39..Qf6+ 40.Qxf6+ Kxf6 41.Rf8+ Ke7 42.Ra8 Rxh2+ 43.Kg5 Rg2 44.Rxa7, etc.
What's more, the alternative move 35.Kg5, which must seemingly be rejected in horror due to the “fatal” check 35...Rc5+, is in fact winning for White:
36.f5 Rxf5+ 37.Qxf5!! (and not 37.Kh4 Qxh2 mate) 37...gxf5 38.Rd7+ Kg8 39.Kg6! and it suddenly becomes clear that Black can avoid immediate mate only by giving up his queen! (By the way, I was very proud of my discoveries until I found out they had already been made in 2002 by the trainer L. Veretnovy from Krasnoyarsk.)
The most striking thing of all is that Alekhine's variation is doubly wrong. In the first place, as we have seen, it leads not to a win, but to a loss! In the second place, after 32.Qd3+ g6 33.Kg4 f5+, Levenfish in annotating the game in the Newsletter of the Petrograd Commune had already indicated a simple drawing line: 34.Kf3! Rc3 35.Rd7+ Kh8 36.Rd8+ with perpetual check. However, he too thought that 34.exf6 would lose on account of 34...h5+.32...Rxh2
“Such a pity!” is Levenfish's lament in his book Selected Games and Reminiscences. “By continuing 32...h5+ 33.Kxh5 (33.Kxg5 f6+) 33...Rxh2+ 34.Kg4 Qg2 35.Qd3+ (if 35.Rd7 then 35...Rh6!) 35...g6 36.Rd7 (or 36.Kg5 Kg7 37.Rd7 Rh5+ 38.Kg4 Rh4+! and mates) 36...Kh6! 37.Rxf7 Rh4+! 38.Kxh4 Qh2+ 39.Kg4 Qh5#, Black would have brought this combative game to a beautiful conclusion.”
Indeed, it's a beautiful conclusion. Why then did Alekhine, in citing the alternative variation 32...g6 33.Rh8+! Kxh8 34.Qd8+ Kh7 35.Qe7! with a draw, restrict himself to the laconic comment, “If 32...h5+ then simply 33.Kxh5!” Was he mistaken yet again?
No, this time his analysis was accurate. It turns out that White is not obliged to play 36.Rd7? (also bad is 36.Rf8? Qh3+ 37.Kf3 Rg2 38.Rxf7+ Kh6 39.Ke4 Rxg3). One of Levenfish's parenthetical variations in fact leads to a draw: 36.Kg5! Kg7, and as before not 37.Rd7? but 37.Qa3! Rh5+ 38.Kg4 Qd2+ (or 38...Kh7 39.Rh8+ Kxh8 40.Qf8+ Kh7 41.Qxf7+ Kh6 42.Qf8+) 39.Qf3 Qh2 40.Qa3 with a repetition of moves.
33.Qd3+ g6 34.Rd7! Kg7
If 34...Rh4+? then 35.Kf3!
With this we conclude our extract from Voronkov's book.