The Players: Alexander (1909-74) was
born in Cork, Ireland, but lived nearly all his life in England. He was one of several chess players involved in the WW2 ‘Ultra’ codebreaking operation at Bletchley. When its Cold War successor, GCHQ, was established in Cheltenham, Alexander moved there; he and several colleagues were the nucleus of strong Gloucestershire teams of the 1950s and 1960s. He was twice British Champion and a FIDE IM as well as an excellent writer on the game. In the last decade of his life he concentrated on postal chess and earning the ICCF IM title in 1970, playing on England’s olympiad team.
Clarke (born 1933) wrote the first books in English on Petrosian and Tal. A British Master OTB, he played many times for England; he had a reputation for being an extremely hard player to beat. In the 1970s he took up CC more seriously and obtained the ICCF IM title (1976) and then GM (1980) before giving up the game for health reasons.
About this game: It first appeared
in ‘Gloucestershire Correspondence Chess 1954-81’, which records the feats of that team in the annual inter-county competition organised by the British Chess Federation. It
was also included in the collection of Alexander’s games, edited by Gol- ombek & Hartston. Alexander himself was probably the source for most of the lines cited in both books.
1 e4 e5 2 Èf3 Èc6 3 ƒb5 a6 4 ƒa4 Èf6 5 0–0 ƒe7 6 d4
Nowadays this variation is rare in master chess because it lacks strategic depth compared with the 6 …e1 line. In the pre-computer era, however, such sharp opening variations often paid off. First, the opponent’s opening knowledge would be tested, and then his analytical abilities. In this game, Clarke passes the first examination but fails the second.
6...exd4 7 …e1
This move allows Black to castle and retreat his È to e8, but on the principle that you cannot have an attack without development, it gives more prospect of a lasting initiative than 7 e5 Èe4.
7...0–0 8 e5 Èe8
8...Èd5 9 Èxd4 Èxd4 10 †xd4 Èb6 11 ƒb3 d5 12 exd6 used to be thought good for White, but the reputation of the move has improved. Instead of 12...ƒxd6 13 ƒf4 or 12... †xd6 13 †e4 (£ƒf4‹), ‘ECO’ gives 12...ƒf6! 13 †e4 cxd6! 14 Èc3 ƒxc3 15 bxc3 d5 16 †e7
Game 27: Alexander-Clarke 141 †xe7 17 …xe7 Èc4= Feistenauer-
Donev, Götzis 1990.
9 c3!?
9 ƒf4 is a more solid alternative.
9...dxc3 10 Èxc3 d6 11 exd6 Èxd6 (D)
XIIIIIIIIY
9r+lwq-trk+0
9+pzp-vlpzpp0
9p+nsn-+-+0
9+-+-+-+-0
9L+-+-+-+0
9+-sN-+N+-0
9PzP-+-zPPzP0
9tR-vLQtR-mK-0
xiiiiiiiiy
This is now the book main line; White is challenged to justify his gambit. Others, according to the theory of those days:
a) 11...†xd6 12 Èd5!Ÿ Castagna- Limbos, Varna OL 1962.
b) 11...cxd6 12 Èd5 (Kholmov- Lein, 29th USSR Ch 1961), but later
this was shown to be playable with 12...ƒe6 13 Èxe7+ †xe7 in Mini™- Masi™, Yugoslav Ch 1972.
c) The Alexander book said “11...ƒxd6 is safest for Black when 12 ƒg5 Èf6 13 Èe4 ƒe7 leads to a very drawish ending”.
12 Èd5
12 ƒf4 was preferred a few years later, e.g. 12...b5 13 ƒb3 Èc4 14 Èd5! ƒd6 15 ƒg5 †d7 16 …e4 È6e5 (16...f6 17 ƒf4 ƒxf4 18 …xf4 …d8 19 †e2!‹ Romanishin- Tukmakov, 46th USSR Ch, Tbilisi
1978.) 17 Èxe5 Èxe5 18 †e2! Èg6 19 Èe7+! Èxe7 20 ƒxe7 ƒxe7 21 …xe7 †c6 22 …d1! ƒb7 23 ƒd5 †b6 24 ƒxb7 †xb7 25 …dd7 …ac8? (25...†b6 — Gipslis) 26 †e6! †b6 27 …xf7! ‡h8 28 †e7! 1–0 TV viewers-Radio listeners, Latvia corr 1978; now if 28...…xf7 29 …d8+! or 28...…fe8 29 …f8+! …xf8 30 †xg7#.
12...…e8
12...ƒe6¢ is theory today, following Timman-Beliavsky, Moscow 1981.
13 Èe5
Alexander did not want to capture on c6 — obtaining positional comp- ensation through Black’s split pawns — because he believed it would lead to a drawn endgame at best. I am sure he knew that Clarke would conduct a positional game very ably and so preferred to rely on his tactical powers to exploit what should have been only a temporary initiative.
13...ƒd7 14 Èxd7 †xd7 15 ƒf4 ƒf8 16 †d3 …xe1+ 17 …xe1 …d8
Clarke hopes for counterplay on the d-file rather than simplification by 17...…e8 18 …xe8 †xe8 and then:
a) 19 Èxc7 †e1+ 20 †f1 “regains the pawn when the ƒ pair gives White the better prospects”, according to Alexander. Computer analysis doesn’t support that view as Black continues 20...†e4 21 ƒxc6 †xf4å, e.g. 22 ƒf3 Èf5 23 Èd5 †d2 24 †b1 ƒc5.
b) 19 ‡f1 is better, e.g. 19...†e4 20 †xe4 Èxe4 21 ƒxc6!? bxc6 22 Èxc7 Èc5 23 ‡e2Ÿ K.Howard- A.Tankel, British CC Ch candidates 1987.
18 ƒc2 f5 19 …d1 ‡h8 (D)
64 Great Chess Games 142
XIIIIIIIIY
9-+-tr-vl-mk0
9+pzpq+-zpp0
9p+nsn-+-+0
9+-+N+p+-0
9-+-+-vL-+0
9+-+Q+-+-0
9PzPL+-zPPzP0
9+-+R+-mK-0
xiiiiiiiiy
20 g4!Here the Alexander book explains: “White appears to have a great deal of pressure but it is another matter finding something to do... Thus this thrust is really the only attempt to increase the pressure”. White has a variety of moves that might be marginally better positionally (20 a4 or 20 a3 for example), but which would offer less hope of victory; again the note to White’s 13th move
applies.
20...g6?!
“Having defended well, it was time for Black to be thinking more actively,” say Golombek & Hartston, while the Gloucestershire book here says, “Better 20...†f7 — the weakening of the long diagonal proves fatal”.
Indeed 20...†f7! was much better, when:
a) The Alexander book correctly says that 21 gxf5! (21 ƒg5! …e8) 21...Èb5! gives Black some initiative through the threats of ...Èb4 or ...Ècd4. 21...†h5 also comes into consideration.
b) Alexander believed that “White
can still regain his pawn and draw by 21 ƒxd6! cxd6 22 †xf5 and this may be the best he has”. Actually, White may be slightly better in that case but 21...cxd6 seems the worst of the three recaptures and after both 21...ƒxd6 and 21...…xd6 Black’s position is certainly not inferior.
Black’s mistakes here and at the next move are forgiveable, however. They stem from the same source: Clarke’s failure to see White’s stunning 22nd move.
21 ƒg5 …c8?
Black should have played 21...…b8 with fairly good chances of escaping with a draw as White’s next move would then not attack the …. Alexander could then choose between:
a) 22 Èe7 Èe5 23 †c3 ƒg7 24 gxf5 and now the computer world champion program Shredder6 comes up with the defence 24...…e8 (If 24...gxf5 25 ƒxf5 †e8 26 …xd6!) 25 f4 Èg4 26 Èxg6+ hxg6 27 †h3+ ‡g8 28 †xg4 gxf5 29 ƒb3+ ‡f8=.
b) 22 ƒf6+ ƒg7 23 g5 could result in a draw after 23...ƒxf6 24 Èxf6 †e7 25 ƒa4 Èe4 26 ƒxc6 bxc6 27 Èxe4 fxe4 28 †d4+ ‡g8 29 †c4+ ‡h8= (analysis with Shredder6).
22 Èe7!! (D)
This is the sort of counter-intuitive shot that required imagination and calculating skill before the era of the chess engines.
“A beautiful and totally surprising combination,” wrote Hartston; “a bolt from the blue which wins in all variations,” says the Gloucestershire book.
Game 27: Alexander-Clarke 143
XIIIIIIIIY
9-+r+-vl-mk0
9+pzpqsN-+p0
9p+nsn-+p+0
9+-+-+pvL-0
9-+-+-+P+0
9+-+Q+-+-0
9PzPL+-zP-zP0
9+-+R+-mK-0
xiiiiiiiiy
22...…e8Alexander demonstrated wins against the alternatives.
a) 22...ƒxe7 23 †c3+ ƒf6 (23... ‡g8 24 ƒb3+ ‡f8 25 †h8#) 24 ƒxf6+ ‡g8 25 ƒb3+ ‡f8 26 …e1 Èf7 27 ƒg7+ ‡g8 28 ƒh8!ˆ. b) 22...Èxe7 23 ƒf6+ ƒg7 24 ƒxg7+ ‡xg7 25 †c3+ ‡h6 26 g5+ ‡xg5 (26...‡h5 27 f4) 27 f4+ ‡h6 28 …d3 and he has no defence to the threat of …h3#.
c) 22...Èe5 23 †c3 ƒg7 24 Èxc8 also wins for White, e.g. 24... †xc8 (or 24...Èxg4 25 Èxd6 ƒxc3 26 Èxf5! †xd1+ 27 ƒxd1) 25 …xd6 Èf3+ 26 †xf3 cxd6 and White emerges a piece up.
23 ƒf6+ ƒg7 24 ƒxg7+!
White prefers to play for mate; there is still one more surprise in store. The Alexander book says 24 Èxg6+ hxg6 25 †h3+ ‡g8 26 ƒb3+ ‡f8 27 †h7 allows Black to fight on with 27...ƒxf6 28 †xd7 …e7! 29 …xd6 cxd6 but I think White is winning that.
24...‡xg7 25 †c3+ ‡f7 26 g5!!
A very pretty echo of the 22nd
move; the Gloucestershire book points out that “the È is now en prise to four
pieces but cannot be taken by any of them, as analysis shows”.
26...…d8
The Gloucestershire book says that “the plausible 26...…f8 would be brilliantly refuted by 27 Èxc6 bxc6 28 †f6+ ‡e8 29 …e1+ Èe4 30 ƒxe4! …xf6 31 ƒxc6+! and wins”, while 27...†xc6 28 †f6+ ‡e8 29 …e1+ Èe4 30 †xc6+ bxc6 31 f3ˆ is a variation from the Hartston & Golombek book.
Other finishes worked out by Alexander were: 26...‡xe7 27 †f6#; 26...†xe7 27 ƒb3+ ‡f8 28 †h8#; 26...Èxe7 27 ƒb3+ˆ; and 26...…xe7 27 ƒb3+ …e6 28 †f6+ ‡e8 29 ƒxe6 †d8 30 †h8+ ‡e7 31 ƒg8!ˆ.
27 Èxc6 †xc6
27...bxc6 leads to long forced mating variations after either 28 ƒb3+ (or 28 †f6+) 28...‡e8 29 †h8+ ‡e7 30 †e5+ ‡f8 31 …xd6ˆ cxd6 32 †h8+ ‡e7 33 †g7+ ‡e8 34 †g8+ ‡e7 35 †f7#.
28 ƒb3+ ‡e8
Clarke prefers a pretty finish.
29 †e5+
The Gloucestershire book says Black resigned here, giving the remaining moves in a note, but the book of Alexander’s games says that Clarke “surely felt it would be churlish to deny his opponent the pleasure of giving mate”. I suppose the most likely explanation for this discrepancy in the accounts is that the remaining moves were sent as a conditional.
29...‡f8 30 †f6+ ‡e8 31 …e1+ Èe4 32 †f7# 1–0