An amusing trap which turns up in many tactics books is one that world
champion Emanuel Lasker fell into during a simultaneous exhibition (diagram
1.)
Lasker had just executed a petite combinaison
himself, sacrificing a rook on f8 followed by g6xh7, queening the pawn.
But after 1...Rc3+ he blundered with 2.Kg4? (2.Kf2 etc.,
using ranks 1, 2 and 3 to approach the rook wins), and now 2...Rc4+
3.Kg5 Rh4! 4.Kxh4 g5+ 5.Kxg5 Kg7 and Black won.
It is quite likely Loman found his move over the
board, but this particular square-vacating anti-promotion sacrifice
had already been shown by the great Troitzky in 1899. (Diagram 2;
you can click #1 to go there.)
1.h3+ As 1...Kxh3 2.Nf4+ and 1...Kf3 2.Rf7+
etc. lose directly, there are two variations:
a) 1...Kh4 2.Rd4+! Bxd4 3.Nxd4 el=Q+ 4.Kg2
and the threat 5.Nf3 mate costs Black his queen.
b) 1...Kf5 2.Nd4+ Bxd4 3.Re7 Be5 4.Rxe5+!
Loman's move. 4...Kxe5 5.f4+ followed by 6.Kf2 and White wins.
There is a slight blemish in this beautiful endgame
study: in variation a), which is really the heart of it, a transposition
of moves is possible. Instead of 2.Rd4+ White can also play 2.Kg2, and
Black has nothing better than el=Q 3.Rd4+ etc., leading back into a). In
1909, Troitzky published a repaired version which later proved to have
even greater defects, and he finally left it out of his collected studies.
Beside Lasker-Loman, I know only two examples of Loman's move from actual
play.
In diagram 3 there followed 1...Rhl+ 2.Kxhl gxf2 3.Rf5!
etc. and White won.
The little known position from diagram 4
occurred during a day of blitz play. Black thought he could do something
better with his Pg3 than to use it as a bind with which to draw. 1...Qf2?!
2.Rxf2 gxf2 He must have thought this was a win. 3.Rxf5! and
Shishko resigned: the pawn is stopped after 3...Kxf5 4.g4+ and 5.Kg2, and
White wins.
The two opponents were fellow pupils at Moscow's
first school for film actors where Alekhine, according to Shishko, failed
miserably.
Although a rook seems to be indispensable to Loman's move, it can be
executed by other pieces as well. (See diagram 5).
How is White going to stop Pa2 from queening? 1.Bc5+!
1.c7? Nxc7 or Kd7 and White has nothing. 1...Ke5 After 1...Kxc6
2.Nb4+ or 1...Nxc5 2.c7 Kd7 3.Nxc5+ etc. White has no problems. 2.c7
Nxc7 3.Bd4+! Kxd4 (see diagram 6)
4.Nc5! Kxc5 5.b4+ Kxb4 6.Kb2 draw.
A logical extension of the theme was to construct the sacrifice and the ensuing square-vacation as an insufficient means of defence, as in the undeservedly unknown composition of diagram 7, which looks like it was directly inspired by Lasker - Loman. 1.c6! Rxc6 If Black does not take the pawn, White wins easily with 2.Rc3, 3.a4 and a march of the b-pawn. 2.Rf8+ (Lasker's move) 2...Kxf8 3.gxh7 Rc2+ 4.Kg3 Rc3+ 5.Kg4 To make progress, White must allow Loman's move. 5...Rc4+ 6.Kg5 Rh4! 7.Kxh4 g5+ 8.Kxg5 Kg7 9.h8=Q+! Kxh8 10.Kg6 d5 (or e5) 11.a4 and Black queens first, but White mates immediately afterwards.
Much better known is the antidote to Loman's move of diagram 8
- among miniatures, it is a classic.
1.b7? Kc6!, but there is another way: 1.Rd8+!
Kxd8 2.b7 And now Loman. 2...Rb4(!) 3.Kxb4 c5+ 4.Kb5! Kc7 But
now Moravec. 5.Ka6! Kb8 6.Kb6 c4 7.a4 and 10.a7 mate.
It took forty years for somebody (an ex-over the
board champion of France) to show a further elaboration; a counter to Moravec's
counter to Loman's counter to Laskers promotion combination. See diagram
9.
1.b5+! Kb7 Or cxb5 2.Rxb5 and the file b5-b1
will be available anyway. 2.bxc6+ Kxc6 3.Rb5! Kxb5 4.c4+ Kb4 5.Kc2 Ka3
6.Kbl Kb3 And now Moravec? No: 7.c5! The crucial move. Now White
will queen with check and win, either after 7...a5 8.cxb6 or 7...bxc5
8.g5 etc.
Loman (1861-1932), a Dutchman who made his living as an organ player and music teacher in London, wasn't just a simul anonymous. He often visited his native country and in the pre-Euwe years he was champion of the Netherlands many times.
A related square-vacating anti-promotion combination is best introduced
by a near-tragedy that happened in diagram 10.
The IBM 1971 was Timmans first grandmasters tournament,
and he had started disastrously, losing his first five games. In round
6 however, three pawns up, he was about to change that. 49.Rg6+ followed
by 50.Ra6 would have been easiest, but Timman saw something much more interesting.
49.f6
a3 50.Rg8?? With 50.e8=Q Rxe8 51.f7 Rf8 52.Rg6, a near-win was still
there. 50...Rxg8 51.f7 (see diagram 11)
51...a2?? Luckily for Timman, Ivkov saw it too. After 51...Rxg2+
52.Kxg2 Kxe7 White would have had to resign at once. Now, after 52.fxg8=Q
al=Q+ 53.Kh2 Kxe7 54.Qg5+, it was Black who resigned. Neither player
had noticed they had crossed an abyss twice. Only later, when somebody
asked Timman: 'Your king was on h1, was it - it wasn't?' he realised
what had happened.
The point of 51...Rxg2+ is that the rook, moving
away with check, vacates the square where the pawn threatened to capture,
and since pawns cannot move to such squares, White would have lost a vital
option. The difference with Loman's move is that there, access is gained
to the square vacated while here, access to that square is denied
to the opponent. This combination is sometimes termed the 'Prokeš-manoeuvre',
after a study by the Czech composer Prokeš (pro-kèsh; you never
know what cyber-age will do with the s-hacek) in which the idea was first
shown, and in its simplest form.. (Diagram 12.)
1.Kg4 e2 (1...d2 2.Kf3 Kd3 3.Ral e2 4.Ra3+
Kc2 5.Ra2+ Kcl 6.Ral+ Kb2 7.Kxe2 etc.) 2.Rcl+ Kd4 (2...Kb3 3.Kf3
d2 4.Rbl+ Kc2 5. Kxe2. Or 2...Kd5 3. Kf3 d2 4. Rc5+!) 3.Kf3 d2 (see
diagram
13)
4.Rc4+! Kd3 5.Rd4+ Kxd4 6.Kxe2 Kc3 7.Kdl Kd3 stalemate.
It is interesting that before composing his original Prokeš-manoeuvre,
Prokeš had already composed an 'anti-Prokeš'-manoeuvre: see diagram
14.
1.f6! gxf6 2.Nf8! Rxf8 3.e7 winning, as the
rook cannot leave f8 with check. Without 1.f6, it could have.
The try 1.Ng5 Kd6 2.Nf7+ Ke7 3.d8=Q+ Rxd8 4.Nxd8
Kxd8 5.Kf2 doesn't work as Black easily draws after Ke7, g6 etc. He would
also have this draw without the Pa7, which is a point as Prokeš did originally
publish the study without this pawn. But as John Nunn pointed out, White
then does have an alternative win in 1.Ng5 Kd6 2.Nf7+ Ke7 3.Ne5! The threat
Nc6+ forces 3...Kd6 or Kd8, and after 4.Kf2 Black is left with useless
rook moves along his first rank; 4...Kxe5 or Kc7 loses to 5.e7, and 4...Ke7
to 5.Nc6+. Black would be helpless against the king's march f2-g6 etc.
The added pawn at a7 seems to save the study. Now,
if 1.Ng5 a5! 2.Nf7 a4 3.d8=Q Rxd8 4.Nxd8+ Kc7 5.Nb7 (5.e7 Kd7 6.Nc6 a3
draws) 5...a3 6.Na5 a2 7.Nb3 Kd6 8.Kf2 g6 and draw.
The Prokeš-manoeuvre can also occur one row away from actual promotion,
as in these two game examples where it was missed, perhaps demonstrating
that it is difficult to see.
In diagram 15,
White had just played c2-c4, threatening an unusual queen's catch.
Black's move makes clear he saw this:
1...f4?? 2.Ba3 fxe3 Unveiling
his trap: 3.Bxc5 exd2 4.Ral dl=Q etc. A stupid trap at that, because it
would lead to a bad endgame. What he also hadn't seen was the Prokeš-check:
3.Rd8+ (1-0) White remains a rook up.
In diagram 16,
it is quite likely Black, like Timman, had just sacrificed a rook,
in this case on b2. After 1...c3 2.Kxd3? cxb2 3.Kc2 Kd4 4.Kxb2 Ke4 5.Ke3
Kxf4 6.Kb4 Kxg4 a draw was agreed. Using the Prokeš-manoeuvre, White
could have won a few useful tempi and the game: 2.Rb5+! Kxb5 3.Kxd3 c2
4.Kxc2 Kc4 5.g5 Kd4 6.f5 Ke5 7.f6 gxf6 8.gxh6 etc.
As with Loman's move, endgame composers have brilliantly elaborated
on Prokeš' idea.
In diagram 17, play starts with a battle
against Black's passed pawn and his two potential passed pawns. 1.Rhl
(1.dxc4+? Ke4 2.Rhl f3 3.Kg3 Kxe3 and Black wins) 1...exd3 2.Kf3 Ncl!
Prokeš' knight's move we have already seen in diagram 14. See diagram
18. Here it threatens 3...d2, which would also be the answer to 3.Rxc1.
3.c6!
Now Rxcl is threatened because after 4...d2 White could give the Prokeš-check
on Rc5+. In the next few moves White and Black alternately try to open
and close a square for this thematic check. 3...Kd6 (3...d2 4.cxd7
elQ 5.d8Q+ Kc4 6.Qd4+ Kb5 7.Qb4+ and 8.Rxel) 4.cxd7 Kxd7 5.Rxcl!
(the Prokeš-check is available now) 5...d2 6.Rxc7+ Kxc7 7.Kxe2 e3 8.f6
Kd6 9.a4 etc. and White wins.
David Gurgenidze (b. 1953), one of those dashing young Georgian composers
who count among the greatest of the modern romantic school, has some very
beautiful Prokeš-manoeuvres to his name.
In his study of diagram 19, White plays 1.g7+
Kg8 2.Nh7, and the double threat of Nf6 mate and gxf8=Q can only be
parried if Black can move his Rf8 out of reach of the Pg7 by checking.
2...Ra5+!
This might be termed a pre-Prokeš sacrifice. If it is accepted the Prokeš-rook
breaks loose: 3.Kxa5? Ra8+ followed by 4...Kxh7 and Black wins. 3.Kb4
(3.Kb6 Ra6+ 4.Kxb7 Rb8+ etc.) 3...Ra4+ 4.Kb3 Ra3+ 5.Kb2 Ra2+ 6.Kcl Rxc2+
7.Kb1 (7.Kd1 Rc1+ and the rook will check on the c-file forever, drawing,
as White cannot enter the e-file on account of Re8+) 7...Rc1+ 8.Kb2
Rc2+ 9.Kb3 Rc3+ 10.Kb4 Rc4+ 11.Kb5 Rc5+ 12.Kb6 and Black's pre-Prokeš-rook
is finally muted; White wins.
A year later, this proved to have been only an exercise for a more ambitious
work (diagram 20).
1.Rh7+ (1.Rxf7? Rc8! 2.g7+ Kg8 3.g6 Rc5+
4.Kb4 Rb6+ 5.Ka4 Ra6+ with perpetual check) 1...Kg8 2.g7! Ra5+ 3.Kb4
(3.Kb6 Ra6+ etc.) 3...Ra4+ 4.Kb3 Ra3+ 5.Kb2 Ra2+ 6.Kcl Rxc2+ 7.Kd1 Without
the Pd5, running back to b6 is no use. 7...Rc1+ 8.Kd2 Rc2+ 9.Kd3!
(9.Kel Rxe2+ is a draw, see move 25. White must first remove the pawn at
d6) 9...Rc3+ 10.Kd4 Rc4+ 11.Kd5 Rc5+ 12.Kxd6 Rc6+ 13.Kd5 Rc5+ 14.Kd4
Rc4+ 15.Kd3 Rc3+ 16.Kd2 Rc2+ 17.Kel Rxe2+ 18.Kfl Rel+ 19.Kf2 Re2+ 20.Kf3
Re3+ 21.Kg4 Re4+ 22.Kf5 Re5+ 23.Kxf6! (and he must remove the Pf6)
23...Re6+
24.Kf5 Re5+ 25.Kxe5 Finally. 25...Re8+ 26.Kf5! (26.Kf6? Re6+
and 27...Kxh7) 26...Kxh7 27.Kf6 Kg8 28.h7+ Kxh7 29.Kxf7 Re7+ 30.Kxe7
Kxg7 31.h6+ and wins.
In diagram 21, Gurgenidze surpassed himself, and, to my taste, almost everything else in endgame studies. The key move, as so often in studies, is obvious: 1.hxg7 Now Bxf7 is impossible because of gxh8=Q but without the rook on h8 it would be a good defense. Again, to obtain the necessary Prokeš-check for his Rh8, Black offers a pre-Prokeš-rook. 1...Rg2+! 2.Kf1! Why 2.Kxg2 doesn't work will be clear around move 19. 2...Rf2+ 3.Kel Re2+ 4.Kdl Rd2+ 5.Kcl Rc2+ 6.Kbl Rb2+ 7.Kal Rxa2+ 8.Kbl Rb2+ 9.Kcl Rc2+ 10.Kdl Rd2+ 11.Kel Re2+ 12.Kfl Rf2+ 13.Kgl Rg2+ (see diagram 22) We are back at move 1 - only without the Pa2. 14.Kxg2! Finally White allows Black to vacate h8. 14...Rxh2+ 15.Kxh2 Bxf7 16.gxf7 Kxf7 17.Nxd7 b3 (17...a3 18.Nc5) 18.Ne5+ Kxg7 19.Nc4 and at last we see what White has been doing. The condition on which he allowed Black to execute his Prokeš-manoeuvre was getting rid of his Pa2 - a Prokeš-manoeuvre in itself. With a2 vacated, the black pawn, long before it arrived on b3, has been robbed of the double move it would have had there. With the pawn still on a2, Black could now have played bxa2. However: a2 being empty, all he has is 19...b2 20.Na3 and White wins.
Prokeš (1884-1966) was one of the great composers of his day. He was also a strong player who, in the twenties and thirties, often represented Czechoslovakia in Olympiads.
Thus the story on square-vacating anti-promotion combinations appeared
in my book Chess Curiosities in 1985. As it turned out, there were several
PS's.
When I first worked on this story, in 1982 for Schaakbulletin,
I had already been infected by the Prokeš theme and, trying to compose
one of my own, I had found an outline for a double Prokeš - something I
had not come across (see diagram 23).
The idea was: 1.Rh1 1.Kf2 e3+ and White loses
his Rh2 1...e3 2.Kf3 Rc1 3.Rxc1 exd2 4.Rc4+ bxc4 5.Kxe2 c3 6.Rf1 Rb1
7.Rxb1 c2 8.Rb4+ Kxb4 9.Kxd2 draw. That was very schematic and incorrect
at various points, and I hadn't been able to make something worthwile out
of it. And in '85, much to my regret, I had to conclude Chess Curiosities
without a Prokeš of my own.
I forgot about it for many years.
Then, in the summer of '92, I visited Jan Timman
in the south of France, near St. Tropez. While the women shopped for bouillabaisse
ingredients and the children played monopoly or drowned in the swimming
pool, we sat over a chess board. Misha Tal had died recently, and for an
article in memory of him, Timman was analysing his famous game against
Keller, Zurich 1959. I tried to help with suggestions and maybe did, but
mostly, I was just admiring the thoughts Tal had had, and the thoughts
Timman had about those thoughts now.
After a while, as Timman is a top-notch composer
as well, we turned to talking chess studies, and I tried to remember my
old double-Prokeš outline, but did not really manage. For the next two
days or so, ('Jan! Tim! Are you coming? Dinner is ready!') we tried to
find a setting and build something from it, but failed. Or, as Timman later
wrote in New in Chess (1994-3): 'Tim Krabbé told me about the double
Prokeš. I am afraid that since his peroration on the subject, my insight
in chess has never been the same. (...) It turned out to be exceedingly
difficult to find the correct technical background. In the South of France
we didn't make much progress either, but my diary entries of that time
seem to suggest that looking for a satisfactory double Prokeš version could
beat any sleeping pill.'
Home again, I found my old double Prokeš-outline
of diagram 23 in my papers, and gave it to Timman. Then nothing happened
for months.
But one day that autumn, Timman called me excitedly
- he had something to show me - a full-fledged double Prokeš. See diagram
24.
As always with Timman, it was an appealing, not
too-unnatural position. White is a piece up, but most of his pieces are
placed awkwardly, and the black pawns are dangerous. After 1.exd3 Rxd2+
Black has a perpetual; 1.Rf1 dxe2 costs at least a piece, and 1.e3 Bxe3
(or Rxd2+ first) isn't very attractive either. And on 1.Kf1 Rc1+ Black
can force a repetition if not more. After 1.Rh1 Rxd2 the two black center
pawns are worth at least a piece, e.g. 2.Re1 Rxe2+ 3.Kf1 Bd4 4.Rf5 Rxe1+
5.Kxe1 e3 and Black wins. On 1.Kh3 Rxd2 is good, but even better is 1...e3,
for instance 2.exd3 exd2 3.Rh1 Rc1 4.Rff1 Rxf1 5.Rxf1 Rxf8
The solution begins with: 1.Bc5! White wants
to trade his hanging bishop for blacks strong one. 1...dxe2 Also
dangerous for White is 1...e3 Then 2.dxe3 dxe2 and 2.exd3 exd2 would both
lose, as would 2.Bxa7 exd2 3.Rh1 dxe2 4.Rff1 d1=Q of exf1=Q+. The only
winning move is then the surprising 2.Rh3! and now exd2 3.Rxd3 or dxe2
3.Rxe3 2.Rh1 The obvious 2.Kf2 doesn't work because of e3+ and White
loses his Rh2. 2...Bxc5 3.bxc5 e3! The best chance. Suddenly it
seems as if White has nothing against the threat exd2 But: 4.Kf3! (see
diagram
25.)
Now White threatens to capture on e3, and 4...exd2
5.Kxe2 doesn't work because e8 is controlled (5...Rc1 6.Rd1 or Rff1.) 4...Rc1!
The theme, finally - exactly as in Timman - Ivkov. 5.Rxc1 exd2 6.Rc4+!
bxc4 7.Kxe2 c3 Threatening Rb1. 8.Rf1 And not 8.Bf7 Rb1 9.Bb3+
Ka3! 10.Rf1 Rxf1 11.Kxf1 Kb2 and Black wins. Now (see diagram 26),
it looks once more as if Blacks is finished, because 8...c2 is answered
by 9.Kxd2 Rb1 10.Rc1 But of course: 8...Rb1! Again - the double
Prokeš. 9.Rxb1 Forced; 9.Rd1 is answered by c2. 9...c2 10.Rb4+!
Kxb4 11.Kxd2 and White wins.
I published this in my chess column, Timman prepared to use it for an
article about the Prokeš-manoeuver but then, as so often happens with endgame
studies, it was cooked. When Timman showed it around in Linares a few months
later, Speelman found that with 8...Ka3!, Black can at least draw.
Timman writes: 'I was so disappointed that I dropped
the whole thing. The time I had invested in it, the racking of my brain!
Maybe finding a totally correct, good double Prokeš was just beyond me,
I thought resignedly. But shortly before Christmas 1993 I decided to give
it one more go. (...) On December 27th I finally managed to get a more
or less satisfactory version. It is still a co-production, as Krabbés
skeleton is still visible after the first move.' (See diagram 27.)
After 1.Bh6 (1.Bh8 dxe2 2.Rh1 Rh7) the solution
was the same: 1...dxe2 1...e3 2.Rh3! 2.Rh1 2.Kf2 e3+ 2...e3
3.Kf3 Rc1! 3...exd2 4.Ke2 Re7 5.Re6 4.Rxc1 exd2 5.Rc4+ bxc4 6.Ke2
c3 7.Rf1 Rb1! 7...Re7 8.Kd1 Ka3 9.Bg5 and 10.Bf6 8.Rxb1 c2 9.Rb4+!
and wins.
Even that wasn't the end of it yet - the Prokeš manoeuvre still fascinated
Timman, and some time later an idea for a very artistic (single) Prokeš
occurred to him. And after months of failed attempts, he finally created
the beautiful study of diagram 28. He dedicated it to Misha Tal,
because the initial idea had occured to him during the 1993 Tal Memorial
Tournament in Moscow, and also because his interest in the theme had been
born the day we had looked at Tal - Keller.
White is a rook up, but the Black passed pawns seem
to be enough for a draw at least. 1.Ke2 Rb1! Because c2 2.Kxd2 Rb1
3.Kxc2 Rxa1 4.Ra8 is hopeless. 2.Rxb1 c2 What now? 3.Rd1 c1=Q 4.Rxd2
Qc4+ is only a draw. White can only win if he creates a checking opportunity
for his threatened rook. 3.Rxg6! Threatening Nf4 mate. 3...fxg6
Of course not 3...Kxg6 4.Rb6+ Now, tempting is 4.g4+ Kxh4 5.g3+ Kxg3 and
White has the check he has been fighting for, but it has been too expensive:
6.Rg1+ Bxg1 7.Kxd2 Be3+! 8.Kxc2 fxg4 9.Ng1 Kf2 10.d6 Kxg1 11.d7 Bg5 12.a6
g3 etc. and Black draws. Therefore: 4.Nf4+ Kg4 (See diagram 29.)
5.Nh5!! The most beautiful move of the study.
White threatens Rb4+ but also Nf6 mate. 5...Bd4 is impossible because of
6.Nf6+ anyway, followed by Rb4+, but 5...f4 is now Blacks best defence.
White then proceeds with 6.Nf6+ Kf5 7.g4+ Ke5 8.Nd7+ Kd6 9.Rb6+ Bxb6 10.Kxd2
Bxa5+ 11.Kxc2 Kxd7 when, according to Timman, he has a technical win with
12.Kd3 and 13.Ke4. As this is not thematical, the main variation is: 5...Kxh5
6.g4+ fxg4 Even now, White's intentions are not clear. Where is the
check? 7.Rh1! The final finesse: not only a check, but a mating
threat too, can be a Prokeš-manoeuvre 7...c1=Q 8.Be1+ Kg5 9.Bxd2+
and White wins.
© Tim Krabbé 1985, 1998