Endgame Study


    

 

EOTM1.jpg (8169 bytes)

31.Rxd5?!

According to Bronstein, the correct approach was to avoid the exchange of queens here:
31.Kg2! Qc1 32.Qb4! Rc2 33.Rxd5 Qxe3 34.Rd8+ Kg7 35.Qf8+ Kf6 36.Rd6+

31....Qc1+ 32.Qxc1 Rxc1+ 33.Kg2 Rb1!

EOTM2.jpg (7829 bytes)

Euwe correctly maneuvers to exchange queenside pawns, so as to transpose into a theoretically drawn rook endgame (four pawns vs three, all on the same side of the board).

34.g4 Kg7 35.h4 b6 36.h5 bxa5 37.Rxa5 Rb7

EOTM3.jpg (7552 bytes)

Euwe has acheived his goal: the queenside pawns are gone, and the position is now theoretically drawn. However, the difference between theory and practice now asserts itself, as Gligoric finds an interesting trick:

38.g5!?

Trifinovic, Gligoric's second in the Tournament, thought this move premature, and suggested White first bring his King to f4 (along with p-f3 and p-e4). However, according to Bronstein, even so Black should not lose with accurate defense.

38...gxh5?!

Euwe tempts fate here. Simplest was 38...h6!, exchanging off two pairs of pawns.

39.Ra6!

EOTM4.jpg (7347 bytes)

Suddenly, though equal materially, Black's position is now precarious - with a shattered pawn structure (the h5 pawn is particularly unfortunate) and King confined to the first two ranks, he must now play accurately to draw.

39...Rb3?!

Though not strictly speaking the losing move, this is definitely pushing things. The clearest drawing method was to place a rook on e6: 39...Re7! 40.Kg3 (40.Rh6 Re5 41.Rxh5 Kg6) 40...Re6! 41.Rxe6 fxe6 42.Kh4 Kg6 43.f4 h6 with a drawn king and pawn endgame.

Once again, we see how crucial transitions between different phases of the game (middlegame to endgame, rook ending to pawn ending, etc.) can be, and how even a great master such as Euwe can screw them up.

40.Rh6 Ra3 41.Kg3 Ra1 42.e4 Rg1+ 43.Kf4 Rh1 44.e5

EOTM5.jpg (7286 bytes)

44...h4?

The losing move. Black could actually shuffle back and forth; should White ever threaten e6 (say, after Kf5, f4 and Rh6-a6-a7), Re1 would hold. Now, however, the g4 square is available to white's King - and this provides Gligoric just enough juice to squeeze out a victory with.

45.Kg4 Rg1+ 46.Kf5 Rh1 47.Kg4 Rg1+ 48.Kf5

Repeating moves to gain a bit of time.

48...Rh1 49.f4

EOTM6.jpg (7310 bytes)

Once again, quoting Bronstein:

From now on, this position will be included in every endgame textbook. The winning method, discovered Gligoric in actual play, is not only pretty and logically consistant, but a valuable addition to theory as well.

Reaching the final position will require White to advance his pawn from e5 to e6 with his rook on the seventh rank, but this is not immediately playable. Gligoric's plan to make it so may be divided into the following stages:
1) Induce the black h-pawn to advance, and then capture it;
2) Bring his own rook to d1 and drive the enemy rook from the e-file;
3) Carry out the final maneuver by pushing the f-pawn to f5 and the e-pawn to e6 with support of the king and rook.

So, here we go!

49...h3

Alternatively, 49...Rh2 50.Kg4 Rg2+ 51.Kh3! accomplishes Bronstein's part 1.
But not 51.Kxh4? Rh2+ 52.Kg4 Rxh6 53.gxh6+ Kxh6 54.f5 Kg7 55.Kg5 f6+ 56.exf6+ Kf7, with a draw.

50.Kg4 Rg1+ 51.Kf3 Rf1+ 52.Kg3 Rg1+ 53.Kf2 Rh1 54.Rf6

EOTM7.jpg (7293 bytes)

The first zugzwang White employs to corral the h pawn.

54...Ra1 55.Kg3 Rh1 56.Kg4 Kg8 57.Rh6

EOTM8.jpg (7369 bytes)

The second zugzwang clinches the unfortunate h pawn's fate. If now 57...Rg1+!? 58.Kxh3 Rh1+ 59.Kg4 Rxh6 60.gxh6 we arrive at an interesting king and pawn endgame:

EOTM9.jpg (6881 bytes)

60...f6! 61.exf6 Kf7 62.Kf5 Kf8 63.Ke6 Ke8 64.f7+ Kf8 65.Kd6! Kxf7 66.Kd7 Kf8 67.Ke6 and wins.
Or 62...Ke8 63.Ke6 Kf8 64.Kd7 Kf7 65.Kd8 Kf8 66.f7! Kxf7 67.Kd7 Kf6 68.Ke8 Kg6 69.f5+ Kxf5 70.Kf7 Kg5 71.Kg7 again winning by an inch.

A remarkable variation! Anyhow, back to the game.

57...h2 58.Kg3 Rg1+ 59.Kxh2 Rg4 60.Rf6 Kg7 61.Kh3 Rg1 62.Kh4 Rh1+

If Black tries to trap White's king on the h file, White can threaten Ra6-a7 and e6 at the right moment, drawing Black's rook off the g file.

63.Kg4 Rg1+ 64.Kf5

EOTMa.jpg (7145 bytes)

White is now ready to oust the rook on the first rank via yet another zugzwang. Note that Black tries to maintain his rook behind the white f-pawn so as force the enemy king to hide on f5 from checks, thus preventing f4-f5 and e5-e6 (part 3 of Bronstein's plan). Note that side checks (with, say, the black rook on a4) would be less effective - White could use his rook to help shield himself.

64...Rf1 65.Rc6 Kf8 66.Rc8+ Kg7 67.Rd8

EOTMb.jpg (7195 bytes)

The final, and decisive, zugzwang! The Black rook must cede the back rank or the f-file. If the latter (say with 67...Ra1), White plays 68.Rd7, threatening e6, and if then 68...Re1, 69.Re7 and nothing can prevent the e-pawn's advance (Kf8 70.Kf6). Finally, 67...h6 68.gxh6+ Kxh6 69.Rg8 wins easily, since the Black king is cut off.

For the still unconvinced, Bronstein gives a few lines after 67...Ra1 68.Rd7:

a) 68...Ra5 69.Kg4 Kf8 70.Rd8+ Kg7 71.f5 Rxe5 72.f6+, with mate next;
b) 68...Kf8 69.Kf6 Ra6+ 70.Rd6!, etc.;
c) 68...Ra6 69.Kg4 h6 70.f5 Kg8 71.Rd8+ Kh7 72.g6+ fg 73. f6.

Thus Euwe chooses instead to give up the back rank, whereupon White is able to oust the Black rook from the f-file entirely, with f5 and e6 soon to follow.

67...Rf2 68.Rd1! Rf3 69.Ke4 Rf2 70.Ke3

EOTMc.jpg (7055 bytes)

Mission accomplished.

70...Ra2 71.f5! Rg2 72.Rd7! Rxg5

Bronstein gives a beautiful alternative loss for Black:
72...Kf8 73.f6 Ke8 74.Re7+ Kf8 75. Rb7 Ke8 76. Rb8+ Kd7 77. Rf8 Ke6 78.Re6+ Kf5 79.e6!

73.Kf4 Rg1 74.e6 Rf1+ 75.Ke5 Re1+ 76.Kd6 h5 77.Rxf7+ Kg8 78.Ke7

And Euwe resigned, fittingly with white pawns on f5 and e6!

EOTMd.jpg (6904 bytes)