Page1 I Page 2

27.Kf3 Nxf5 28.exf5!?

A DIFFICULT DECISION

This was a key decision, and I'm still not sure whether this or 28.gxf5 was better. My thought process was that I would welcome the trading of all the rooks, as then the h-pawn would fall and white would win easily. I also believed it would be easier to create a passed pawn after the text, as g4-g5-g6 is hard to stop.

On the other hand, gxf5 gives the white king quicker access to the h4 pawn via g4. In the game, white had to allow black to liquidate his doubled f-pawns to get to that square (by playing g4-g5).

After the game, I almost had a heart attack when I asked my opponent what he would play after 28.gxf5. He trotted out the logical 28...Rg7 29.Rg1 R8g8? when 30.Rg4! is very strong for white. The point is that white threatens simply R1g1, and 30...Rxg4 31.hxg4 followed by Rh1 and Kg3 rounds up the h-pawn.

28...R7h8 29.Re1

White has to be careful about making the g5 advance before he's prepared. In this case, 29.g5? Rhg8! creates strong counterplay. Once again, patience is key for white.

29...Be7 30.Re2 Rhg8 31.Rae1 Rh8 32.Bd4

NO UNNECESSARY TACTICS!

Here black is again baiting white to play g5, and I almost did. I wasn't certain that white could win after 32.g5 Rhg8 33.Bxf6 Bxf6 34.Rxe8 Rxe8 35.Rxe8 Kxe8 36.gxf6. Even though white will get a passed h-pawn, his king doesn't have easy access to black's queenside. At the time, I couldn't see a clear win, so I avoided the line.

Grandmaster John Nunn has recently written a useful book titled SECRETS OF PRACTICAL CHESS (click to see reviews by Watson and Silman). One of his tips is to remember the acronym DAUT – don't analyze unnecessary tactics. That's useful advice coming from a tactician of Nunn's ability. The above line is an example. I couldn't see clear to the end and didn't see any reason to spend an inordinate amount of time trying to work it out.

White's perspective is that he doesn't need to burn his bridges. He can slowly improve his position and then seek a winning line.

32...Rhg8 33.Bf2 Rh8 34.Re4

It is now clear that black is going to try to sit tight. White has been improving the scope of his pieces and black is relegated to shuffling his rook along the eighth rank.

White's last is a useful "creeping" move. White may play Rg1 and/or Be3 and aim for g5, when the rook, after a pawn exchange, would attack h4. White can also choose to bring his king to the queenside if circumstances warrant since the Re4 prevents a discovered check.

34...Kd8 35.Be3 Kd7 36.Re2

CAT AND MOUSE

One of my favorite books is ENDGAME STRATEGY by Shereshevsky (click to see reviews by Watson and Silman). In there, one of his key admonitions is "Do Not Hurry!" Before white undertakes a committal advance (i.e., the intended g4-g5) he further improves his position. The point of this move is to protect the second rank and remove the rook from a dark square. This rook will likely end up on the g-file, and white doesn't want to give black any counterplay based on getting his bishop into play via an attack on the rook. Karpov is a master at this sort of patient build-up.

36...Bd8

Black tires of dealing with white's firepower on the e-file and seeks an  exchange of rooks to lessen the pressure.

37.Rxe8 Rxe8 38.Bd2 Rg8

Black cannot afford a trade of rooks, because then the h4 pawn will be defenseless.

39.Rg2 c6

TOO LITTLE, TOO LATE

This is black's first real attempt at counterplay in a long time. From white's perspective, he does not want to give up his pawn on d5 because it severely cramps black's king. White will allow the pawn exchange, because he can use his bishop to prevent the black king from advancing and attacking the isolated pawn on d5. In the endgame, king activity is very important. If your king is further advanced or more mobile than your opponent's, you have an important imbalance in your favor.

Meanwhile, it's time for white to get back to his primary objective, which is to get his kingside pawns rolling. His next move is obvious – it's the only way for white to convert his kingside pawn advantage, and it frees the g4 -square to attack the h4 pawn as well.

40.g5! cxd5 41.cxd5 Kc7 42.Be3

KEEPING BLACK'S KING PASSIVE

The bishop move keeps the black king passive. Otherwise black will play ...Kb6 and ...Kc5 when suddenly black has some threats that must be considered. It is much better to take a move to prevent this. Again, patience and preventing unnecessary counterplay are very important in developing winning technique.

42...Kd7 43.g6(!?)

NOT SIMPLEST

When I played this, I envisioned the zugzwang position that arises on move 50 (it really wasn't that hard since most of the moves are forced, as we'll see). While it is a win, it's a bit subtle. An easier, more practical approach would have been 48.Bf2 fxg5 49.fxg5 Bxg5 50.Bxh4! f6 51.Bxg5 fxg5 52.Kg4 with an  easy win. To tell you the truth, I developed tunnel vision. I saw the winning ideas behind this move and I didn't stop to look for anything else. That's not a good idea!

43...fxg6 44.fxg6

Even here, 44.Rxg6 was an easier approach.

44...f5!

Black must prevent white from playing f4-f5, and as an added bonus it protects the h-pawn.

45.Bd4!

MASTER OF THE a1-h8 DIAGONAL

Likewise, white must seize this diagonal. If black can play ...Bf6 he has excellent chances to hold the position.

45...a6

Unfortunately, 45...Ke8 (attempting to forestall white's plan of g7 and Rg6) doesn't work because white can play 46.Re2+! and the black king doesn't have access  to f7. Then 46...Kf8 47.g7+ Kf7 48.Re6 leads to the same ultimate position as the game. Is there anything better to do with the queenside pawns? Not really. Black can't afford to play ...b6, since the white king is given an easy route into the queenside via the white squares. He can't really leave them where they are either, because later white will play Bd4 and bring the king over to the queenside and threaten to simply take the a7 pawn and win on that side of the board.

This brings up another very important point that is discussed at length in Shereshevsky's fine book – the principle of two weaknesses. Shereshevsky (and others) point out that, for one side to win, they must usually create two weaknesses in the opposing camp. Generally, one alone is not enough because the defender can bring all his forces to bear on that one particular problem. If, however, a second weakness can be created, the defense is usually unable to deal with both problems. This is such a case.  The first weakness, of course, is the pawn on g7 that is poised to queen. As we're seeing, however, that alone is not enough to win. If the queenside were somehow totally blocked off from the white king and bishop, the win would be problematic. It is not, however, and it constitutes the second weakness that white can exploit.

46.g7 Ke8 47.Rg6!

This is the key move that stymies black. White will play Rh6-h8 and black's king and rook will be frozen in place. Then the white king and bishop will be free to raid the queenside. Black can't afford to play an immediate 47...Kf7 because 48.Rxd6 and all of black's pawns start falling off.

47...Be7 48.Rh6 Kf7 49.Rh8!

DEAD MAN WALKING!

Unfortunately for black, he finds it very hard to shake the white bind. He can never exchange rooks, both his king and rook are needed to control the g8-square, and his bishop has few moves as well. Basically, all he can do here is make a bishop move or play ...Rd8. That gives white the time to leisurely pursue a winning king advance, most likely via the queenside.

49...Bd8 50.Ke2

The winning plan begins to take shape. White will bring his king to the queenside. At the key moment (i.e., when the bishop is on e7, the white pawn is on a4 and the white king on b4), he'll exchange the rooks and play his king to a5. Then Kb6 is threatened and Bd8+ is answered by Bb6. Black can't trade bishops, but after ...Be7 Bc7 for white followed by Kb6 decides the game.

Can black do anything to counter this plan? Not really. He can play ...b6, but white just waits until black has to play ...Be7, and then he trades rooks and plays Bxb6 followed by Ka5 and takes the a-pawn.

He could try to play his pawns to b5 and a5, but white will bring his king to d3, trade rooks, and play Bb6, winning the a-pawn. This idea of trading rooks and winning the bishop ending is the key that my opponent said he missed. Note how little activity the black bishop has compared to the white bishop.

In short, black has no way to prevent white from winning on the queenside. Instead of getting to see this process, black now played...

50...Ba5?

Hoping to activate the bishop. Unfortunately, after

51.Rxh4, 1-0.

Black resigns, as he now realizes that white can just play Rh8 and advance the h-pawn all the way to h7.

Page1 I Page 2