Saturday, March 1, 2008

Laying Out An Endgame Plan

While I'm still on the topic of endgames, have a look at the following position.

This is a game in which I am Black. I lost this game because of poor endgame skills. As you can clearly see, Black has the advantage here because of the bishop pair and White's Knight is on the edge and needs to be brought back to the center asap.

White has the open c-file but this control is largely illusory. White is also saddled with a bad bishop on b2.

Now White just played the move 33. Rc2. (see 1st diagram on right)

In my game, I played a poor move, got duly punished and lost.

What should I do next?

What I should have done is plan a coherent long term strategy.

The most obvious move that comes to mind is of course 33.... Rc8 challenging the c-file and swapping rooks.

But is there a better move? I relooked at the position after the game and saw something.

First, there's no hurry to swap Rooks just yet. What Black needs to do is clamp down the position on White's camp and prevent any counterplay from White. Prophylaxis here pays off.

The a and b pawn structure is not fixed. We need to fix this pawn structure because if the Rooks come off (which is very likely) than it is important to find out how to deal with the pawn structures.

Black has problems with the light squared bishop (which is limited in activity), this bishop needs room to roam. The central pawn chain is rock solid and White's Knight can't attack it just yet.

The only thing is to find activity for this bishop.

How do we activate this bishop on b7?

Simple. We open up the light-squared a6-f1 diagonal. Is there any other downside to opening this diagonal? Not at the moment but it has to be remembered that the Knight on h3 will surely make its way to the Queenside.

The Knight is most likely to take the following path in red (as shown in the 2nd diagram on right).

The squares that need to be guarded are c3 and a2 (highlighted in green). Note that it is possible for White to push the b-pawn and the Knight can then squeeze into b3 to cause chaos in Black's position.

And the only way to do is to push the b-pawn. 33... b4 was the correct prophylactic move. If White plays 34. axb4 axb4. If White plays a4 the pawn chain is fixed in the same way and the Knight cannot use c3 or a2.

Are there any other threats or positions that might change this plan?

Not at the moment but we have to be careful of the pawn triggers by White, mainly f4, g4, h4 or a combination of the three. The good news is that this is not likely to happen soon. Any pawn breaks is easily guarded by Black's dark-squared bishop on d6 which is in a perfect position and the Black king can easily move to g6 if the need arrives to render more help to support the kingside pawns.

Only after the moves b4 can Black now swap Rooks with Rc8.

How is Black then to win? With the a3-f1 diagonal open, Black must at the opportune moment swap off the Knight with the light-squared bishop. This bishop can wreck some havoc by going to d3 and then b1 to prevent the Knight from going to a2 . In order to do this, White must make a concession on the kingside pawn chain sooner or later. And the plan is very easy. After trading off the Knight and Bishop, start moving the Black King up along the board and use the h-pawn to break open the position and get to a winning ending.

The move b4 is not easy to see, I admit. I certainly didn't.

Update: I've added the game below - highlighting, main lines and various variations (as suggested by Fritz) on continuations of this game. Cheers. :)


  1. honest. b4 was my move in about ten seconds. :) maybe i just got lucky.

  2. I saw b4 pretty quickly too, but just because that's the only move that locks the knight out. That is the only square the knight can really get too.

    For the record I see the position as about equall. I don't see the bishop pair as much of an advantage in this position where the pawns are likely to get locked up. If that happens basically you'll have one "bad" bishop because all of the pawns are on its colors, and another bishop that will be filled with sadness due to the fact that it all of the opponents pawns are on it's color.

  3. Rc8 looks natural, though as you pointed out it does not need to be played right at that point. It's not like it can get in and do any damage. Also there's not a second rook to come to c1 and toatlly dominate the open file.

    I like your knight path map. Sometimes in positions like this it's simply a matter of counting how many moves it takes to get a particular piece back into play, and identifying the crucial squares that you want to keep it away from.

    I agree with Wang's accessment here, that even with b4 it's a rather drawish position. I don't see that the bishop pair is enough of an edge to make progress here. In some ways I'd rather have the knight.

    I'd be curious to see the rest of the actual moves played in order to see if you still could hold after Rc8. At first glance I don't see this as the losing move.

  4. wang:

    thank you both for your comments. my initial thoughts were that it's rather drawish as well. but after analysing with Fritz (I've added Fritz's suggestions at the end of the blog), the position seems to be winning. it does seem like there's a draw somewhere.


    Fritz gave the analysis of Rc8 as the same as b4 although i've not analysed it thoroughly if White plays b4 first. would be interesting to find out.

    March 1, 2008 10:29 PM