Fortresses!

Tomorrow morning we will have a final look at Endgame Play before sending it away to the printer. I am physically and emotionally exhausted to a degree I have not felt since the mid-1990s – where I somehow still thought that consuming a lot of alcohol was a fun way to waste your life away. This time I have also gained a few kilos, but I have something to show for it!

I do not want to relate Endgame Play to other endgame books. It is the endgame seen through the prism of the Grandmaster Preparation series, where only one book will have a different style: Thinking Inside the Box, which is meant to be the underlying theoretical book. There are more exercises than usual in EP and I think some of the chess is really nice, but this will be up to others to decide.

The most interesting chapter in the book is probably the one on fortresses. Mainly because I do not know of any real good material about fortresses. I looked In Dvoretsky’s Endgame Manual and I did not feel it made me much wiser on this topic. Maybe someone else has written excellently about fortresses; I am just not aware of it (nor am I saying that I have done so – I just say that I have tried to make a few observations about fortresses as a strategy).

While finishing the book I came across the following fortresses. They should all be draws.

The last one I did not put in the book. It is well-known for many, but it was still lurking around in the databases associated with the book right till the end.

Zhou, Yang-Fan – Jack Rudd, London 2012

White could have drawn with: 62. Kd2! Bd3 63. Kd1 Kc3 64. Ke1!, where the white king both avoids being forced into zugzwang and stays close enough to rush for a1 when Black takes on a5. Black can take the pawn on b4 and prevent the white king from making it to a1, but in that case the stalemate of the king will actually be stalemate and not force White to play b4-b5.

In the game White did not know about this idea it seems:
62. Kb2? Kd3 63. Kb3 Bc4+ 64. Kb2 Kd2 65. Kb1 Kc3 66. Kc1
8/8/p7/P7/1Pb5/2k5/8/2K5 b – – 0 66[/fen]
66…Be2!
66… Bd3 67. Kd1!
67. Kb1 Bd3+ 68. Ka2
68. Kc1 Bc2!
68… Bc2 69. Ka1
69. Ka3 Bb3 70. b5 axb5 71. a6 b4#
69… Kb3 70. b5 axb5 71. a6 Be4 72. a7 b4
0-1

13 thoughts on “Fortresses!”

  1. From a recent game of mine, an interesting task for the interested reader: Is the following a fortress?
    White: Kg1, Qb4, pawns g2, g4
    Black: Kg8, Rh6, pawns b5, d6, f7
    Black to move
    Clearly a draw without white’s pawns and without black’s b5, d6 pawns. How does it look now?

  2. Jacob Aagaard

    The position you are talking about looks like this.

    [fen size="small"]6k1/5p2/3p3r/1p6/1Q4P1/8/6P1/6K1 b - - 0 1[/fen]

    Black to move.

    I would be afraid that White could somehow achieve the following position:

    [fen size="small"]6k1/5p2/2Kp1Qr1/6P1/8/8/6P1/8 b - - 0 1[/fen]

    White wins here. The question is of course if the white king will make it across the e-file. I think zugzwang should take him there, but as it is the middle of the night, I think I will go to bed rather than trying to find out.

    But a quick guess is that White wins.

  3. Jacob Aagaard

    I was thinking about this fortress while falling asleep and I have found a zugzwang position that breaks it.

    [fen size="small"]3Q4/5pk1/3pr3/6P1/5K2/8/6P1/8 w - - 0 1[/fen]

    White can always achieve this position with Black to play (Qe7-d7-d8 for example). The rook will have to go to g6 after which the white king goes to d5.

    I should maybe add that White wins easily if Black gives up the d-pawn. We can imagine this position:

    [fen size="small"]6k1/5p2/5Qr1/5KP1/8/8/6P1/8 b - - 0 1[/fen]

    where White wins after 1… Rg7 2. Qd8+ Kh7 3. Qe8 and the zugzwang is deadly.

  4. In positions of like the second one you have to account for a potential mating attack with king and connected passed pawns. In particular if one of the pawns is an a- or h-pawn. So if you replace the pawn on c5 by the the one on a5 it may no longer be a fortress. You play b6, then you triangulate with your king so that you reach a position with Black’s king on a6 and your king on c4 with Black to move. An Kb7 you can play Kb5! and you win by a6+ and Kc6. Black will queen, but you’re going to mate him.

  5. Yang-Fan’s too young to remember Karpov-Korchnoi, of course. Superannuated punters like myself would have got that right – there’s got to be something to be said for age.

    Is your third one a fortress, exactly, Jacob? Black just has a threat White can’t prevent, doesn’t he?

  6. Jacob Aagaard

    @John Cox
    Take away the bishop and White would win. But of course, we can have a long debate about fortress definitions. I would refer to my book to start with :-).

  7. @Jacob Aagaard
    My opponent got me into Zugzwang like this: white: Kd7, Qd4, black: Re6, Kg7 (pawns as above) and now Kh7 and then triangle with the king: 56. Kd8 Kg8 57. Kc7 Rg6 58. Kd7 Re6 (Kh7 Ke7) 59. Qxd6 and white won easily because of the g2 pawn. Your approach with the queen on f6 seems easier, though.

  8. @Jacob Aagaard

    Well, that’s true, of course. The bishop could be on other squares as well, I think, as long as it controls g3?

    To me a fortress means a position where the defender can defend passively and the attacker can’t progress, which this isn’t. But, as you say, it’s just a matter of definition, and I look forward to reading yours following my upcoming birthday.

  9. Jacob Aagaard

    My definition has much more to with fortresses as a technique than as a thing that is either there or not. It makes more practical sense. Also, talking about when fortresses do not work is quite interesting.

    One of the observation of the book is that a lot of bishop endings are fortresses; even with few pieces on the board.

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top