Keene’s Bold Rook Lift

This game is a beautiful example of an “Isolated Queen’s Pawn” (IQP) position and resulting attacking prospects. The game is from the 1975/1976 Hastings ENG tournament (round 13). Raymond Keene has the White pieces and Anthony Miles has the Black pieces.

1. Nf3 Nf6 2. c4 c5 3. Nc3 Nc6 4. e3 e6 It is still too early to determine which opening this game will eventually shape into: English Opening (1. c4) lines, Queen’s Indian Defense lines, or even something else through transposition. What we can say for sure, thus far, is that both sides are playing solid chess.

5. d4 d5 6. cxd5 Nxd5 7. Bd3 cxd4 8. exd4 Be7 After 5. d4 d5, it wasn’t even clear which side was going to get the isolated pawn! For this game, it looks like White got saddled with it. Typically, an isolated pawn (isolated pawn on any file) is viewed as a weakness. The reason is because it can no longer be defended by another pawn (the pawn is “isolated” because the friendly pawns on the neighboring files are absent; in this case, the White d-pawn is “isolated” because of the fact that White has no c-pawn or e-pawn). This is all correct analysis, but not always is an isolated pawn a weakness: sometimes its compensation can even make it a strength!

It reminds me of a book from Irving Chernev (The Most Instructive Games of Chess Ever Played: 62 Masterpieces of Chess Strategy) where he quotes Siegbert Tarrasch (Game 18: The Isolated Pawn [p. 81], 1965), “He who fears an isolated Queen’s Pawn should give up chess.”

The reason for this statement is because any isolated pawn is potentially weak, but this doesn’t mean the player should fear getting it in light of compensation. Thematic to many isolated Queen’s Pawn positions, the player with the isolated pawn will either play for an attack (utilizing the open lines that the isolated pawn brings as well as the space advantage usually present) or find a way to trade off the isolated pawn. Usually the side without the isolated pawn will either try to blockade the pawn (to fix the weakness) or to exchange many pieces and enter an endgame. The reason for this is because pawn structure and positional weaknesses seldom lessen when pieces are exchanged, but attacking prospects usually do.

In this case, White has the isolated [Queen’s] pawn. If Black can exchange into an endgame, then the positional weakness of the isolated pawn should be apparent in many cases. For this, White wants to avoid exchanges and play aggressive attacking ideas while they have open lines for attack (and in this game, White also has a space advantage that an isolated Queen’s Pawn often creates).

9. O-O O-O 10. Re1 Nf6 After some more developing moves, we still have a position in established “opening theory” here! This position typically arises out of the Semi-Tarrasch in the Queen’s Gambit Declined (Mainline of the Pillsbury Variation for the Semi-Tarrasch).

The usual move order is: 1. d4 d5 2. c4 e6 3. Nc3 Nf6 4. Nf3 c5 5. cxd5 Nxd5 6. e3 Nc6 7. Bd3 cxd4 8. exd4 Be7 9. O-O O-O 10. Re1 Nf6 and we have the identical position this game reached through transposition.

In the current position, a computer may evaluate this as roughly equal. In practical play, White scores very well from here though. 11. Bf4, 11. Bg5 and 11. a3 all score very well for White according to my opening database.

11. Bg5 Nb4?! This move might be questionable because it may be pursuing the wrong plan for this moment. Black is probably planning on …Nb4 and then …Nbd5 (to blockade the isolated pawn and fix it as a weakness to attack later). This appears to “gain time” because White must spend a tempo moving the Bishop (likely to b1) to keep the Bishop Pair. This plan sounds good for Black, but it may be a little premature. NM Sam Copeland recommends 11…b6 as an improvement; the idea is preparing …Bb7 when the fianchetto Bishop will be active. The move …Nb4 isn’t terrible (perhaps a good move later), but right now: keeping the focus on piece development and opening principles may be better.

Keene himself (White pieces in this game) commented on this position by also stating: “Black’s plan to dominate the blockade square d5 instantly is possibly too straightforward.”

An alternative move for Black (although not developing thematically) is 11…h6!? with the intriguing point of 12. Bf4 Nb4. The difference from playing …Nb4 on move 11 is that the White Bishop from g5 is now on f4 where it won’t put as much Kingside pressure as when the Bishop influenced the d8-h4 diagonal. Whenever a pawn move near the castled King is made though (such as …h6 in this variation), one must be careful about sacrifices that expose the King. Here …h6 appears fine, but Black would always have to be on the lookout for future sacrifices on the h6 square.

12. Bb1 b6 13. Ne5 Bb7 This is all a very logical continuation. White plays Bb1 to save the strong Bishop Pair and then Ne5 looks like an active Knight move; similarly, Black plays the Bishop fianchetto idea for development that we foreshadowed earlier. What is one of the most ambitious attacking ideas for White on the next move though?

14. Re3!! Keene’s own words on this position were: “A difficult and bold move. When I played it I already had to be sure that my kingside attacking chances would be sufficient compensation for my lack of queenside development and the clumsy position of my rook on the third rank.”

Keene gave this Re3 idea a double exclaim (“!!”) and then cites a chess game from 14 years earlier with a similar Rook lift. Keene admits that in that earlier game, Filip - Pogats 1961, that game ended in a draw and that “neither Miles nor I had any notion of this until I discovered the theoretical background after our game.”

In the referenced game, Miroslav Filip had the White pieces and Jozsef Pogats had the Black pieces. That game reached the Re3!! “theoretical background” from the moves:

1. d4 Nf6 2. c4 e6 3. Nf3 d5 4. Nc3 c5 5. e3 cxd4 6. exd4 Be7 7. cxd5 Nxd5 8. Bd3 Nc6 9. O-O O-O 10. Re1 Ncb4 11. Bb1 Nf6 12. Bg5 Bd7 13. Ne5 Rc8 14. Re3!!

However, the 1961 game ended in a draw after Black’s 24th move.

Back to the Keene game where White won spectacularly, the text continued:

g6 15. Rg3! Rc8 Keene’s thinking about 15. Rg3! was that, “It looks odd to aim the rook against a granite wall (g6) but the granite has faulty foundations. In contrast, the more natural 15. Rh3 achieves nothing.”

In light of Keene’s analysis: 14…g6 may indeed have “faulty foundations” but I believe that Miles (Black pieces) played the best move with 14…g6. Sometimes, the defending player must make a concession like this as the lesser of two evils. How else can Black hold the position if not for the …g6 idea here? I am sure Miles was well aware of the weaker dark-squares created around the King and such, but where is there an improvement? For instance, 14…Nc6?? 15. Bxf6 Bxf6 16. Bxh7! Kxh7 17. Qh5+ Kg8 18. Rh3 and White has a winning position because the threats of Qh8# or Qh7# will cost Black massive amounts of material with 18…Bh4 19. Rxh4 Qxh4 20. Qxh4 and Black does not want to play a game down 5 points of material against a strong player like Keene.

Alternatively, something like 14…Nfd5?? (trying to exchange Bishops and lessen White’s attackers) fails due to: 15. Bxe7 Qxe7 16. Bxh7+! and it is still clear that White has more than enough attackers present to crash though. This line would end in similar fashion to the 14…Nc6?? variation; White will play Qh5+ and Rh3 with little Black can do to prevent this.

16. Bh6 Re8 17. a3 Nc6 18. Nxg6! Even though the attacking idea of 14. Re3!! was “bold”, the move 18. Nxg6! felt like the climax to the attack for me. Positional and tactical motifs fuse together into what becomes an amazing attack for White.

Solid analysis mentioned by NM Sam Copeland points out that 17. Nxg6?? is “Too early.” He gives the critical defense for Black: 17…hxg6 18. Bxg6 fxg6 19. Qb1 Ne4 20. Nxe4 Rc2! - +

hxg6 19. Bxg6! fxg6 White must continue the attack or else the sacrificed material will give Black a winning endgame; the only way White can win is if they can finish Black off before the endgame stage rises. The key to this attack cashing in is if the White Queen can get to the g6 square.

The first candidate move to look at is probably 20. Qd3?? but this is a blunder due to 20…Ne5 and the pin on the d-file is key for Black. 21. dxe5 Qxd3 22. Rxd3 Nd5 and Black is much better. Not only is Black up in material, but Black also has great positional pressure mainly due to the controlled c-file and flexible piece movement.

The next candidate move to investigate is the one that I thought was best when I saw this game for the first time. I saw the d-file pin, so I liked Qc2 with the idea that the White a-Rook can swing into the action with smooth development soon.

20. Qc2? is a mistake. I overlooked the c-file pin! 20…Ne5 21. dxe5 Ne4 and perhaps White is barely surviving, but we don’t want “barely surviving” when we had a win! We want a better result for White than the draw in the Filip game.

20. Qb1!! This is the only move that keeps the attack going and decisively wins!

Ne5 21. dxe5 Ne4 Now White is much better because the c3 Knight can safely capture on e4 - thanks to no pins on the c-file or d-file that eyes the White Queen.

22. Nxe4 Kh7 Leave it to Stockfish to come up with a creative way for Black to temporarily prevent the White Queen from infiltrating on g6. It recommends 22…Rc1+ 23. Qxc1 Bxe4 24. Qf4 Bf5 and at least the White Queen is shut out for the moment. Unfortunately, computers do not yet understand the value of complexity/sharpness as this “solution” is merely temporary since White would have an easy endgame being up by four points of material. Contemporary chess engines tend to always pick the “slow and painful” route that loses because they are purely objective and see that as the “best” option available. A human player however, will usually try to play complex ideas when losing; the hope is to “muddy the waters” so the other player may slip up too. A human player realizes that a losing endgame for us is too simple for a strong opponent to beat you with.

Defending the g6 pawn with the King is a very “human” defense. If White misplays somewhere, then at least g6 will be defended and then g6 can’t be safely captured.

23. Nf6+ White already has Black in a mating net.

Bxf6 24. Qxg6+ Kh8 25. Bg7+ Bxg7 26. Qxg7# White’s attack ends the game with an impactful checkmate. This attack was made possible because of “Keene’s Bold Rook Lift” and an argument could even be made that this attacking potential began even earlier: back on move 8. exd4 with resulting in an Isolated Queen’s Pawn (IQP)! Hopefully we can all learn to make use of isolated pawns like Keene does. These isolated pawns are directly weak (since they can’t be defended by nearby pawns), but they indirectly create open lines and typically bring a space advantage - both of these effects serve as elements for a potentially powerful attack. Remember these assets of an Isolated Queen’s Pawn and perhaps you can be lucky enough to repeat a brilliant and “Bold Rook Lift” like Filip or Keene.