Bird by Bird, Part 3A
We now move on to the third part of my series on the Bird Variation of the Ruy
Lopez, and here it will get a little bit more challenging for Black. Michael
Goeller notes, in his comment to "Bird by Bird, part 2," that he has been
playing the Bird Variation for a long time but always stuck to the classical
variation with ... Bc5. The Chess Authorities look down their noses at ... g6
and ... Bg7, which is supposed to be bad. So the main point of my next two posts
is going to be that this variation is actually quite playable for Black.
The argument against this line is that Black has already wasted a tempo by
playing 3. ... Nd4, and now he proposes to waste another tempo by developing his
KB in two moves instead of one. Can Black really afford to lose two tempi in the
In the end it will come down to concrete analysis, but as long as we're
speaking in terms of generalities, here is my view.
- I just don't like ... Bc5 very much because I don't know what my bishop
is doing on this diagonal. It's like a spectator at a sporting event who is
seated behind a pillar and can't see anything. On g7 it puts pressure on the
long diagonal, which becomes particularly significant if White chooses to
challenge Black's center with c3 (as he often does).
- As I said in my first "Bird by Bird" post, Black intends to prove that
White's pieces on the queenside (particularly the bishops) do not
communicate well with the kingside. This is particularly true if White
chooses to play f4 (as he often does).
- The moves ... g6 and ... Bg7 are intended to be provocative, like waving
a red flag at a bull. White, especially if he knows theory, will see them as
sub-standard moves and will try to punish them. When your opponent, as
White, plays the move f2-f4, then you know that the red flag trick has
By the way, Goeller also points out that the line I am advocating, with 5.
... g6, was played in the 19th century by Joseph Henry Blackburne, so we should
call it the Blackburne Variation (or Blackburne Subvariation of the Bird
Variation?). Today I will talk about what happens if White charges at the red
flag by playing f2-f4. I guess we should call this the f4 Sub-sub-variation of
the Blackburne Subvariation of the Bird Variation. But I'll skip all the "subs"
and just call it the f4 Variation.
So the f4 Variation begins as follows: 1. e4 e5
2. Nf3 Nc6 3. Bb5 Nd4 4. Nxd4 ed 5. O-O g6 6. d3 ...
White can play 6. f4 here, but it's difficult to see what the point is in
doing so. He can't checkmate Black without the participation of his queenside
pieces, so he will need to play d3 anyhow. However, I did play one tournament
game where my opponent adopted this hyper-aggressive approach, and what happened
is kind of interesting. My opponent was Jeremy Fremlin, a teen-aged player in
Santa Cruz who used to play a lot, but I haven't seen him in the last three or
four years. I think his rating was around 1600, but he was really closer to 1800
strength. This game was from 2001:
6. f4 Bg7 7. f5?! Be5 (Sure, I'll take the dark
squares!) 8. Qg4? (Definitely not the best square
for the queen. Fritz recommends 8. Qf3, but Black is fine after 8. ... Qf6.)
8. ... c6 (I'm not sure if this is really
necessary; 8. ... h5 right away also seems good.) 9. Ba4
h5 10. Qf3 g5 (It's always nice for Black when he can close up the
kingside this way. Usually he can't do it because it creates a hole on h5.)
11. d3 g4 12. Qe2?? (Better was 12. Qf2, but Black
still has an advantage after 12. ... g3! 13. hg Nf6 and the knight has a
beautiful outpost on g4.) 12. ... Bxh2+! 0-1
Fremlin-Mackenzie (final position)
This is one of the most bizarre games I've ever played! Twelve moves into the
game, Black has only one developed piece and White has three. Black then
sacrifices his only developed piece, and White resigns! It's mate if White
takes the bishop, and pretty god-awful if he doesn't.
But now let's get to the main line, which is 6. d3 Bg7
7. f4 (diagram).
Starting point of the f4 Variation.
In this position I believe that Black should not wait around: he should punch
the bull in the snout as soon as he possibly can. Therefore, I recommend
7. ... c6, which leads to two subvariations: (A)
8. Bc4 (the more common move)
and (B) 8. Ba4 (more subtle). If you're
curious, my record in variation (A) is +3 -2 =3, and my record in variation (B)
is +2 -1 =1, but this is perhaps too few games to really mean anything.
(A) 8. Bc4 d5
Of course. Again, there is no point in waiting around. I did play 8. ... Ne7?
in one of my earliest games, and should have gotten destroyed after 9. f5! gf
10. Bg5! Black now has no good way to get in his ... d5 break; if he does it,
his king will get caught in the center. If he castles, White will play ef and
f6. However, I got lucky -- my opponent played the weaker 10. ef d5! Black
blocks the bishop's diagonal, and now 11. f6 is not particularly dangerous. In
fact, it opens up the g-file for Black's rook, and Black went on to win in
9. ed ...
This is what White almost always plays (6 times out of
7 tournament games) but I wonder whether 9. Bb3, as played by Bob Rowley against
me in 1999, might actually be better. The point is to keep the tension and in
particular to keep the option of f4-f5 open for White. As we've seen in the line
above, f5 not only threatens f6, it also gets the f-pawn out of the way of the
QB. As I've said above, one of my main objections to the Red Bull Variation is
that White blocks in his own QB. So playing f4-f5 should be a high priority for
Rowley-Mackenzie continued 9. Bb3 Ne7 10. Nd2 h5 11. Nf3 Be6. The computer
likes White's position (+0.7 pawns), but I felt as if I had a reasonable game
here. The game ended in a draw. Curiously, the move that the computer likes is
one that barely seems playable to me: 9. Bb3 Nf6!?, which does nothing to
prevent f4-f5. But according to Fritz, 10. f5 de 11. de gf 12. ef O-O!? is
copacetic for Black.
Has the computer lost its mind?
Even though I told you that Black wants to goad White into attacking on the
kingside, this seems to me a little bit extreme. Black castles into a
compromised kingside, where White already has a pawn looming on f5 and has great
diagonals for his bishops -- and yet Fritz somehow thinks this position is even.
Well, after 30 minutes of thought my copy of Fritz 9 gives White the tiniest of
advantages, 0.29 pawns, after 13. Bg5 h6 14. Bh4.
Until I understand the computer's reasoning better, I say the jury is still
out on whether 9. ... Ne7 or 9. ... Nf6 is better after 9. Bb3.
Anyway, let's continue with the main line, 9. ed. Black's response is
9. ... ed
And now White has a choice of three checks: 10. Re1+, 10. Qe2+, or 10. Bb5+.
But in the first two cases White normally follows up with the bishop check
anyway, so for practical purposes they are all the same variation.
10. Bb5+ Kf8!
This is the key, and very distinctive, move of the f4 Variation. I
avoid playing 10. ... Bd7 for two reasons. First, after 10. ... Bd7 11. Qe2+ Ne7
the possibility of f5 worries me (perhaps after 12. Bxd7+). By sidestepping with
the king, Black keeps control over the f5 square with his bishop. Second, Black
realizes that White's bishop on b5 is in a very exposed location. In fact, Black
is threatening to trap it right away with ... Qa5. We'll see that this theme of
the undefended bishop comes up over and over again. Although with ... Kf8 Black
forfeits the right to castle, he plans to castle by hand with ... Ne7, ... Bf6,
and ... Kg7.
11. Qe2 ...
This position is more likely to arise by transposition (10. Qe2+ Ne7 11.
Bb5+). Other possibilities for White:
11. Re1 Ne7 12. c3? was played in Bekker-Mackenzie,
2008. Here we both overlooked the fact that I can win a piece with 12. ... dc
13. Nxc3 (13. d4 saves the piece but after 13. ... cb 14. Bxb2 Qb6 15. Qe2 Be6,
Black has "the pawn and the compensation," to quote GM
Dzindzichashvili) 13. ... Qb6+.
Oops, there goes the bishop.
I actually missed this too, and instead I played 12.
... Qb6, which also gives Black a wonderful position. After
13. c4 Be6 14. Nd2 Rd8 15. Qc2 Nf5 16. Nf3 a6 17. cd Bxd5
18. Bc4 Bxf3 19. gf (diagram) Black has a dream position, and went on to
win the game.
Bekker-Mackenzie. A dream position for Black.
Going back to the diagram after 10. ... Kf8,
Michael Aigner (a well-known Bay Area master whose
blog should be well-known to all of my readers) was aware of the danger to
his bishop and played the prophylactic move 11. Ba4.
That game, played in 2000, continued 11. ... h5 12. Bd2
Bf6 13. Qf3 Bg4 14. Qf2 (diagram).
After the game, Michael said that he felt White had a long-term advantage
because the doubled d-pawns are a weakness that won't go away. But in my
opinion, Black has several compensating factors. The doubled pawns create more
files for his rooks (the c- and e-files) There is a beautiful square for the
knight on f5, although the bishop will first have to vacate g4. If White should
ever play h3, then the dark squares on the kingside become a real weakness, and
Black can try to play Bh4 and Bg3. Finally, those allegedly weak d-pawns have a
real cramping effect on White's position; they keep his white-squared bishop
locked up on the queenside and they keep the dark-squared bishop from going to
e3. All in all, I think the position is in a state of dynamic balance.
Here, unfortunately, I made a mistake with 14. ... h4?,
and I think this is really why I had problems later. (I eventually lost the
game.) After this move the dark-squared bishop becomes a completely passive
piece, and the h4 pawn becomes a potential target. It would have been better to
complete my development with 14. ... Ne7, and a sample line could go 15. Bb4?!
Kg7 16. h3 Be6 17. Nd2 Nf5 18. Nf3 b5!
Oops, there goes the bishop (again!).
Oh well, this is just fantasy chess, but I wanted to remind you about the
combinative possibilities that arise from White's undefended light-squared
Okay, let us continue with the main line, 11. Qe2, and we'll follow the game
Maser-Mackenzie 2003 (by transposition), which actually went 10. Qe2+ Ne7 11.
11. ... Ne7 12. Nd2? ...
Maser is blissfully unaware of the danger to his b5 bishop. He should have
played 12. Ba4, transposing to something like the Aigner-Mackenzie game.
12. ... Qa5!
Here we go again! White's bishop is trapped. However, Maser comes up with
some ingenious counterplay:
13. Re1 Bf6 14. g4! (diagram)
Oops, there goes the bishop! (Again.) (Or does it?)
If Black takes the bishop on b5, he has to give the piece back after 15. g5.
However, Black has a completely safe and sound alternative here, which is 14.
... Bxg4 15. Qxg4 Qxb5. Not only is Black a pawn up, but he has the same sort of
solid, unassailable king position with the knight on f5 and the bishop on f6
that we saw in the Bekker game.
Alas, I wasn't satisfied with winning a pawn -- I got greedy, and I insisted
on winning the whole piece. Mostly, I just didn't believe that White's attack
was worth a whole piece. And so I played
14. ... Bh4? 15. Qe5 Rg8 (diagram)
Now White plays an amazing move. With a bishop and a rook hanging, he doesn't
defend either of them, but instead plays a move that opens up a file for me to
attack his king!
16. f5!! Bxe1?
If 16. ... gf 17. Nf3 Rxg4+ 18. Kh1 Bxe1 19. Bh6+ Kg8 20. Ng5! (better than
the immediate 20. Qxe7, although that also wins) 20. ... Rxg5 21. Qxe7
threatening mate on f8 and also attacking the rook on g5. Wow! The computer says
17. .. Qxb5 (instead of 17. ... Rxg4+) is Black's only move to stay in the game,
but then after 18. Nxh4 Rxg4+ 19. Kh1 White is clearly better, with Black's king
position shattered and the knight on h4 taboo.
17. f6! Be6 18. Nc4! Qxb5 19. Bh6+ and White
went on to win in smashing fashion. If you're interested, I gave this game as an
example in my ChessLecture called "Tactical Motifs 203: Mating Nets."
You might be surprised that I would show you a game where Black got crushed
in an opening where I'm advocating the Black side. Well, there's some method to
my madness. First of all, we have to learn from our mistakes. It's important to
understand what can go wrong if you misplay the position. Second, I wanted to
show this game because it's probably the most brilliant attack that has ever
been played against me (even though I think my opponent was forced into all
those sacrifices by sheer desperation). And third, from the point of view of
opening preparation, the main thing is that Black could have had an easy
advantage after 14. ... Bxg4.
So, to sum up part (A) of the f4 Variation, we have seen that White can
get severely punished if he pushes his pawns too rapidly without developing his
queenside first. We've seen that f4-f5 followed by Bg5 (or Bh6, in the
Maser-Mackenzie game) is a dangerous plan for White, and so Black should do his
best to prevent the f5 break. That means he should reply to Bb5+ with the
unexpected ... Kf8! White must be very wary of threats to his
exposed white-squared bishop, and in fact he has to lose a tempo with Ba4 to
keep the bishop safe. Black has plenty of time to bring his king to safety with
... Ne7, ... Bf6, and ... Kg7. The game will then take on a positional
character, with White trying to exploit the static weakness of Black's doubled
d-pawns, while Black uses his space advantage (given to him by those same
doubled pawns) to create dynamic activity for his pieces, especially the knight,
which will go to f5 and then possibly to e3 or g3.
I was going to cover part (B) -- the retreat 8. Ba4 -- in this entry, but I
can see that the entry has already gotten extremely long, so I will stop here
and talk about 8. Ba4 next time.