Page1 l Page2

Joseph asks:

I recently played an OTB tournament game last weekend as White. It started:

1.e4 e5 2.Nf3 and then my opponent played 2...f5!?

What nonsense is this?

The game continued 3.Nxe5 (I figured taking this pawn was best.) 3...Nc6!? (More nonsense, leaving the diagonal open.) 4.Qh5+ g6 5.Nxg6 (Can't take the knight because the rook hangs.) 5...Nf6 6.Qh4 hxg6!?!? 7.Qxh8 Qe7. I figured the queen might get trapped and so played 8.Qh3. Eventually I got my other pieces out, consolidated my material and won but it was quite a struggle and quite frightening to see my opponent play these bizarre opening moves with such confidence.

What is the point of 2...f5?

Silman replies:

The line you faced is known as the Latvian Gambit. There is a whole community of players who live for this opening, but the fact that no grandmasters use it speaks volumes for its true lack of soundness.

Nevertheless, it is tricky, and offers Black good practical chances against an unprepared opponent. Though various “refutations” have appeared, fans of the gambit always find ways to keep Black alive (or so they claim), until gambit debunkers come up with new ways to kill it. Then Latvian aficionados fix it again, the debunkers kill it, and…well, it's a never-ending cycle.

As for the line Joseph faced (1.e4 e5 2.Nf3 f5 3.Nxe5 Nc6), this actually has some fans since the rapid development Black often achieves gives rise to complications that are far from easy to solve. For example, in Joseph's game 4.Qh5+ g6 5.Nxg6 Nf6 6.Qh4 hxg6 7.Qxh8 Qe7 leads to a very difficult position where Black has definite compensation (Even if it's not quite enough, Black's practical chances are excellent.).

White's best (instead of Joseph's 8.Qh3) is 8.d3, which scores pretty heavily for White. Nonetheless, John Nunn's 4.d4! seems to put 3…Nc6 on ice (see the Watson analysis below).

Below is an abridged version of IM John Watson's analysis of White's most testing line. It appears in the new book, SURVIVE & BEAT ANNOYING CHESS OPENINGS by Schiller & Watson. John was good enough to let me share it with jeremysilman.com readers.

Personally, I feel the Latvian is just bad. True fans of the gambit should consider this a challenge. If, after examining Watson's apparent refutation, they feel that the Latvian Gambit still lives, write me and together we can search for the truth of the matter (my “truth” being that it deserves to die).

JOHN WATSON'S ANALYSIS:

1.e4 e5 2.Nf3 f5

The Latvian Gambit has a long and storied history in chess literature. Although it is almost impossible to find in the repertoire of a professional player, amateurs, and especially correspondence players, have long found the tactical labyrinth of the main lines to be highly appealing.

Taking on the Latvian in the main lines requires a bit of knowledge and familiarity with the critical lines is essential. On the other hand, there really isn't that much that you have to memorize in order to get the upper hand.

3.Nxe5

The straightforward move, and the favorite of top masters. White grabs an important center pawn and challenges Black to get it back without compromising his position. The first threat is 4.Qh5+.

3…Qf6

Black has five, rarely encountered, alternatives:

1) 3…Nc6 (Black aims for quick counterplay after 4.Nxc6 dxc6, but 4.d4! is the line that John Nunn suggests will refute it. This move emphasizes center and rapid development. Black will normally choose one of the following moves:

1.a.) 4…fxe4 5.Nxc6 dxc6 6.Qh5+! forces the King to move, in view of 6…g6? 7.Qe5+.

1.b.) 4…Nxd5 5.dxe5 d6 (5…Qe7 6.Qd4 and White is a pawn up, threatening Nc3-d5 – Nunn) 6.Bf4!, and at this point Dzindzichashvili says, “Black's position flat out sucks!”

1.c.) 4…Nf6 5.Nxc6 dxc6 6.e5 with a clear extra pawn.

1.d.) 4…Qh4!? 5.Nf3 Qxe4+ 6.Be2 and Black's Queen is exposed to further attack on the e-file, e.g., Nunn gives 6…Nf6 7.0-0 Be7 8.Re1 with a clear advantage for White and 8…0-0? loses to 9.Bc4+. This is easy, straightforward chess!

2) 3…Nf6 4.Bc4 Qe7 (Black's other try is 4…fxe4 5.Bf7+! Ke7 6.Bb3 d5 7.d4 [7.0-0!?] 7…Qd6 [7…exd3 8.0-0! And Re1] 8.Bf4 Qe6 9.c4! c6 10.Nc3 Qf5 11.Qd2 and Black's King placement makes his position a real mess, Grott - Scott, corr. 1989) 5.d4 d6 when a cut idea is 6.Bf7+ (6.Nf7 Rg8 7.Nc3 fxe4 8.Nd5 Nxd5 9.Bxd5 is also good) 6…Kd8 7.Bb3 dxe5 8.dxe5+ Nfd7 9.e6 regaining the piece with wide open lines against Black's exposed King. Bailing out with 9…Qd6 10.Bg5+ Be7 doesn't help after 11.Qxd6 cxd6 12.Bxe7+ Kxe7 13.exd7 Bxd7 14.Nc3 with a pawn and attack.

3) 3…d6 4.Qh5+ g6 5.Nxg6 Nf6 6.Qh4 when Black doesn't have enough compensation after both 6…Rg8 7.Nxf8 and 6…Nc6 7.Nxh8 Nd4 8.Na3.

4) 3…Qe7 4.Qh5+ (4.d4 is also good) 4…g6 5.Nxg6 Qxe4+ 6.Be2 Nf6 7.Qh3+. With g2 covered, white has a solid advantage, for example 7…hxg6 (7…Rg8 8.Nc3 Qc6 9.Nxf8) 8.Qxh8 Qxg2 9.Rf1 Ng4 10.d4 or 10.Bxg4 Qxg4 11.d3, winning.

5) 3…fxe4? 4.Qh5+ g6 5.Nxg6 Nf6 7.Qe5+, winning.

4.Nc4 fxe4 5.Nc3

As is so often the case, the emphasis is on straightforward development.

Now Black has A) 5…d6?, B) 5…c6?!, C) 5…Qf7, D) 5…Qe6, E) 5…Qg6.

A) 5…d6? 6.d3!

6.Qh5+ g6 7.Qa5 is also good, as given by Kosten.

6…exd3 7.Bxd3 Qf7

7…Be6!? is suggested by Kosten, but White can just play 8.0-0 and now what? For example, 8…Nd7 9.Ne4 Qe7 10.Ng5.

8.0-0 Be7 9.Re1 Nf6 10.Nb5! threatening both 11.Nxc7+ and 11.Ncxd6+.

B) 5…c6?! 6.Nxe4 Qe6 7.Qh5+! g6 8.Qe5 Qxe5 9.Nxe5 d6 10.Nf3

Even better is 10.Nc4! d5 11.Ned6+ Kd7 12.Nf7 dxc4 13.Bxc4.

10…d5 11.Nc3 and Black has no compensation whatsoever for the pawn, Kozlov - Grivainis, corr. 1977.

C) 5…Qf7!?

This is an odd move with the point that 6.Nxe4? will be met by 6…d5, but the blockading move 6.Ne3! stops …d5 and threatens Bc4, so Black must make another non-developing move.

6.Ne3! c6

6…Nf6 7.Bc4 Qg6 8.d3! breaks Black's position down.

7.d3 exd3 8.Bxd3 d5 9.0-0

Dzindzichashvili likes 9.Qe2 for White, but we'll stick with castling. There are six moves worth examining for Black, but none solve the problems caused by Black's pathetic development. White's idea is attack, attack, attack!

9…Bd6

Trickiest, though Black has also tried:

C.1) 9…d4 10.Bc4 Qd7 11.Qh5+ Kd8 12.Rd1 or even 12.Ne4! dxe3 13.Rd1.

C. 2) 9…Nf6 10.Re1 Be7 11.Nf5.

C. 3) 9…Bc5 10.Na4 (10.Ng4!?) 10…Bd6 11.c4 d4 (11…Ne7 12.Nc3 Be6 13.cxd5 cxd5 14.Nb5! Nc8 15.f4! with a terrific attack that Kosten likes for White; 11…Be6 12.Nc3 Ne7 13.cxd5 cxd5 14.Nb5 transposes to 11…Ne7.) 12.Nc2 c5 13.b4! b6 14.Be4 Bb7 15.Bxb7 Qxb7 16.bxc5 bxc5 17.Nxc5! winning, Kozlov - Macgee, corr. 1989.

C. 4) 9…Bb4 10.Ncxd5! cxd5 11.Nxd5 leads to a winning attack.

C. 5) 9…Be6 10.f4 (10.Re1!?) 10…Bd6 11.f5 Bd7 12.Ng4! Ne7 13.Bg5! with an attack.

10.Re1 Ne7 11.Nexd5! cxd5 12.Nb5

Page1 l Page2