Search Our Site
Search The Web

READER's letters:
war zone in the glek variation of the four knights

Josiah Stein shares the following:

In our recent annual Green Bay Open in Green Bay Wisconsin, master Alex Betanelli trashed expert Matt Conant with the following shocker! We had a large projection screen transmitting the moves.
Matt Conant (2030) vs. Betanelli (2300 something)
Board 1, Round 4
1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 4.g3 Nxe4



With only a class-A perspective I thought that Black was playing for a loss. Joining my friend Matt for a cigarette we just shrugged our shoulders and admitted that Alex had guts. Shortly after the smoke, I was discussing the game with another expert friend and he pointed out to me that there was a similar variation with colors reversed that was considered bad for White, however the extra move 4.g3 makes a ton of difference since the knight can't go to g3!
5...d5 6.Nc3 d4 7.Nb5 a6 8.Na3 e4 9.Ng1 d3!
White can't be better here, can he?
Not the kind of move Matt is used to making, but he already was wary of his position.
10...Qf6 11.c3 Ne5 12.Bh3 Bc5
White is now clearly lost. Strolling around looking at the games, I bumped into Alex who chuckled to me in his still present Russian accent, "Which extra white Knight is better?"
13.f4 exf3 14.Kf1 f2, 0-1.
Food for thought! 

Jeremy Silman replies:

This is all theory. The variation in question (with 1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 4.g3) is known as the Glek Variation (he also has a line in the KID named after him). One normal continuation is 4 d5 5.exd5 Nxd5 6.Bg2 Nxc3 7.bxc3, etc.

However, the move 4...Nxe4!? is extremely interesting, and has been used by grandmasters like Magnus Carlsen and Macieja.

After 4...Nxe4!? theory continues: 5.Nxe4 d5 6.Nc3 d4 7.Nb5


1) A tricky way to bail out is 7.Bg2 dxc3 8.bxc3 when we've transposed back into the line 4...d5 5.exd5 Nxd5 6.Bg2 Nxc3 7.bxc3.

2) 7.Nb1 is thought to be too passive, though Alexander Ivanov used it to win a 60-minute game in 2000. It was also the move of choice in a couple (low quality) postal games in 2004, so perhaps the Knight retreat has a future.

3) 7.Ne4 f5 8.Neg5 e4 9.Bc4 exf3 10.Bf7+ (10.Nxf3 Qe7+ 11.Kf1 Be6 12.d3 Bxc4 13.dxc4 Qd7 14.Kg2 0-0-0 was comfortable for Black (1/2 in 24) in Smirin Macieja, Czech Republic 2003/2004.) 10...Kd7 11.Be6+ Ke8 12.Bf7+ Ke7 13.Bb3 Kf6 14.Nf7 Qe8+ 15.Kf1 d3 16.Qxf3 Nd4 17.Qxd3 Nxb3 18.Nxh8 Qc6? (perhaps a bit too tricky. Instead, the simple 18...Nxa1 seems to be winning) 19.Kg1? (White had to try 19.Qxb3 Qxh1+ 20.Ke2 Qe4+ 21.Kf1 Be6 22.Qc3+) 19...Nxa1 and Black went on to win in Sengupta - T Petrosian, Kochin 2004.

7...a6 8.Na3 e4 9.Ng1!?


I haven't seen this played before, but it makes perfect sense.

A bad idea is 9.Qe2? since 9...Bxa3! 10.bxa3 0-0 simply favored Black in Al Modiahki - Hakki, Teheran 2001. The game was eventually won by White after Black missed many chances.

White's best move was thought to be 9.Nh4. However, after 9...Bxa3 10.bxa3 0-0 11.Bb2 Re8 12.Be2 Bh3 13.Bg4 Bxg4 14.Qxg4 Ne5 15.Qh3 Nf3+ 16.Kd1 Qd5 White's position doesn't seem like much fun to play.

9...d3 10.Qh5?!

I don't think this is correct. Critical is 10.cxd3 exd3 11.Bg2 intending to cover the e-file with Na3-c4-e3, and also getting ready to castle kingside after the g1-Knight is developed. After 11.Bg2 I rather like 11...Be6 with an unclear position.


Certainly the most natural, but is it best? Perhaps 10...g6!? deserves the nod. Then 11.Qh4 Be7! 12.Qxe4 (12.Qh6!?) 12...Bf5 13.Qe3 0-0 14.Bxd3 Re8 and White's in for a rough ride.


White cracks. Both 11.Qh4!? and 11.Nc4 were better.

11...Ne5 12.Bh3 Bc5 13.f4 exf3 14.Kf1 f2, 0-1.