Zaven Andriasyan, The English Attack Against the Taimanov Sicilian: A Guide for White (New in Chess, 2015). 175 pp., $24.95/€21.95. Reviewed by Dennis Monokroussos.
The "English Attack", which could as easily and with more propriety be called the Byrne Attack, is the generic name given to the White attacking formation in the Open Sicilian with the moves Be3, f3, Qd2, 0-0-0 and (generally) g4. Byrne played this against the Najdorf (with ...e5) in the early '70s, and then English players took up the mantle in the '80s against the Najdorf/Scheveningen approach with ...a6 and ...e6. It has since been used against the Taimanov as well, so with the exception of a chapter on early divergences the book offers a white repertoire in the variation 1.e4 c5 2.Nf3 e6 3.d4 cxd4 4.Nxd4 Nc6 5.Nc3 Qc7 6.Be3 a6 7.Qd2 Nf6 8.f3 Bb4 9.0-0-0.
The first thing to say is that this is a committal move order by White, one which has become less popular nowadays. It's more common to start with 8.0-0-0, waiting for Black to commit himself first. If Black plays 8...b5, then 9.Bf4 has been much more successful than transposing to the English Attack with 9.f3. If Black plays 8...Be7, then White may eschew 9.f3 for 9.f4, though there's nothing wrong with the former move. In case of 8...Bb4, then everything goes back to normal and White plays 9.f3.
To his credit, Andriasyan offers some coverage of 8.0-0-0 in the early divergences chapter, but only considers 8...Be7 9.f4; the 8...b5 line isn't addressed. Is this such a big deal? Not if we're talking about geopolitics, relationships, or the saving of one's soul; in the context of creating a strong and flexible repertoire against the Taimanov Sicilian, however, it matters. One thing I almost always do when examining a new opening book for one side is to see what it says about the recommendations from the previously most recent book advocating for the other side. So what does Andriasyan have to say about Alexander Delchev's and Semko Semkov's 2014 work The Most Flexible Sicilian or John Emms' 2012 book The Sicilian Taimanov: Move by Move?
There are plenty of places where the works overlap, but there are important gaps. Emms focuses on the move order with 8.0-0-0, but sure enough, when White plays 8.f3 he recommends 8...b5, and after 9.0-0-0 he goes for 9...Ne5. As far as I can tell, there's no place where this transposes to anything in Andriasyan's book, so here, White players, you are out of luck unless you revert to an 8.0-0-0 move order (or do some further work to be ready for 8.f3 b5 9.0-0-0 Ne5).
Delchev & Semkov (D/S) likewise recommend 8...b5, but to the relief of Andriasyan's readers their variation will eventually merge to a line covered in the new book, after 9.g4 Bb7 10.g5 Nxd4. So all is well, and will this take care of the problem in the Emms book as well? Not so fast.
Both D/S and Emms note that Black can also play 9...h6 (rather than 9...Bb7), and this won't transpose to Andriasyan. That's not to say that 9...h6 is best. After 10.Nxc6 both Emms and D/S think White has an edge after 10...dxc6, as Black's h-pawn gives White a hook for his attack, but D/S does think that 10...Qxc6 might not be so bad, intending to continue with ...Bb7, ...Qc7, ...Be7 and ...d5.
Emms fans might also try another of his suggestions to avoid getting pulled into Andriasyan's repertoire, and that's 8...Ne5 (not move 9, but move 8). If White plays 9.f4 he'll have wasted a tempo relative to 8.0-0-0 Ne5 9.f4, and there's even the wild 9...Bb4 (after 8.f3 Ne5 9.f4). (The latter is very warmly embraced by the computer.)
One might wonder why Andriasyan gives 8.f3 instead of 8.0-0-0, as it seems to give Black more options while taking away White's. I'm not sure I have an answer to this, as White almost invariably ends up castling on move 9 against any sane move. Here are two hypotheses. First, White is concerned that 8.0-0-0 would allow 8...Ng4; second, White might get a little headstart on his attack in the line 8.f3 Be7 by playing 9.g4.
Against the first hypothesis: no one plays 8.0-0-0 Ng4. There are 22 games where that was played (out of 2977!), most of them more than a decade ago, and there's a reason for this. After 9.Bf4 Nge5 (9...e5?! 10.Nd5 Qd8 11.h3!+/- as in Arizmendi Martinez-Collutiis, Valle d'Aosta 2003) 10.Bg3 Nxd4 11.Qxd4 f6 12.f4 Bc5 13.Qd2 Nf7 14.e5 (Luther-Banikas, EU Cup 2002 and five other games) Black's position stinks. As per the second, you could make a case for that if that was what Andriasyan recommended after 8.f3 Be7, but he doesn't - he follows the 9.0-0-0 crowd.
The foregoing is therefore a bit disappointing, or at least mysterious, but as White can just reverse his 8th and 9th moves the problem is easily solved. How does the book fare in the main lines, where the overlap is obvious? Here I don't want to go into too many details - there ought to be a reason to buy the books, after all - but I will offer some conclusions. Often Andriasyan refers to the same games as D/S and/or Emms, but they disagree slightly about the assessment, at least as a practical matter.
For instance, in the line 8.0-0-0 Be7 9.f4 b5 10.e5 b4 11.exf6 bxc3 12.Qxc3 Bxf6 13.g4 h6 14.h4 Bb7 15.Rh2 Rc8 16.Qd2 D/S recommends 16...Bxd4 as "perhaps safest", and they follow the game Saric-Yu, Wijk aan Zee 2014 through move 27 and say "Black has no problems in this endgame". Andriasyan gives a couple more moves from the game and offers a different conclusion: "An absolutely equal position, according to the computer, but to me White seems better, a view supported by the fact that the very strong Chinese GM [DM: 2677 at the time of the game] lost this position eventually". There is something to this sort of evaluation, but it must also be acknowledged that one can examine the further course of the Saric-Yu game to improve on Black's play. Black went seriously wrong on his 33rd and especially 34th moves, and it seems that if he keeps one rook on the 8th rank he is probably fine. One idea is 33...Bf1, aiming to swing the bishop around to f5.
Similarly, Emms also examines the variation given above through 16.Qd2, but proposes 16...Be7, saying it looks about equal. Again Andriasyan goes further, taking the analysis another 14 moves - all of it original - and ends in a position about which he suggests that although White is a pawn down in a 4th phase position (both sides have a queen and both rooks), White's attack gives him good practical chances despite the engine's claim of equality. In this case, I agree entirely, and since his analysis is original White players will have a big practical edge over those relying on Emms.
Another important difference-of-opinion/practical-chances-vs.-computer-evaluation line is 8.f3 b5 9.0-0-0 Be7 10.g4 Nxd4 11.Bxd4 Bb7 12.g5 Nh5 13.Be5 Qxe5 14.Qxd7+ Kf8 15.Qxb7 Bxg5+ 16.Kb1 Qb8 17.Rd7 Qxb7 18.Rxb7 Bf6 19.a4 Bxc3 20.bxc3 bxa4. Through here Andriasyan and D/S are following an important game Ter Sahakyan-Polgar, Yerevan 2014, and both works offer improvements for their respective sides along the way. In particular, Andriasyan proposes and investigates 21.Bc4 as a possible improvement over Ter Sahakyan's 21.Bxa6, and after 21.Bxa5 g5 22.Bb5 Kg7 (Andriasyan(!) also takes a deep look at 22...Nf4) 23.Rd1 D/S offers 23...Nf4 as an improvement on Polgar's 23...Rhb8. Sticking to Polgar's move, Andriasyan follows the game through White's 26th move, again acknowledging the engine's claim that the position is equal but insisting once again that things aren't so clear:
[S]uch positions are extremely difficult for Black to play, especially when a pair of rooks have been exchanged. The c-pawn becomes very dangerous. I would not recommend Black to enter this position, where he has to play very accurately for a long time, just to make a draw.
Indeed, Polgar did not succeed in making a draw.
Finally, after the moves 8.f3 Bb4 9.0-0-0 Ne5 10.Nb3 b5 Andriasyan has two chapters considering four possible White moves, 11.Bd4 and 11.Qf2 in one chapter, 11.Kb1 and 11.Qe1 in the next. In both cases he thinks Black can equalize after the first move with precise play, but thinks the latter move in each pair rectifies the problems with the former move and offers White an advantage. His analysis of 11.Qf2 looks quite decent, and that's where I would steer his followers for now, because in the 11.Qe1 variation there seems to be an important omission.
After 11.Qe1 Be7 12.f4 Ng6 13.e5 Ng4 14.Ne4 Nxe3 15.Qxe3 there are 33 games in the database, and in 30 of them Black castled short. Inexplicably, Andriasyan only covers 15...Nh4, which was played just once and by a comparatively lower-rated player (2388) against an elite GM (Jakovenko, rated 2753). The move wasn't very good and isn't recommended by the computer, and unsurprisingly Black got crushed. The omission of 15...0-0 is a serious one. To be fair, Andriasyan does cover 14...0-0, and this is more common than 14...Nxe3. (Unfortunately, it doesn't transpose because White meets 14...0-0 with 15.Bc5.) But 14...Nxe3 is still an important move, and it's also the move covered by Alejandro Ramirez in his ChessBase DVD on the Taimanov. (More bad luck for club players hoping their source will beat their opponent's source.)
It is time to conclude the review. If you are a serious player who uses the Taimanov on a regular basis, you will want this book, if only out of self-defense. And if you're a serious player looking for a way to meet and beat the Taimanov, this book will help. There is a lot of original analysis, and as we've seen Andriasyan is a player with a strong sense of when to trust the computer's evaluations and when not to. It is a worthwhile book, notwithstanding the gaps and the 8.f3 vs. 8.0-0-0 problem. Recommended to players rated 2000 and up with a serious interest in the material covered.