Small Encyclopedia of Chess Openings, 3rd Edition (Sahovski Informator, 2010). 671 pp., $69.95. Reviewed by Dennis Monokroussos.
Last year, I reviewed the electronic version of the second edition of the Small Encyclopedia of Chess Openings (SECO 2), this time around we’ll examine the physical version of the third edition (SECO 3).
SECO 3 is a one-volume openings encyclopedia like Modern Chess Openings (MCO) or Nunn’s Chess Openings [sic - it's by Nunn, Burgess, Emms and Gallagher] (NCO), but in keeping with all the Sahovski Informator products (such as the Informant, which comes out three times a year) there is no verbal commentary; all the communication is with the standard chess symbols that have been in use for generations. There are advantages: language isn’t an issue and it’s easier to compress more material into a smaller space. The latter is especially important here, as even the “small” ECO still runs to 671 dense pages. (It’s small only with respect to the full ECO, which comprises five volumes, each of similar thickness.)
There are disadvantages too, of course. The lack of natural language makes the book relatively useless for understanding general themes in certain openings. An author can express in several sentences each side’s major plans in a given variation, but that’s not available to the SECO’s readers. Maybe one can abstract some of the key ideas by replaying all of the lines and footnotes for a certain variation, but that’s not a given. The lines might stop too soon, for one thing. Another difficulty is symbol choice – and this is true in ChessBase as well, with their symbols. Suppose a player makes a new move, which is signified by the symbol “N” (for “novelty”). Let’s say it’s not a great novelty, and as a result the player is slightly worse where he could have had equality. Normally one would append += (assuming the novelty was Black’s), but guess what? You can’t. You’re forced to choose one symbol or the other, either N or +=, but not both. You can put += after White’s next move if you’d like, but that’s unclear – perhaps White could have had a clear advantage with some other move, for all we know. So the unwillingness on the part of the Informant (and ChessBase) people to string certain classes of symbols together lessens somewhat the power of their symbol system to convey information.
Another important matter, even for a dense 671-page tome, is determining what material to include. Should the focus be on the state of the art in professional chess, or should variations that are fairly likely to occur in club play receive more attention than they would if SECO were meant for professionals only? How deeply should the lines go, and how many variations should be included when the main lines are complicated? And what about “solved” variations that are no problem if one knows the theory, but are practically impossible to solve on spec – should they be included or not? These are serious editorial challenges, and I’m not sure what the right answer is, if any. (Maybe SECO could be released in two editions: one for masters and up, the other for club players? That would be interesting, but is such a model commercially viable?)
With this background laid out, let’s have a look at some specifics.
Moller Gambit:
Last time around I protested that the most common continuation was ignored, and they repaired the omission! Here’s the line: 1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.c3 Nf6 5.d4 exd4 6.cxd4 Bb4+ 7.Nc3 Nxe4 8.0-0 Bxc3 9.d5 Bf6 10.Re1 Ne7 11.Rxe4 d6 12.Bg5 Bxg5 13.Nxg5 h6 14.Qe2 hxg5 15.Re1 Be6, and now SECO 2 gave only 16.Re3 g4 17.h3 c6 18.dxe6 f5 19.hxg4 d5 20.Bd3 Qd6 21.Qf3 Qh2+ 22.Kf1 f4 23.Re5 0-0-0 unclear, as played in the game A. Perez - J. Olivera, Cuba 1998. In my review I noted that White far more commonly tries 16.dxe6, when after 16…f6 17.Re3 c6 18.Rh3 Rxh3 19.gxh3 g6 Black is in good shape, especially if he can play …Kf8-g7 without White achieving something major in the meantime. Lo and behold, SECO 3 gives just this line, continuing 20.Qf3 Qa5! 21.Rd1 Qf5! with a clear advantage for Black as in Silver-Matsuura, Brasil [sic] 1998, from Informant 73.
This is a definite improvement, but I wonder about the Perez – Olivera main line. If Black is clearly better after 16.dxe6 f6 17.Re3 c6, why doesn’t he just meet 16.Re3 with 16…c6? White doesn’t seem to have anything better than 17.dxe6, but after 17…f6 we’ve just transposed to Silver-Matsuura. (Indeed, in games since 1998 Black has chosen 16…c6 pretty regularly, with excellent results.)
This is a backwater of theory, at least for the big boys, that’s true. Still, I regularly examine how sources treat the Moller Gambit, for at least three reasons. The least interesting for purposes of this review is that it’s a line I’ve followed since I was a kid, so I'm always curios to see if anything new has happened. Second, it’s pretty common in blitz and at lower levels, so it’s fairly important for club players that the line be covered correctly. Third, it’s not difficult to cover it correctly. The key ideas have been known for a pretty long time now, so I see it as a test of competence for an openings source. If they can’t get something this simple and well-defined correctly, then chances are there will be trouble elsewhere.
Various Najdorf Sicilian Lines:
In my review of SECO 2, I complained that its coverage of the Perenyi line in the 6.Bg5 Najdorf (1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bg5 e6 7.f4 Be7 8.Qf3 Qc7 9.0-0-0 Nbd7 10.g4 b5 11.Bxf6 Nxf6 12.g5 Nd7 13.f5 Nc5 14.f6 gxf6 15.gxf6 Bf8 16.Rg1) was too superficial. In particular, I wrote that “only 16...Bd7 is covered, and after 17.Rg7 Bxg7 18.fxg7 Rg8 19.e5 0-0-0 20.exd6 only 20...Qb7 is given, not the probably better 20...Qb6. But more importantly, 16...h5 is omitted, and that was an important move even back then [in 2007, the publication date for SECO 2].” In the third edition, 20…Qb7 is still the only 20th move given in the 16…Bd7 line, but now 16…h5 is covered and given as the main line.
After 16…h5, 17.a3 is the main line, when 17…Rb8 18.e5 Bb7 19.Qg3 d5 20.Re1 Qb6 is given as unclear, as in the game A. Stuart – D. Pillhock, corr. 1994, while 17.Rg7 is also given, with the continuation 17…b4 18.Nd5 exd5 19.exd5 Nd7 20.Nc6 Bb7 21.Bh3 Bxc6 22.dxc6 Ne5 23.Bd7+ Kd8 24.Qe4 Qb6! 25.Be6 Kc7 26.Bf7 Nc6 27.Bd5 Bxg7 28.fxg7 Rhg8 with a clear advantage to Black (Vasquez Schroeder-Vallejo Pons, Tripoli (m/5-blitz) 2004.
This latter game was important, but White can improve with 27.Kb1, which is equal. More importantly, after 17.a3 Rb8 18.e5 Bb7 19.Qg3 d5, the move 20.Re1 is completely out of fashion. In fact, it was never in fashion, except that it was the very first try, in the 1994 correspondence game cited by SECO 3. In the 27 games to reach the position after 19…d5 since then, only once more was 20.Re1 tried; the overwhelming choice is 20.Kb1, which has scored very well and is the computer’s first choice. Maybe there aren’t any games in the Informant with this, but it would have stood out with a routine check of other databases.
Turning to another Najdorf line, the Poisoned Pawn, there was absolutely no coverage of 10.e5 in SECO 2, and the omission has been rectified. Likewise, the Byrne Variation (6.Be3 e5 in the Najdorf) was badly undercovered in SECO 2, but is covered in far more detail this time around. Still, it’s inadequate for serious preparation. Let’s take the Vallejo Pons Variation (as Milos Pavlovic labels it in his brand new The Cutting Edge: Sicilian Najdorf 6.Be3). The variation, which is reached after 6.Be3 e5 7.Nb3 Be6 8.f3 Be7 (Black’s last two moves are given in reverse order in SECO 3) 9.Qd2 0-0 10.0-0-0 Nbd7 11.g4 b5 12.g5 b4 13.Ne2 Ne8 14.f4 a5 15.f5, takes up 18 pages in Pavlovic but just one row on the main page and three footnotes in SECO 3. One can view the depth of contemporary opening theory as luxurious or nauseating, but it is what it is, and you simply can’t play a variation like this against a prepared opponent with a third of a page’s worth of game excerpts.
Marshall Gambit:
Turning to another important line, one not covered in my previous review, let’s see how SECO 3 handles the Marshall Gambit, which is popular at all levels. It would be unfair to compare SECO 3 with Jan Gustafsson’s recent DVD on the Marshall (because the latter is more recent and the product of an elite GM’s secret workshop), but let’s see how it fares against the worthwhile works by Milos Pavlovic (Fighting the Ruy Lopez, Everyman 2009) and David Vigorito (Understanding the Marshall Attack, Gambit 2010).
A first observation is that SECO 3 doesn’t even MENTION Marshall’s 11…Nf6 or the recently revived 11...Bb7 (after 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 0-0 8.c3 d5 9.exd5 Nxd5 10.Nxe5 Nxe5 11.Rxe5). Neither line is setting the world on fire, but most sources (including Pavlovic and Vigorito) at least mention the Capablanca-Marshall game in reference to the first line, while players like Short, Kamsky, and correspondence SIM and Marshall specialist Tim Harding have tried the second.
Turning to the old main line (11…c6 12.d4 Bd6 13.Re1 Qh4 14.g3 Qh3 15.Be3 Bg4 16.Qd3 Rae8 17.Nd2), I checked its coverage of 17…Re6 against Pavlovic and Vigorito. Generally speaking, it has just about all the game references it should (always, or seemingly always, in this section, to games published in the Informant), but to really understand what’s going on one should have access to analyses beyond the bare game scores. Perhaps these are present in sufficient detail when one finds the games in the Informant – but of course you must have the relevant issues of the Informant in the first place!
Still, this is the Small ECO, and it just can’t have everything. I did notice one omitted game that should have been included, however. After 18.a4 Qh4 19.axb5 axb5 20.Nf1 Bf5 21.Qd2, SECO 3 gives 21…Rfe8 and 21…Be4, and rightly so. Both moves are important. The line with 21…Rfe8 is continued through move 30, giving the drawn game Svidler-Jakovenko, Foros 2008 in full. 21…Be4, on the other hand – with a game citation – is the extent of SECO 3’s coverage of the latter move. This isn’t too helpful, as it’s a very tricky option and both Pavlovic and Vigorito provide a fair amount of coverage. The game reference in SECO 3 is to a 1989 correspondence game Rubinchik-Vitomskis, and surprisingly for a 22-year-old game going back to the pre-computer era (meaning before strong engines were available to the general public), the analysis by Vitomskis and the late Latvian GM Alvins Vitolins holds up pretty well.
What isn’t present in SECO 3, however, is any mention of 21…Bh3, which led to a speedy win for Black in Shirov-Aronian, Bilbao 2009. SECO’s publication date is 2010, but what that really means depends on the publisher. Chess Stars is perhaps the very best in this respect: although their books must first be translated into English and shipped from Eastern Europe, and even so their books regularly go from the author's keyboard to the buyer's hands within a couple of months. There are other publishers, some of very high quality, whose works come out at least six months or so after the author finished writing.
Another gap in SECO 3’s coverage of the Marshall is the line 12.Re1 Bd6 13.g3 Bf5. In fact it doesn’t even mention 12.Re1, but does give 12.g3 Bd6 13.Re1 which transposes. There’s not much about this – only 13…Re8 is given, with Anand-Aronian, Morelia/Linares 2008 cited through White’s 18th move. 13…Bf5 is not mentioned at all, even though it’s an up-and-coming line that has been known since well before the book’s publication date. (See for instance the spectacular game Naiditsch-Gustafsson, European Ch. (Dresden) 2007, 0-1 in 25 moves of home prep.)
Finally, here are two shorter examples, from different openings. First, browsing the section on the Zaitsev Ruy, I saw the famous game Anand-Adams from San Luis 2005 mentioned in a footnote. Amusingly, it only gives 23.Qd2!?, with an unclear position, which is about as helpful as telling a starving person that food exists somewhere. Even a player of Adams’ extremely high caliber was helpless when faced with this, and after a single error he was crushed by a brilliant sacrificial attack. Again, if one has the Informant (or another relevant source) they can see the antidote, but "23.Qd2!? unclear" is a lousy place to leave the reader.
Example #2 is from the Botvinnik System. I remembered van Wely-Smeets, Dutch Ch. 2005 as a famous and huge improvement for White over what was considered a draw for Black based on the (also) famous game Azmaiparashvili-Shirov, Madrid 1996. Curious to see if the van Wely game would be mentioned, and if there was any subsequent improvement, I looked and saw 25…Bf5! 26.f3! with compensation for the material, referring to the van Wely game and sending interested readers to – you guessed it – the relevant issue of the Informant.
In fact, I was quite interested to do so, because “compensation for the material” did not strike me as the way things were understood at the time. Yes, it’s true that White had compensation for the material, but (exaggerating somewhat) this was like saying that McDonald’s had sold several thousand hamburgers. Well, yes, and a few billion more besides. To say that a player has “compensation” for a material disadvantage, without any further explanation, should suggest that the compensation is roughly sufficient, such that one cannot reasonably assert that either side has a substantial edge. This game, however, was taken to be something like a bust of Black’s play, so such a neutral evaluation was big news to me. So I started checking sources.
First, I looked up the original Informant notes by van Wely himself, and he doesn't give the "with compensation" evaluation, nor does his analysis offer any grounds for thinking that Black is okay. This squares with my own pretty deep analysis of the game around the time it happened, and it is further corroborated by GM Christopher Lutz in ChessBase Magazine 110 and David Vigorito in Play the Semi-Slav. What about subsequent practice? It turns out that six other masochists have tried this with Black, and all six wound up with the same big fat 0 for their troubles. No points to SECO 3 on this one!
In summary, one simply cannot use this volume as their one and only reference work for the opening. Club players won’t get the guidance they need to understand what to do, while stronger players focused on trendier lines will only find barebones outlines. That doesn’t mean that such a book is without value. It can help a player figure out what lines are considered more critical and what the key games are, and that’s a real time-saver. Also, there are many lines where it’s enough to know what to do up to around move 15-20, where the alternatives aren’t all fraught with tactical danger, and in those cases a quick browse of SECO 3 may well suffice.
In conclusion, there’s really no point in a lower-level club player buying such a book: there’s too much detail and too little guidance. If you’re below 1550, you’re much better off buying a book with a lot of “talk”, focusing on generalities, like Paul van der Sterren’s Fundamental Chess Openings or Carsten Hansen’s Back to Basics: Openings. Above that level, while it shouldn’t be used as a standalone source or the last word on any line, it could prove useful as a first source, to get a general idea of what’s important and to start organizing one’s research.
(For U.S. buyers who have made it this far and might (still) be interested, ordering info is here.)