Reviewed by John Watson

Everyman's OFFBEAT KING’S INDIAN (“OKI”) should provide a complement to Gallagher’s new PLAY THE KING’S INDIAN by presenting in detail what Gallagher didn’t have room for, i.e., the numerous side variations that White can choose and Black’s strategies against them. Normally back cover copy shouldn’t be taken seriously, but the claim “All unusual King's Indian Defense lines are covered” is particularly ironic since only a couple of the many truly unusual lines are covered. The entire contents of the book are devoted to only 4 variations: (1) the Averbach, a traditional, main-line Grandmaster opening whose theory has been studied extensively for 50 years and is hardly “offbeat”; (b) the “Makogonov System” 5.h3 (after 1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 d6), a variation which is fairly popular today. For some reason the related Bagirov variation with 5.Nf3 0-0 6.h3 has not been included; (c) 5.Bd3 (as practiced by Seirawan), a sort of “mainstream irregular” mixture that was more popular a few years back when there were also more King’s Indian players; (d) 5.Nge2, which is rare enough to be called “offbeat” as long as it doesn’t transpose into other variations.

Here’s a short list of what would be missing from a fairly complete treatment of irregular lines, taken from the chapters of my own book, THE UNCONVENTIONAL KING’S INDIAN (“UKID”) published in 1997:

1. All seldom-played Bg5 systems, including [after 1.d4 Nf6 2.c4 g6 3.Nc3 Bg7] 4.Bg5 (Smyslov); 4.e4 d6 5.Bg5; and 4.e4 d6 5.Nf3 0-0 6.Bg5 (Zinnowitz). Even the Torre with 1.d4 Nf6 2.Nf3 g6 3.Bg5 would be useful.

2. The above-mentioned 5.Nf3 0-0 6.h3.

3. All Bf4 systems, including 1.d4 Nf6 2.Nf3 g6 3.Bf4 Bg7 (London) with 4.e3 or 4.c3 intending Nbd2); 1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.Nf3 (4.Bf4) d6 5.Bf4; and 1.d4 Nf6 2.Nf3 g6 3.Nc3 d5 4.Bf4 (Barry).

4. All irregular g3 systems, including at least 4 double fianchettos variations (g3 and b3) with and without c4; lines like 1.d4 Nf6 2.Nf3 g6 3.g3 Bg7 4.Bg2 0-0 5.0-0 d6 6.Nc3 (the Martinowsky System, difficult for Black to meet); and lines with g3, e4, and Nge2.

5. A host of other ideas such as: attack by 1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 d6 5.Be2 0-0 6.g4 and 6.h4, or here the positional 6.Be3, and various Colle Systems with an early e3.

So the subject matter is limited but that just means that we get a specialized book about the four systems. The authors Panczyk and Ilczuk (“P&I”) state their philosophy in the Introduction: “We have tried to demonstrate how to treat this or that scenario, as well as what to avoid. This is not a compendium of knowl­edge about all the systems but a guide to the ideas which may occur, a manual that explains strategic principles and eases the journey around tactical themes.”

We’ve heard that one before. But the book turns out to be primarily a set of games with a dense forest of game fragments used as subvariations. I don’t see much truly useful advice on strategy, although the authors do provide commentary and descriptions about several types of positions. When it comes to analysis, most of the lines come from games themselves. So, for example, we get a fragment that favors White, and in the next sentence the authors say “but [move x] would have improved for Black.” Some moves follow and it turns out that they are not independent analysis but rather from another game, and that game was improved upon by another game. A strange way to present theory, because the “best” moves are assumed to come from a chain of earlier games not necessarily touched by analysis. Panczyk and Ilczuk do suggest various continuations for some lines. Apart from using the 1- or 2-move unexplained parenthetical entities that characterize opening books these days, they also make a stab at presenting some original suggestions and they evince true interest in investigating some positions. But too many times we are left with only a game fragment that is so badly played as to be meaningless.  In the end I think that Panczyk and Ilczuk do the opposite of what’s promised in their Introduction. They use the considerable space at their disposal to list games and to present a thorough and well-organized compilation of the actual practice of these systems. Unfortunately, I don’t find much thematic unity to this dense flow of examples, and one gets the impression of a lot of database manipulation and a lack of depth.

Let’s look at a few examples. I don’t understand why authors generally don’t make a dedicated search for other books dealing with their subjects, but they usually don’t. P&I list only 4 books with the Encyclopedia E, two 1980 books of no real value for these variations, and a Polish book called “Chess from A to Z.” But the theory on the Averbach is extensive and full of ideas. Everyman’s predecessor published a book by Petursson on the Averbach, New In Chess Yearbook lists a series of articles on it and articles abound. NICY also lists no less than 13(!) surveys on 5.h3. OKI’s publisher Everyman put out a whole book on 5.Nge2 by Forintos and Haag (2000) and NICY has an article on 5.Nge2. NICY has a feature article listed on 5.Bd3. And there are repertoire books that have to address these lines such as the ones mentioned above and books by Martin, Marovich (2 books) and others. My guess is that I’ve listed a very small fraction of the outstanding literature and none of it is listed. It would save the reader a lot of time to confirm that material from various sources had been incorporated or not. I’ll compare two lines of OKI with the two older books mentioned above, i.e., Gallagher’s from 1996 (updated by his new one) and mine from 1997. The Offbeat King’s Indian’s first chapter on 5.Nge2 has a mass of game examples, but doesn’t address the literature. Sometimes this changes the importance of a line. For example:

1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 d6 5.Nge2 0–0 6.Ng3 c5 7.d5 e6 8.Be2 exd5 9.exd5 can be met by 9...Ne8 10.h4!? Nd7 11.h5 f5 12.hxg6 hxg6 13.Bh6

Now OIK gives only 13...Bxh6(?) 14.Rxh6 Qg5 15.Qd2! Qxd2+ 16.Kxd2 Kg7 17.Rah1 with advantage, citing a Topalov game. But my book and even earlier Marovic’s PLAY THE KING’S INDIAN gave 13...Ne5, = Marovic, 14.Qd2 Nf7 15.Bxg7 (15.Be3 Bd7 16.a4 Rb8, =) 15...Kxg7, e.g., 16.0–0–0 Qg5 17.Qxg5 Nxg5 with moves like ...Bd7, ...Nc7 and ...Rh8 to follow. We both think that this is equal.  Regardless of the verdict, this was the line to investigate.

In the Bd3 variation P&I miss the key resource in one of the main lines:

1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 d6 5.Bd3 0–0 6.Nge2 Nc6 7.0–0 e5 8.d5 Nd4 9.Nxd4 exd4 10.Nb5 Re8 11.Re1

Now after 11...Ng4 12.h3 P&I give only 12...a6 13.Nxd4 Nxd5 and 12...c6, both better for White and therefore leading one to think that Black should avoid the line. But it doesn't take that long an investigation to realize that 11...Bd7! gains a key tempo in the line 12.Nxd4 Nxd5. I gave this with games and extensive analysis (including 12th move alternatives) in 1997. In this year’s PKI Gallagher says that 11...Bd7! “is a much better move and to be honest I hadn’t realized that Black could get such an easy game in such fashion.” He quotes 2 games to support that. Remember that Gallagher and I are devoting far less space than P&I to variations like 5.Nge2 and 5.Bd3; this again illustrates the danger of grabbing games out of a database without giving them enough serious thought.

Panczyk and Ilczuk do their best job with the Averbach Variation (5.Be2 0-0 6.Bg5). From what I see the authors have put more of their energies into those chapters. They draw upon hundreds of games, as in other sections. But this variation isn’t an “offbeat” one and has thus produced years and years of good games based upon accumulated playing experience and analysis. Thus games played with it tend to be more sophisticated and reliable than in the other variations in OKI.

Gallagher recommends the 6...Na6 variation, which I have both played and taught. Both he (in a 12-page chapter) and P&I (in 19 pages) offer an interesting layout of material. I found good coverage and some interesting differences in the line with 1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 d6 5.Be2 0–0 6.Bg5 Na6 7.f4 c6, for example: 8.Nf3 Incidentally, 8.Qd2 Nc7 9.e5 (Instead of 9.Nf3 transposing to a main line) is an unique move order that I don’t see mentioned anywhere. Quite possibly 8...d5!? is the reason. 8...Nc7 9.Qd2 The current main line is 9.Bh4, when both 9...d5 and the dynamic 9...b5!? are apparently satisfactory. On the other hand 9.d5!? (stopping ...Ne6) 9...cxd5! 10.cxd5 isn’t much feared, but it's not easy to handle and in fact gets different answers from the two sources:

Gallagher gives 10...Nh5 “!” intending ...f6 and capturing on f4. He follows the eccentric but logical analysis by Seirawan that goes 11.f5 Nf6! 12.fxg6 hxg6 13.Qd2 Na6!, heading back to c5. This is a typically paradoxical modern line with Black taking 6 knight moves to get to a6 and f6! But I’m suspicious of the simple 11.Qd2, to which Gallagher replies 11...f6 12.Bh4 Bh6 13.g3 e5!? 14.dxe6 Bxe6 with double-edged play. I think the burden of proof is on Black after simply 15.Nd4 and I don’t trust this line. A similar game is quoted by Panczyk and Ilczuk without the insertion of 9...cxd5 10.cxd5. It went 9...Nh5 10.Qd2 f6 11.Bh4 Bh6 12.g3 e5 13.dxc6 (a possibility eliminated by Gallagher’s 9...cxd5!) 13...bxc6 14.c5 exf4 15.cxd6 Ne8, and now they suggest 16.Rd1 with a clear advantage.

The best solution to 9.d5 seems to be the immediate : 9...cxd5 10.cxd5 Na6. Panczyk and Ilczuk follow a game Sturua-Kempinski, Leon 2001: 11.Nd2 (a strange alternative would be 11.Bxf6 exf6 12.Qd4!? Re8 13.0–0) 11...Ne8 12.Rb1 (12.Rc1) 12...f5 13.0–0 Nc5 14.exf5 Bxf5 with a good game. White’s 12th move isn’t accurate, but the authors seem to have discovered the best way for Black to go.

Versus the older 9.Qd2 Gallagher’s repertoire follows the better-established solution 9...d5 10.Bxf6 exf6 11.exd5 Here P&I cite a game with 11.cxd5 cxd5 12.e5 Bg4 13.0–0 fxe5 14.dxe5 (I think that 14.fxe5 f6 15.Rad1 fxe5 16.Nxe5 Bxe2 17.Nxe2 Qd6 is fine for Black) 14...f6. Now they claim that   15.Rad1 “would have secured White an edge.” Does this mean that White gains a forced advantage after 9.Qd2 d5? That would be big news indeed, and it’s incumbent upon the authors to either improve upon this line or elevate it to a high status in the chapter. In fact, I don’t think that Black stands badly in the final position after 15...fxe5 16.fxe5 (16.Nxe5 Bxe2 17.Qxe2 Qd6 is very comfortable) 16...Bxf3 17.Bxf3 Bxe5, for example, 18.Nxd5 Qd6 19.g3 Rad8 20.Qe3! Rxf3! 21.Qxf3 (21.Rxf3 Nxd5) 21...Nxd5 22.Qf7+ Kh8 23.Rf2 Qc5 24.b4 Qe3 25.Rxd5 Qe1+ with perpetual check. 11...cxd5 12.c5 Bf5!? (“!”-Gallagher) 13.0–0 Be4 Black might also consider simply 13...Re8 14.Rad1 Qd7. 14.b4 Ne6 15.Rad1 15.Rac1 is played in Gallagher’s main game but is less logical. 15...f5 16.Ne5 f6 After citing 3 games to get to this position, P&I stop here, saying “with chances for both sides.” But since this is the main line of the variation, they really should take it further. 17.Nf3

17...Bh6 18.g3 g5 Here Gallagher says, “Black has good play.” He should also have followed up on the position. The important continuation is 19.fxg5! Nxg5 19...fxg5 20.Ne5 Kg7 21.Nxe4 isn’t working. 20.Kh1, and White seems quite a bit better, for example, 20...Qd7 21.Nxe4 fxe4 22.Nh4 Nf7 Or 22...Qh3 23.Nf5. 23.Qc3 Bg5 24.Nf5 Nh6 25.Nd6 f5 26.Qb3 etc. So once again the question arises: does this whole line favor White? For once it seems that both books have failed us. Fortunately we can bail them out by offering 17...g5! 18.fxg5 (18.g3 gxf4 19.gxf4 Qc7 20.Nh4 Bh6, or 19.Nh4 Qc7! 20.Nb5 Ng5! – threatening checkmate – 21.Nf3 Qf7) 18...fxg5 19.Kh1 Qe7 intending ...Rad8 and ...g4. This should favor Black.