Martin Breutigam was the author of the excellent ChessBase CD on the Chigorin Defense to the Queen's Gambit. With his new CD on the h3 variations of the King's Indian, he again enters into a subject that I have both written about and played quite a lot, so it's hard to resist reviewing his work. Breutigam covers 1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 d6, and now 5.h3 or 5.Nf3 0–0 6.h3, the latter accounting for by far the most material. The CD also contains training questions, an h3 database, and an opening tree.

Breutigam speaks from the White point of view, and indeed one of the CD’s strong points is that he wades through some incredibly complicated material and 6000 games to locate what he thinks are White's best lines. To me, he has done so with uncanny judgment, and there are only a few cases when I have meaningful differences as to what White should play. I do think that he misses Black's best lines more often, but I may be wrong and anyway no author could possibly get everything right in this morass. It should be mentioned that Breutigam also has an excellent section on the somewhat neglected 1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 d6 5.Nf3 Bg4 6.Be2 0-0 (normally arrived at by 5...0-0 6.Be2 Bg4). This is an attempt to avoid the h3 lines, and I think a particularly interesting line for the practical player in that case is 7.Be3 and 7...Nfd7 8.h4!?, which has enjoyed much success; or 7...Nc6 8.d5 Bxf3 9.gxf3 Nb8 (or 9...Ne5 10.f4 Ned7 11.h4) 10.h4 with attack. The author was under no obligation to include this section and it reflects his dedicated attitude throughout towards improving this CD.

Returning to h3 lines, Breutigam emphasizes the phenomenal success of the leading grandmaster Michal (new spelling!) Krasenkow when playing 5.Nf3 0-0 6.h3, but also points out that “Evgeni Bareev, Vassili Ivanchuk and Alexander Beliavsky...also [play h3 systems] more or less often. Even the World Champions Garry Kasparov, Vladimir Kramnik and Viswanathan Anand have already employed this system.” Breutigam gives credit to both Bent Larsen and Lubosh Kavalek for their earlier investigations into these lines. I think that he should also have mentioned the many contributions of GM Vladimir Bagirov, in my opinion the leading theoretician of such systems who spent decades playing and analyzing them. In fact, I named the 6.h3 variation “the Bagirov System” in my book.

At any rate, Breutigam continues: “Still, to date, as was mentioned above, hardly anything worth mentioning has been written on this line. This is confirmed by a look at three otherwise recommendable books about the King's Indian: Efim Geller devoted four pages to the variations with h3 in his old standard work (KONIGSINDISCHE VERTEIDIGUNG, Schachverlag Rudi Schmaus, Heidelberg 1980). Eduard Gufeld's GEWINNEN MIT KONIGSINDISCH (Sportverlag Berlin, 1990) devotes zero (!) pages to it; at least Joseph Gallagher's repertoire-book THE KING’S INDIAN (Everyman Chess, London 2002) has twelve pages on it.” [jw: The latter book is actually STARTING OUT WITH THE KING’S INDIAN (click HERE to see a review of that book) so it appropriately devotes more space to explanation and not too much to serious theory]. Unfortunately, Breutigam is unaware of my own 1997 book THE UNCONVENTIONAL KING’S INDIAN (“UKID”), in which I devote 44 pages to h3 systems, including 28 on the main line 5.Nf3 0-0 6.h3. I became interested in the latter line in 1980 when Dzindzihashvili slaughtered me with it (later I drew with him on the White side) and I immediately took it up. I have 37 of my tournament games on both sides of 5.Nf3 0-0 6.h3 in my database (with various others lost to posterity), so I know a good deal about these lines and can guarantee the reader that Breutigam has done a terrific job of covering the material. I would have avoided a few painful losses had I had his advice at the time.

I noted a couple of lines that interested me and may be worth mentioning. In his main text after 6...e5 7d5 Nh5 8.Nh2 (all other alternatives such as 8.g3 and 8.Nd2 are skipped in the general discussion, but are satisfactorily represented in the database of games), Breutigam doesn't mention my main recommended line in UKID, i.e., 8...a5!?, played as long ago as 1947 and brought to general attention by Uhlmann. Then the play generally goes 9.g3 (or 9.Be2, when Shabalov-J Watson, Las Vegas 1993 went 9...Nf4 10.Bf3 Na6 11.Be3 Nb4 -- 11...f5!? -- 12.0–0 Nbd3 13.Qc2 f5. This was equal and led to an early draw) 9...Na6 10.Be2 Nc5 11.Bxh5 gxh5, intending 12.Qxh5?! f5!. White should therefore play 12.g4! hxg4 (I think that 12...Qh4 is also satisfactory) 13.hxg4 Qh4, and now a key game went 14.Be3 h5 15.Kd2!? (a novelty given “!” by Bagirov since earlier tries had failed) 15...hxg4 16.Nf3 Bagirov-Shaked, Linares 1997. Here instead of 16...Qd8?!, Black had 16...Qf6! with the idea ...Bh6, leading in my opinion to a small advantage for Black. One might want to keep the underrated 8.Nd2 in mind here.

The book has particularly good coverage of systems in which White plays Bg5, provoking ...h6, and then retreats to e3. In my opinion this is the best way for White to attempt to gain an advantage after three of the main lines following 1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.Nf3 0–0 5.e4 d6 6.h3: 6...e5 7.d5 Na6 8.Bg5 or here 7...a5 8.Bg5, and 6...Na6 7.Bg5. In the latter variation I was curious to see that 7...Qe8 8.g4 c5 has been answered by 9.Bg2!? in Krasenkow's games, with a strange sort of Maroczy Bind structure resulting from 9...cxd4 10.Nxd4. I'm not sure that I trust this to yield an advantage, but the older option of 9.d5 is still hard for Black to meet, as far as I know. One of my own games went 9...e6 10.Qe2!? Nc7?! (10...exd5 11.Bxf6 Bxf6 12.Nd5 is promising) 11.e5! dxe5 12.d6 e4 13.Bxf6 exf3 14.Qxf3 Bxf6 15.Qxf6 Qc6 16.dxc7!? Qxh1 17.Rd1 with a strong attack, although I failed to prosecute it accurately and Black achieved a draw in Watson-Gulko, Philadelphia 1993.

The coverage of 6...e5 7.d5 Na6 8.Bg5 and 7...a5 8.Bg5 is superb, discussing nearly every possible plan for Black. A key line is 1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 d6 5.h3 0–0 6.Bg5 Nbd7 7.Nf3 e5 8.d5 h6 9.Be3 Nc5 10.Nd2 a5 11.a3 Ne8 12.b4 axb4 13.axb4 Rxa1 14.Qxa1 Na6 15.Qa3 with a small but definite advantage, Beliavsky-Khalifman, Ubeda 1997, eventually won by White. In my book, I didn't mention 11...Ne8, but instead gave what I think are superior alternatives: 11...Nfd7 (since 12.b4 f5! looks equal), and 11...Bd7, when 12.b4 axb4!? (best is probably 12...Na6! with the idea 13.Qb1 c6) 13.axb4 Na6 14.Qb3 Nh5 yields unclear play.

I should add that Breutigam provides many suggestions and new ideas throughout. He also makes it fairly easy to assess which lines are the most desirable. My general impression from this CD (and from experience) is that White tends to be better in every main line in which (after ...e5) Black plays the moves ...Nc5 and ...a5, and then moves his f6 knight and plays for ...f5. But I personally had the most trouble versus flexible systems with an early ...c6 and ...Bd7, usually without an early ...Nc5, when Black retains options on both sides of the board (just as White normally does). This ...c6/..Bd7 plan is also recommended by Gallagher in several key lines. Anyone examining Breutigam's work will see that there is still a tremendous amount of leeway for creativity with all the h3 lines. I would therefore recommend not only this excellent CD but also the systems described on it. They are ideal for those looking for an anti-King's Indian System that is not even close to becoming worn out.