I have looked forward to Iossif Dorfman's THE METHOD IN CHESS for some time. Dorfman is considered a leading trainer, most famously for his stewardship of Etienne Bacrot. I heard that many in Europe consider him superior in that field to super-trainer Mark Dvoretsky; given that high praise, reading about his chess philosophy and insights was an intriguing prospect. In particular, this work got favorable marks from many reviewers, and is even listed among the 5 finalists for the Chess Café book of the year award.

What a surprise I was in for: this book is a hodgepodge of assertions that are either too broadly interpretable or simply banal. I see little actual content to the purported method, mostly because Dorfman doesn't establish enough limits or uses for the loose generalities it seems to consist of. Worse still, the great bulk of the book is merely a collection of Dorfman's games, and amazingly, the notes to most of the games consist primarily of variations with no reference to Dorfman's theories whatsoever!

In the foreword, Dorfman announces that he has created a “new theory”, adding that “the reader will see a number of rules being formulated for the first time, generalizing the processes taking place during play. Rules enabling certain well-known postulates to be explained, and others to be looked at more critically.” A fair enough “resume” (as he calls it), although he concludes with the obscure (and boldfaced) quote: “I regard chess as being an equivalent (adequate) exchange” (Botvinnik). Hmmm. Anyway, in the Theoretical Section Dorfman defines “dynamism” as variations [changes] in “the hierarchy of strategic factors”, so that “plans and ideas are transformed”. “To foresee the modification of the hierarchy of strategic factors is nothing other than being able to define critical positions.” Crucially, he proposes, “analyzing critical positions on the basis of their static state. This aim is served by the proposed static balance.” This shifting status of static [enduring] and dynamic factors is the main theme. The player needs to find a critical position (a turning point in the play). Then he or she makes an assessment of the static balance. “If for one of the players the static balance is negative, he must without hesitation employ dynamic means, and be ready to go in for extreme measures.”

This is all quite reasonable-sounding, and clear enough if one concentrates; but as one might expect, such an abstract and indefinite approach leaves an awful lot up to the player's judgment. It also handily preserves plenty of room for post-facto theoretical justification. One might assess the static balance as negative, go dynamic, and later find that it was much better not to do so (if only to pursue the lesser evil). Is that supposed to be impossible? Dorfman bravely tries to help us by defining “critical positions” as those in which a decision must be made about (a) possible exchanges or (b) possible changes in pawn formations; he also includes (c) “the end of a series of forced moves”. As he says, “To sense that a position is critical is already a great success”, which is not encouraging to someone trying to think in this manner. I don't think he ever adequately addresses how to do this, especially because the topic almost never comes up in his own exemplary games! This idea -- that developing a sense for critical positions is essential -- has been taught by Dvoretsky and presumably numerous others for many years, and I'm not sure that Dorfman has added much with his above list, since the delicacy of the required recognition is so obviously more subtle than the list indicates.

Dorfman also provides a method for “drawing up the static balance”. He “suggests” (I think that “proposes” might have been a better translation at several points) “a regressive scale of static factors, arranged in order of their importance.” This word “regressive” is used repeatedly, but as someone who experienced the word in a scientific and problem-solving context over many years, I can't see how any facet of Dorfman's static evaluation, much less his method as a whole, is “regressive”. At any rate, he then turns to more traditional elements of positions such as king position, material, pawn formation, good and bad pieces, weak squares, etc. The breakdown of these elements into categories is, as far as I can see, absolutely standard and thus can be found in any number of instructional books.

So the theoretical edifice is a little shaky, but let's poke around and see how he fleshes it out. Right away I have problems with dogmatic statements such as “Rule: the exchange of a bishop for a knight can only be justified when the pawn formation is fixed” (this is repeated at least twice, and much more strongly than as a mere guideline). Finding exceptions to this rule is of course easy – imagine if Bacrot, or Dorfman himself, took it seriously. In fact, this rule doesn't even apply to Dorfman's own examples that he uses to illustrate it, unless one misses easy improvements for the side yielding the bishop. Then we have oddball characterizations such as that of an “outpost”, which turns out to be “a square on a half-open file in front of an enemy pawn, situated on the 6th (for White on the 3rd) rank.” That is bizarre enough, but the steps for exploiting its advantages are also strange: (a) create the outpost; (b) place a piece on it (“usually a knight”); (c) “in the event of the exchange of this piece, recapture with a pawn, exposing an enemy backward pawn”; (d) “create pressure on the backward pawn”; (e) “force its advance and attack the resulting weaknesses”. When you think about it, this only applies to specialized cases, and extremely often such a procedure would give the owner of the “outpost” (so defined) a distinctly inferior game. On top of that, step “c” seems to assume that, say, Black's adjacent pawns are on the second rank, e.g., that in the case of a knight on d5, they are not on c5 and e5 so that White would suffer from a “dead spot” on d5. Even when a pawn is on the second rank, say on e7 or c7, a pawn recapture on d5 may well allow Black the opportunity an advantageous for ...f5 or ...b5.

I find myself quite often disagreeing with the variations that Dorfman supplies; he tends to skip the opponent's favorable opportunities, sometimes obvious ones. Then there are the irritating editorial mistakes such as misnumbering of moves and move typos. It's not that difficult to do move-checking these days.

Revealingly, the section on the “Practical Application” of the method consists solely of 64 games by Dorfman. One might be more convinced if it turned out that other top players' games reflected use of the method. Or if his games were linked to his text and theories, rather than limited for the most part to comments about the players, opening theory, and wordless variations. Then, as if those contests weren't enough, he has an “Appendix” of “several games”; this turns out to be 29 more almost noteless Dorfman games. And in all these 93 games, amazingly, I can't find a single negative comment about Dorfman's play! Seriously. Perhaps a few exist somewhere, but the point should be obvious.

I have no doubt that Dorfman is a good teacher/trainer, but people with a particular skill do not always write good books. Unless you're looking for “The Lightly Annotated Games of Dorfman With His Mostly Inapplicable Pedagogical Observations”, you should stay away from this book. Perhaps his recently released follow-up will be better, but for that to be true he would have to have brought himself down to earth, and to have widened his base of examples.

.