John Watson Book Review (14)
Comments on the Great Dvoretsky Project
IM John Watson - Monday 13th September 1999
Attack and Defence; Mark Dvoretsky, Artur Yusupov, and contributors; 288 pages; B.T. Batsford, 1998
Attack and Defence; Mark Dvoretsky, Artur Yusupov, and contributors; 288 pages; B.T. Batsford, 1998
[Preface: Numerous readers have asked me about the Mark Dvoretsky's books, so I thought I'd respond by taking the lazy man's way out: re-use of a book review I wrote late last summer! What follows is a revised version of an article I wrote for the Illinois Chess Bulletin ('ICB'). Under the editorship of the multi-talented M.L. Rantala, that magazine has been one of the finest 'regional' chess publications I've ever seen. I am most grateful to her for permission to reprint the review. I have modified the review somewhat, mainly by eliminating certain portions which referred to the ICB itself or to specifically American chess issues. As for the decision to dredge up old work, it seemed silly to start over again on this subject, and I was also motivated by a heavy travel schedule in the coming weeks. How Mig manages to produce so many consistently entertaining and informative columns in such a short time is beyond me; all I know is that readers of this column will be have to get used to my slower pace, and suffer the occasional long gap between reviews.
Looking over what I wrote then, the only point which I want to emphasize in advance is that I think every Dvoretsky book which has been translated into English is a high-quality effort and a real contribution to chess literature. I am critical of certain aspects of the book under review here (Attack and Defence); but that criticism has to be taken in the context of the series as a whole. When I contend that this particular book is the worst of the series, but that it is still a very good book, and well worth reading, that should indicate what I think about Dvoretsky's output as a whole. I drew upon Dvoretsky's work regularly in my own recent book on modern chess, and in my research, I found him to be one of the few writers with original and meaningful insights about general middlegame issues.]
Perhaps the most significant publishing events in the chess world during this decade has been Batsford's publication of a series of books by Russian trainer IM Mark Dvoretsky about various and sundry aspects of chess training and theory, including books on tactics, technique, and positional play. By my count, Dvoretsky now has 7 books in English, 5 of them in this Batsford series officially co-authored with GM Artur Yusupov (a Dvoretsky student), and two of them by Dvoretsky alone. In fact, all seven books are primarily Dvoretsky's work, since he has written about half of each book, created all the exercises, and collected contributions from not only Yusupov, but a wide variety of other masters.
Because I find Attack and Defence to be atypical of this series, I would like to talk about Dvoretsky's contribution as a whole before turning to that particular volume. It's interesting that Dvoretsky's books have made such a splash among masters and top players, including some who seldom read chess books. To a large extent, their enthusiasm about Dvoretsky is due to the fact that his books have that rare quality of addressing the problems facing masters and even grandmasters, rather than talking down to the general chessplaying public. Dvoretsky's examples tend to be complex and hard to pigeonhole, and his constant preoccupation is the improvement of already-master-strength players. This has led to the amusing situation that masters tend to recommend his books to their students, who in turn complain that they are too heavygoing! Apparently, Dvoretsky's books are also difficult to review: most Internet sites with chess book reviews have somehow avoided commenting upon even a single one of them. The fact is, these are difficult books.
Let me first outline what I view to be the strengths and weaknesses of this series. There are many positive features, but I think a few stand out. First, Dvoretsky's books are truly aimed at training, and do not simply list, for example, a set of positional principles or advice on how to calculate. Each volume is replete with well-chosen, concrete exercises for the student (Dvoretsky has been collecting these positions for years), along with instructions on how to best learn from them. As trainer to many of the world's GMs, Dvoretsky knows that only by hard work and practice with real positions, using a clock, can one improve one's thinking habits, regardless of how many handy tips or how much philosophy about chess thinking you've been fed. Of course, this positive feature of his books also limits the number of players who will enjoy them or find them of realistic value. No everyone has time to devote to this sort of study, and many players would rather read chess books for enjoyment (which, I hasten to add, is a wholly legitimate attitude). For those players, however, I'd recommend throwing out their how-to-get-better books and starting with the annotated games collections of top players.
Another outstanding aspect of Dvoretsky's project, more prominent in some books than others, is that he is one of the few chess writers to actually contribute to general chess theory, and to middlegame theory in particular. Dvoretsky and some of his contributors understand that modern chess varies considerably from that played by the classical masters, and he uses his own conceptual frameworks to explain how modern ideas and techniques apply. In a few cases, Dvoretsky has himself defined the theoretical territory involved. He has talked more about prophylaxis in chess, for instance, that all of the other writing I've seen on this subject combined. He is a renowned authority on opposite-colored-bishop positions, from the opening through the endgame. His books have expanded the discussion of good and bad minor pieces, the exchange sacrifice, active roles for the king in the middlegame, and many other original and specialized topics. Dvoretsky's notion of the 'superfluous piece' is both unique and highly relevant to modern chess; it is now starting to be referred to by leading GMs in their annotations. And Dvoretsky's contributors provide their own new ideas as well.
I'd like to single out one other excellent feature of Dvoretsky's books. That is his consistent and sometimes heretical insistence on the primacy of concrete analysis over positional principles and other abstract considerations. This is the fundamental philosophic stance which Soviet masters and writers have adhered to from the time of Botvinnik on, and the one which most distinguishes modern chess from that of the classical masters. Dvoretsky (with the explicit approval of Kasparov) both re-emphasizes and thoroughly explicates the details of this approach. His timing could not be better: the top ten players of the world, for example, are all practitioners of the concrete, analytical approach, and are less inhibited by traditional principles than any previous set of leading players. I feel that this accounts, in part, for the greater dynamism and creativity we are seeing in today's top-flight chess.
So are Dvoretsky's books perfectÿ Not quite. To me, the main problem with this series is his tendency to merely fill space, without tying the material strongly enough to the topic under discussion. It seems, for example, that every time Dvoretsky and/or his students have thoroughly analyzed a position, whether it be an adjourned game, a post-mortem analysis, or someone else's book or magazine article, Dvoretsky wants to include the analysis in one of his books. That would be perfectly appropriate if the game was an instructive example of the subject being addressed; but I find that both Dvoretsky and his students (but not Yusupov) invent general 'lessons' or 'themes' to justify including these games. Worse, they sometimes, in the justification process, start dishing out the same tired and simplistic generalities which Dvoretsky is so careful to refute (or at least warn the reader against) elsewhere in his presentation. To some extent, this is a problem of having multiple authors; but Dvoretsky himself is responsible for much of this 'filler' material. The quality of such filler varies--sometimes, the game involved is fascinating--but he gropes awkwardly to justify its inclusion. Readers familiar with Secrets of Chess Tactics may remember how he manages to work some static endgames which interested him under the rubric of 'Attack and Defense'; in this latest volume by that same name, he performs similar legerdemain. Then there are instances such as his virulent trashing of an article he doesn't like, or his presentation of some favorite piece of lengthy analyses without tying it to the lesson at hand. The point is that, some of these examples, however well-presented, only dilute the important theoretical and philosophical messages his books bring us.
Okay, let's finally move on to his latest volume: Attack and Defence. Since I am about to wax critical, keep in mind that I think Dvoretsky's books have made an unprecedented contribution to both chess training and modern theory. Attack and Defence is, in fact, another very fine effort, with a multitude of rich examples and analysis. But in my opinion, it contributes only mildly to the subject of how to analyze, and very little at all to the theory of attack and defense (the purported topic). Maybe Dvoretsky has simply written too much, and the well is temporarily dry; he says that this is the last of his five-part Yusupov collaboration. This time, there are seven contributors, and its hard to discern any unity of material or purpose in their essays. The 'filler' problem, moreover, seems to have grown considerably, perhaps in an attempt by Dvoretsky to use up as much of his admittedly interesting material as possible. Let's step through the book to see what I mean.
The Introduction begins on a dubious note, with some bragging about the great success of his students (we don't find out how good they were in the first place, or how they were chosen; not randomly, one suspects!). Then there's a rather tired lecture (one not given by Dvoretsky before, as far as I know) about how chess players today study too much opening theory, rather than picking up 'astute observations and conclusions about the game as a whole'. Aside from the fact that the specific observations he points to at this juncture contradict those in the rest of his books (see my comments on Blumenfeld below), the two students he mentions with by far the greatest success are Svidler and Zviagintsev, theory hounds who have obviously devoted untold hours to opening study! Dvoretsky's proud example of an early Zvjaginsev game, for example, has the 13-year-old playing a new and thoroughly-prepared move on the 15th move of an obscure variation of the Slav Defense. Considering his current rapid ascendancy, it's not clear that Vadim would have done better by poring over some psychological essay instead.
Chapters 1 and 2 are discussions of techniques of analysis and in particular, of Kotov's methods. Both Dvoretsky and Krasenkov criticize Kotov in rather important areas, and they offer their own versions of how to analyze. These essays are convincing and well thought out; but mostly, they only outline what common sense tells you about how a strong player actually thinks. Dvoretsky gives the reader more 'tips' about what to remind oneself of during analysis, whereas Krasenkov provides a clearer discussion of when and how to use one's intuition (based on the time left on your clock). This is a well-written section of the book, and certainly worth including, but it's also a bit obvious.
Chapters 3 and 6 are two essays from the 1930s by the psychologist Benjamin Blumenfeld. Neither has much to offer, in my opinion, despite the constant boldfacing Blumenfeld puts on his rather mundane comments. The first essay begins with an innocuous explanation of why people blunder, suggesting that they should concentrate harder, stay alert, and think more systematically to solve the problem. Fascinating. Benjamin's next, more controversial, argument is that people tend to confuse themselves by calculating too much, so they should limit calculations and 'opt for simple solutions'. In non-forcing situations, he says, "calculations would be confined to a few short lines"; and in general, "calculation is only necessary to verify that the ideas are correct." My first question would be: what ideasÿ I can just see Kasparov, Shirov, or Polgar looking at one or two moves, collecting their 'ideas', and making a move. Of course, ideas, especially creative ones, are the fruit of analysis, not something inherent in one's 'picture' of a position. The irony here is that most of the rest of the book, not to mention Dvoretsky's whole series, is concerned with precisely the opposite point: that the general features of the position are relatively worthless, and often deceptive, without the most disciplined examination of concrete lines of play. That examination, says Dvoretsky, leads to the creative and unexpected solutions which allow one to defeat strong players. In later chapters of this same book, for example, Dvoretsky strongly approves of even unsound sacrifices when they increase ones practical chances, whereas Benjamin says to always opt for clarity. Overall, Benjamin's advice seems to me an excellent prescription for 'How to stay a low Expert for the rest of your life.' I don't think I need to point out who is on the winning side of this argument in contemporary chess!
In fact, in the very next chapter, on the sources of chess intuition, Dvoretsky re-emphasizes a primary theme of his series: "...intensive analysis of concrete situations--this significantly develops and enriches our intuition." He then goes on to advance the brave but dubious theory that the intuitive side of Kasparov's game is flawed! Since there is only one example given, I'm not sure that the reader will be convinced (maybe Kasparov's intuition was just flawed on that particular dayÿ). This theory is interesting to juxtapose with the somewhat more plausible one of Chapter 14: that Kasparov tends to be relatively weak at defense. The reason for this, Dvoretsky opines, is that Kasparov's openings are so good, he is seldom in trouble and therefore rarely gets practice defending! There might be a some truth to this, although I would still only call it a working hypothesis. Of the four difficult positions Dvoretsky cites, for example, all against Bareev and Anand, Kasparov won one, lost one, and drew two; given that the goal of the game is to garner points, it seems that there may be method behind the madness of such 'weak' defense!
Let me conclude by briefly reviewing other aspects of the book. There's a whole chapter of quite appealing and well-annotated wins by Sergei Dolmatov. Again, what relevant general lessons they contain is not clear to me, but the games and notes are great. Amusingly, Dolmatov at one point gives himself an '!' for correctly following the principle of avoiding pawn exchanges when he is material ahead; but then, one move later, he attributes his mistake to following that same principle (i.e., he should have exchanged pawns, but didn't)! Again, Dvoretsky's books make a good case for being wary of such principles in general.
Other chapters are mostly annotated games organized by theme, e.g., 'Missed Brilliancy Prizes', 'Virtuoso Defence', 'Opposite-wing Castling', 'Games by Pupils', and 'Analysis of a Game'. I can't deny that these games tend to be fun and instructive, in the general sense that games annotated by strong players tend to be instructive. I also think that the annotations make a clear and powerful case for the virtues of the analytical method which Dvoretsky advocates. What's not clear to me is whether these games say much about the nature of attack and defense, or about how one goes about attacking and defending. Again, it's the lack of focus and originality that makes this volume, for me, somewhat unsatisfying.
In the end, I feel that Attack and Defence is a rather disconnected collection of games and positions, and perhaps the worst book of this series. On the other hand, since this may well be the best series of instructional/middlegame books ever written, last place isn't so bad! Certainly the annotations and comments by strong players put it on a level with a very good games collection, and serious students will benefit from the high-quality exercises. All in all, this book is well worth reading; but for your first Dvoretsky experience, I would recommend another volume (e.g., Positional Play or Training for the Tournament Player) instead.