TWIC is 30. First issue 17th September 1994.
John Watson Book Review (64)
Once More into the Breach
IM John Watson - Sunday 30th January 2005
Garry Kasparov on My Great Predecessors Part II; Garry Kasparov ('with the participation of Dmitry Plisetsky'); 478 pages; Everyman 2003
This review was written some time ago and by now the book in question has been extensively commented upon by many fine reviewers. Nevertheless, I feel that 'Garry Kasparov on My Great Predecessors Part II' can do with some further examination. In this volume, Kasparov investigates the careers of World Champions Max Euwe (1935-1937), Mikhail Botvinnik (1948-1957, 1958-1960 and 1961-1963), Vasily Smyslov (1957-1958), and Mikhail Tal (1960-1961). Overall, the book strikes me in much the same way as Part 1 (with some important exceptions, as shown below). Again, Dmitry Plisetsky, who doesn't even appear on the cover or in the Introduction, seems to have done most of the work. For example, he has apparently written the many historical accounts that define the book and he almost certainly collected the games that were used, probably together with the existing notes to them. As the chief researcher, one can also assume that he was responsible for one of the most intriguing parts of this work: the abundance of quotations from great players. These are spread throughout each chapter, giving the World Champions' opinions and observations about each other's styles and play, as well as comments about other players under discussion, notably Keres, Bronstein, and Geller. We even find some Champions' critiques of themselves. Their commentary and that by other analysts often provide the majority of the verbal notes to games, with Kasparov providing the rest. Every section ends with a fascinating string of reflections by great players. All of this entertains us and greatly enriches the book. The analysis of moves themselves also contains much input by the participants and other analysts; here of course Kasparov himself has the final say and provides the bulk of the important and original work. As in the previous volume, any pretence to scholarship goes out the window when no specific source is given for the great majority of quotations or the books from which analysis is drawn. I hope that this fault can be righted in volumes to come.
Kasparov's introductions and notes to the games include some of his own thoughts on various topics including his assessments of why a certain player prevailed or lost, with some very interesting but too brief discussion of the evolution of chess. He again provides us with sweeping theories, sometimes rather absurd. For example, he has a discussion of the champions' contributions to opening theory, by which contributions he strangely seems to assess and define each great era of chess thought. In the midst of describing the progress of chess he says, 'Smyslov, Tal, Petrosian, and Spassky did not achieve any radical progress in the opening and developed theory along the lines of Botvinnik'. He describes Fischer as creating 'the next evolutionary leap'; Fischer 'sharply expand[ed] the range of openings studied'. To me, this is a very odd characterization of the player who had the narrowest repertoire of these five. Then he says, 'Note what an interesting line can be traced: Steinitz (1st champion) – Botvinnik (6th) – Fischer (11th), that is, a revolution accomplished by every 5th champion!' Needless to say, a trend based upon a sample space of two intervals is hardly compelling. Kasparov, by the way, was the 13th champion; we'll see what this means when we get to the final volume.
Apart from the games themselves, the book sticks mainly to the recitation of events. That in itself can be fascinating and inspiring for players who have never absorbed much of the history of chess. Among other discoveries, for example, I certainly learned about or was reminded of fascinating material regarding the Interzonals and Candidates tournaments matches. Still, we welcome the great exception to such mere recitation when Kasparov discusses Botvinnik. At the end of the chapter devoted to the latter there are no less than 7 pages of discussion by Kasparov about his great teacher. Based upon personal experience, he provides numerous insights, both positive and critical, into the various sides of Botvinnik's personality. One surprising point (claim?) that Kasparov makes is, 'I am the only genuine pupil of Botvinnik. On the whole he met the others only at training sessions, whereas he had regular and close contact with me for a full 14 years.' I should mention that in this excellent section Kasparov is assisted and uses material by Sosonko, without attributing much of it (which is all too typical of this book). The personal touch with regard to Botvinnik bodes well for future volumes, since we have Petrosian, Spassky, and Karpov ahead of us. And the section on Fischer should be extremely interesting regardless.
As in Part 1, the descriptions of players' styles are very similar to the stereotypes invariably presented in the rest of chess literature, often using the very same phrases and descriptors. Maybe that's all there is to be said, and obviously those stereotypes exist for a very good reason, i.e., they must reflect the most characteristic aspects of a player's overall style and personality. But I find it inconceivable that a player of Kasparov's chess intelligence couldn't with some effort have discovered anything subtle or interesting in the shades of stylistic differences, or strengths and weaknesses of these players that are not often noted. One thinks of Robert Hübner, who takes his conclusions about champions (however overcritical I find them) from a careful study of the players' own games and commentary; his results repeatedly challenge conventional views and reveal subtle tendencies that the players exhibit.
Before moving on to the redeeming quality of the book – its analysis of the classics – I want to register a very strong criticism about the choice of material. Yes, one can always complain about what should or shouldn't have been included, and normally that would be purely a matter of taste. Mig Greengard whines about people who engage in this exercise and say that they should write their own book (so much for the world's leading review publications!). Nevertheless, as my friend John Tomas pointed out, the book is 'essentially only a history of USSR chess (with the exception of Euwe)', noting out that Fine disappears from the 1930s (he's not even in the Index!), and Reshevsky is almost completely ignored. These are two of the most egregious examples of bias against Western players. To demonstrate Kasparov's prejudices, I'll examine those two in detail.
For starters, consider the 37 pages on Geller. He was of course an important and great player, but to get an idea of his strength in world chess, let's look at Jeff Sonas' figures on his historical ratings and rankings. Sonas has calculated yearly estimated ratings for every player back into the 19th century, helped of course by modern FIDE ratings. For 1950-59, he uses two ratings per year, and from 1960 onward he splits the year into 4 parts. This leads to a list of world rankings for each year. I will count players as having achieved their top ranking whether it was for the whole year or fraction of one. These figures yield the following information about Geller. He was ranked 3rd in the world only once (in 1963), 4th two times, and 5th once. Geller was in the top 10 for at least part of 16 years, but he was typically 8-10th in his prime, for example, he was ranked 10th in one part of the year seven times, and similarly ranked 9th seven times and 8th eight times. He had a historical rating of 2700 or above for only 1¾ years.
Now referring to those same rankings, we find that Reshevsky was in the top 10 for no less than 24 years. He ended the year in ranked 1st in the world twice, was ranked 2nd three times, was 3rd-highest ranked in the world in 11 different years, and held 4th place five times! He spent 13 years at 2700 or above. Nor can his record be denigrated as a product of less important tournaments. In the second half of the 1930s alone (the period covered most by the two Predecessors volumes put together), Reshevsky took clear first in Margate 1935 ahead of Capablanca, won strong U.S. Championships in 1936 and 1938 (and of course many of them later), finished equal 3rd in the famous Nottingham 1936 tournament, shared first place at Kemeri 1937, and won Hastings 1937-8. This represented his golden era, but Reshevsky remained one of the world's elite throughout the 1940s and much of the 1950s. For example, he tied 2-4 in the famous 1953 Zurich Interzonal World Championship qualifier and in the U.S.-U.S.S.R. match of 1955 Reshevsky (on board one) won his 4-game mini-match versus Botvinnik 2.5-1.5. The rest of the U.S. team scored a miserable 4.5-23.5! Reshevsky closed out his career at a level that reminds one of Smyslov or Korchnoi. For example, he qualified for the 1964 Amsterdam Interzonal at age 52 and at age 55 won 6-8th place in the 1967 Sousse Interzonal, thus earning a spot in the Candidates. He qualified for and played in two more Interzonals, including Portoroz 1973 at the age of 61. Reshevsky was still a strong player going into the 1980s, and just months before reaching age 70 he achieved a 3-way tie for first in the 1981 US Championship, a Zonal Tournament. He then failed to qualify in the playoff that followed due to inferior tiebreaks!
This is surely the best record of any player who never played in a title match. Only Keres comes close. Reshevsky was certainly as talented as any great player apart from a few World Champions. And while he was not an expert in ultra-theoretical openings (nor was Capablanca or Petrosian), he found an amazing number of new moves and treatments in more positional openings, most obviously Queen's Gambits of nearly every stripe, Nimzo Indians, Gruenfelds, Sicilians, and Double King Pawn openings. A look at the index of his games collection (compiled by Steve Gordon; McFarland 1997) reveals that he played nearly every major opening, including risky ones – I estimate that well over 80% of the 500 ECO codes are represented. Reshevsky also greatly enriched the game by his brilliant methods of gaining and extending middlegame advantages (much as Karpov did) and producing amazing defensive ideas in difficult positions. Yet Kasparov not only denies Reshevsky a section in his book, he essentially ignores him. Reshevsky is 'represented' in Predecessors 2 by 4 losses to Soviets. Outside of those games (in the notes to which he isn't given much attention), I can only find references to Reshevsky consisting of 7 partial-sentence mentions and a paragraph about his blitz match versus Smyslov at age 80 – a total of perhaps ¼ of a page! Compare this with Geller's 37 page section of the book (and more games outside of that section); it's hard not to infer a serious bias towards the Soviets on the part of Kasparov.
[See the end of this review for some feedback from Kasparov about his later change of the material in Predecessors and other interesting commentary regarding this issue.]
One might propose that this just a case of extreme favoritism towards Geller. Clearly not, because other Soviet players get similar treatment. Bronstein is accorded his own section of 29 pages, quite apart from the 24-page exposition of his match with Botvinnik! Bronstein was a great player of course, but he was never ranked first in the world, and he got as high as 3rd only three times (during a period that chess at the top was relatively weak, according to Kasparov), 4th place five times, and 5th three times. He was in the top 10 for only eleven years. Yes, Bronstein played a World Championship match (Reshevsky had a lesser opportunity in the 1948 Championship tournament). But Bronstein himself says in his book that his good friend Boleslavsky agreed to take draws in the last two rounds of the Candidates Tournament to allow Bronstein a chance to catch up to him. As indicated above, his subsequent match with Botvinnik is itself more than adequately covered by Kasparov.
The extremely lopsided treatment accorded these players cannot be just a matter of taste; it reflects a prejudice towards Kasparov's predecessors in the West. Another example is Keres, who was of comparable strength to Reshevsky yet receives his own section of 26 pages (with more game material elsewhere). Admittedly Keres was a much more successful player than Geller or Bronstein, but his detailed treatment invites comparison with Reshevsky and others. Reuben Fine was another famous player from the United States. How much attention does Fine get from Kasparov? Quite simply, none at all; there isn't a single game by him in Predecessors 2! How can this be? Kasparov's bias in favour of the Soviets is again evident. One could try to argue that Fine's career wasn't long enough, but his total exclusion is nevertheless incredible. After all, Fine was either the #1 or #2 player in the world for 4 of the 5 years between 1936 and 1941. Among other tournaments, he won Hastings 1935/36, tied for first in Amsterdam, and came in 3rd at the famous 1936 Nottingham tournament. He won Zandvoort 1936 (a point ahead of Euwe and two ahead of Keres), Leningrad and Ostende in 1937, Margate 1937, and of course finished 1st with Keres at the famous AVRO 1938 tournament (ahead of Botvinnik, Euwe, Alekhine and Capablanca). At the 1937 Stockholm Olympiad he played behind Reshevsky but got the best score on 2nd board and led the U.S. team to victory. On a theoretical level, he contributed opening ideas that broke new ground and he wrote one of, if not the, greatest endgame books to appear up to that time. Would a Soviet player who was ranked #1 in the world, sporting such a record and reputation, escape Kasparov's notice? Of course not, and you will hardly be surprised by now that the super-grandmaster Najdorf is granted only 2 losses and a sentence or two. And there are further snubs, e.g., I haven't even mentioned Eastern European players.
What I've said so far is mostly critical of Predecessors, but others have been more withering in their recitation of specific inaccuracies, poor analysis, and biases. Plisetsky and Kasparov apparently continue to attribute analytical contributions to Soviet chess analysts although they were prominently featured elsewhere. The book's advocates generally use the defence that all that is unimportant and beside the point. Matthew Sadler says, and Jan Timman tends to agree (his words), that one can discount historical and analytical errors by comparison with an error-free work by a lesser grandmaster "because it's Garry!" I don't think that's even useful to say until we see how a quality chess writer would do given more time and care than Kasparov has expended. But that's a matter of speculation. They are certainly right that a Kasparov can give us insights that few other players in the world can. That is extremely valuable for our understanding of chess, and in itself makes this an important work. However, the continual use of ultra-famous games (some of the match games are an exception) wastes another opportunity for originality. Mig Greengard, for example, fumes about the critics, arguing that the recitation of standard history and too-familiar games serves the majority of the chess community who are ignorant of such material. To prove his point, he travels to a Barnes and Noble bookstore (this is a nationwide chain store in the United States) and finds only a few books that deal with chess history. Fair enough. But this raises the question of whether Predecessors 2 is a classic of chess literature (as has been claimed more than once), or a book to fill the needs of Barnes and Noble customers, quite a different matter. In my review of Predecessors 1, I tried to draw the distinction between a book that absorbs and entertains us, a valuable thing in and of itself, and one that can be considered a great book, i.e., that really breaks new ground and/or makes a significant and truthful contribution to the body of chess knowledge.
By this criterion, I think that the answer is muddled. In a historical not to mention scholarly sense, Predecessors 1-2 introduces error and confusion to the literature without adding much to what is already well known. That is not to denigrate its popularising effect, which is laudable, but the book offers little new material and is in some respects a step backwards. On the other hand, in the sense of advancing our knowledge and understanding of classic games, I think that errors in attribution and even analysis are probably outweighed by Kasparov's extensive and original annotations, in particular the many new moves and raw analysis that he contributes. I'm unfortunately not familiar with the particulars of this volume, e.g., whether a majority of analysis is still found in other sources or how many analytical errors exist. There have been critics in this regard. One reviewer says that for several games Kasparov's analysis is inferior to that in other widely available sources (oh, for a Bibliography!). In addition, Robert Hübner’s articles in Schach are said to contain many accurate criticisms of the analysis; alas, I haven't read them, but the two I have seen look convincing. Finally, some clear errors (even a couple of simple wins) call into question Kasparov's avowed application of computer engines throughout. For all that, I have to say that it counts for an awful lot when the most famous games of history are subjected to thorough and sometimes ground-breaking analysis by the greatest player in history. These are, after all, many of the games that we grew up with and treasure most. I was especially impressed by improvements at the most famous and critical junctures of these games. Kasparov's scrutiny produces both extraordinary combinations and subtle details. Of course only a tiny percentage of regular players (or for that matter professional ones) will actually play over his labyrinthine notes. The majority will simply read the narrative and follow the drama. But this doesn't lessen Kasparov's objective contribution in the slightest; the literature will be forever enriched by his analysis. Moreover, the originality of his investigation must be a key factor in assessing the worth of this work, and no one seems to deny that quality in it.
My own impression from looking at a few games is that the notes, whether accurate or not, could well repay a year or two's study. I did, however, notice the elementary fallacy of annotating by result. To be clear: Kasparov doesn't analyse by result. His notes, which are the important part, are objective. But his criticism of moves, admittedly a lesser matter, seems inconsistent and influenced by the outcome. This was evident in three of the four games I looked at for this review.
Here's an example from the renowned game Tal-Portisch, Bled 1965, the 2nd game of their Candidates match:
1.e4 c6 2.Nc3 d5 3.Nf3 dxe4 4.Nxe4 Bg4 5.h3 Bxf3 6.Qxf3 Nd7 7.d4 Ngf6 8.Bd3 Nxe4 9.Qxe4 e6 10.0–0 Be7 11.c3 Nf6 12.Qh4 Nd5 13.Qg4 Bf6 14.Re1 Qb6 15.c4 Nb4
Here Tal played the famous (and virtually forced) sacrifice:
16.Rxe6+ fxe6 17.Qxe6+ Kf8?!
This dubious symbol and the other symbols are Kasparov's. Since the 2 alternatives to this move lead to a draw and equality, it appears that Tal's whole conception was sound. But as it turns out, if you believe Kasparov's own notes, Portisch's 17...Kf8 should have been equally good.
18.Bf4 Rd8
Kasparov provides analysis to show that this is by far the best move.
19.c5 Nxd3! 20.cxb6
20.Bh6 forces a draw (Kasparov) after 20...Qxb2 21.Qxf6+ Ke8.
20...Nxf4 21.Qg4 Nd5 22.bxa7 Ke7?
Kasparov thinks that after 22...g6! (preferred by both Tal and Portisch) 23.Re1 Kg7 24.a8Q Rxa8 25.Qd7+ Kh6 26.Qxb7 , 'it is hard for White to count on anything better than a draw'. If that's so, then 17...Kf8 wasn't bad at all. Play might continue 26...Rxa2! 27.Qxc6 Rd8 and it will be hard for White to make progress.
23.b4!
23...Ra8?
Kasparov says, 'Perhaps the only chance of defending was 23...Nc7! (Aronin).' But then I (jw) think that 24.Re1+ Kf7 25.Qg3 Nd5 26.b5! ultimately favours White by a considerable margin, assuming that some analysis I did holds up. A crucial line is 26...Ra8 27.b6! Bxd4? (but a4-a5 is a problem and in any case White is much better) 28.Qg4, winning, e.g., 28...Bxb6 29.Qe6+ Kf8 30.Qd7 Bc7 31.Rb1 etc. Compare the game, which is not as dire for Black. Kasparov's '?' seems unobjective.
24.Re1+ Kd6 25.b5
25...Rxa7?
Let's see: We are only two moves from the end of the game and according to Kasparov, Portisch has already made '?' moves on both moves 22 and 23 (not to mention '?!' on move 17). Yet in spite of the suggestion that he has played so badly, Black still had an improvement that Kasparov himself suggests: 25...Rhd8! 26.b6! Nxb6 27.Qf4+ Kd7 28.Rb1 Rxa7 29.Rxb6 Kc8, which Kasparov calls 'more tenacious'. Since Portisch's 25...Ra7?? loses instantly, Kasparov might have done well to continue a bit further. I have taken a stab at this fascinating position. White's most obvious and probably best try is 30.Rb4! (or 30.Qf5+ Kc7 31.Rb1 Rxa2 transposing, whereas ; 30.Rb2 Rxd4 31.Qf5+ Kc7 32.Re2 h6 gives Black's king plenty of cover and it's not easy for White to make progress.) 30...Rxa2 31.Qf5+ Kc7 32.Rxb7+! Kxb7 33.Qb1+ Kc7 34.Qxa2 Rxd4, but then Black's c-pawn makes the situation unclear and very interesting. In view of the greater drawbacks of the earlier alternatives to Black's 22nd and 23rd moves as we have seen above, it looks as though both those moves were in fact the best ones in the position! That follows mainly from Kasparov's own analysis. This unjustified assignment of too many question marks to the loser's moves is almost always a case of annotating by result. On the other hand, I found nothing at all unobjective about the actual analysis that Kasparov gives for this game, which is too long and complex to present here. The actual game continued:
26.Re6+ Kc7 27.Rxf6! 1–0.
There follows Qg7+.
Everyman has done an exceptionally good job with these books. Both Predecessors volumes are hardbacks, well put together with a sturdy binding and attractively typeset. A two-column format works to perfection and this is physically the highest-quality work that I've seen from them. What to conclude? Like most readers I find Predecessors entertaining and absorbing as browsing material, in a way similar to the 'Greatest Tournaments in the History of Chess' CD that I reviewed in the previous column. Therefore, in spite of my skepticism regarding the extent of their contribution, I can recommend these books to the few players who don't already have them on their shelves.
Postscript: Recently, and well after I wrote the first draft of this review, some interesting exchanges took place on Nigel Davies' website http://www.chesschamps.com/. The first is dated May 18, 2004: When Raymond Keene protested the near-omission of Reshevsky, Davies says that Kasparov initially responded that [in Davies' words] he 'questioned his creative achievement in chess, saying he did not think he had contributed anything notable'. Keene replied with a string of Reshevsky achievements and noted his positional creativity. Kasparov countered that you could 'hardly compare Reshevsky to Geller from a theoretical point of view...In terms of critical opening theory Geller was way ahead.' This seems to me a limited view of creativity! At any rate, Kasparov goes on to say that 'I will have more to say about him in the Fischer chapter of Volume III'. [Apparently, this turns out not to be the case, as described below.]
It seems as though Kasparov has now changed his mind and significantly adjusted his material, probably having taken into account such criticisms as Keene's (or perhaps even a much earlier draft of my own review?). In the process of adding an extra Volume 4, he seems to have made room for Reshevsky, albeit in a rather bizarre way. On the same site on July 3rd (a week ago), Kasparov explained his new division of the series (with an extra volume) by saying that he could have moved 'Stein, Polugaevsky and Portisch to volume IV and try to find connections to justify having them there, where they don’t really belong', or to do what we are doing...To put Fischer in a new book, move Larsen with him and add material on Reshevsky and Najdorf.' He goes on to say that there will now be 21 or 22 fragments or games by Reshevsky!
Obviously Kasparov would have told Keene about this new and apparently substantial treatment of Reshevsky if it had already been planned at the time of their exchange. His response instead was a dismissal of Reshevsky's importance or contribution. Kasparov's reasoning is still strange, however. The peak of Reshevsky's career was in the late 1930s, and his world-class period extended through to the mid-1950s. Predecessors Volume 2 covers the 2nd half of the 1930s thoroughly (without including Reshevsky), and runs through the Tal's reign in 1960-61 and his later career. It includes: Bronstein, whose peak period is around 1948-58; Keres, who was extremely strong from the late 1930s well into the 1960s; and Geller, whose peak period was around 1960-67. So the period covered by this volume obviously includes and runs well beyond the years in which Reshevsky was on top. Okay, that's an omission; so do the authors quietly fit him into the next volume? No, because putting Stein (peak early 1970s), Polugaevsky (peak late 1960s to early 1980s), and Portisch (peak early 1970s to late 1980s) into a volume with their contemporaries 'wouldn't fit'. Thus they too must precede discussion of Reshevsky, who is now shuttled off to a Fischer-led 'Best of the West' Volume IV!
I guess all's well that ends well, and let it never be said that book critics can't have a beneficial effect. Nevertheless, I'm still left wondering what happened to poor Fine.
