What Timman Knew

Chess analysis is a conversation.

Jan Timman passed away last week, but like many chess luminaries he lives on through his games and writings. After I learned of Timman’s death I paged back through his first book, The Art of Chess Analysis, published in 1980, the year of my birth. The Art of Chess Analysis is an anachronism disguised as a game collection: to read it today is akin to playing blitz with an analog clock or dusting off a stack of sealed move envelops. It can only be understood in the context in which it was written, a time before chess engines ruled the earth, but its lessons are still deeply valuable to anyone who wants to master chess.

I was introduced to The Art of Chess Analysis in the late 90s by my chess coach, Ray Freeman, and we poured over some of it together. I was unfamiliar with the great games played in the post-Fischer era, but I already understood the importance of game analysis. I’d spent more than a few hours analyzing some of my more interesting efforts, but Timman’s approach was much more intense:

Far more useful was Botvinnik’s advice to analyze games at home and then publish the analysis. As he put it: ‘During play your analytical work is continually being tested against your critically-minded opponents, but in home-analysis it is very easy to be unobjective. To fight this tendency and to get away from poor analysis it is useful to publish your individual analytical work. Then you are subject to objective criticism.’

So I began to analyze games. Luckily, there was an independent magazine in Holland, Schaakbulletin, which was eager to publish this analytical work. In the framework of ‘The Game of the Month,’ a more-or-less thorough analysis was published in every issue. I limited myself to games between top players, hoping for as much critical comment as possible. In that respect the result was disappointing: only after the game Fischer-Petrosian, which I had worked on for about forty hours all told, did two reactions arrive (both of which have been gratefully worked into this book).

Can you imagine the publication of such a book today? Not only does it fail to promise the reader rapid success through memorization of some dodgy gambits, it says the quiet part aloud: “the game Fischer-Petrosian, which I had worked on for about forty hours all told.”¹ That’s a lot of time to spend on a single game!

Even more fascinating is Timman’s admission that he didn’t write the book for you, the reader, but for himself! Your job is to work carefully through all his analysis, ask some questions, find a hole in a variation or two, and send your ideas back to the author to help him continue his own chess development. This is both an incredible invitation — to do the same kind of analytical work as one of the world’s top grandmasters — and a daunting challenge — if it took Timman forty hours to work through Fischer-Petrosian, how long is it going to take me? No wonder he only received two responses.

Whether or not he was writing into a void, Timman understood that you could only go so far on your own. Every great player needs an analytic partner or, better yet, a team to challenge their bad ideas and validate their good ones. Compared to the Soviet Union, the west was a chess wasteland, and Timman had to seek out chess conversations wherever he could find them.

The idea of analysis as conversation predates Timman, of course, nor is it unique to chess. All scholarship exists as a set of arguments within a larger dialogue, referencing its predecessors as both foil and context. In this way, Fischer’s analysis of his rook ending with Botvinnik in My 60 Memorable Games is framed as response to Botvinnik’s previously published analysis. Fischer, usually concise, spends pages trying to refute Botvinnik’s claim of a draw, and the result is one of the most fascinating moments in the book. Similarly, many of Dvoretsky’s manuals reflect his own students’ work on training positions, highlighting their varied approaches and ideas. Even Tal found it necessary to split himself into two personas to write about his chess career, imagining himself as both journalist and master.

The key to this kind of analytical work, the reason it proved so compelling, is that no one knew the right answer. There was a time — and here we return to the idea of The Art of Chess Analysis as chess anachronism — when it was conceivable that anyone with a good idea could add to the conversation in a meaningful way. When I annotated a bunch of my own games in the late 1990s, I sent a copy to my sparring buddy and future-IM David Pruess, who sent me a set he had annotated in return. I don’t remember trying to poke holes in his analysis, but I was impressed by how different his understanding of chess (dynamic, aggressive) was from mine (stodgy, materialistic), and gaining insight into his view of the game was very important for my own improvement. Chess analysis as conversation is about moves and variations, but it’s also about contrasting approaches and styles.²

The world in which Jan Timman wrote The Art of Chess Analysis is well and truly gone. This is a John Henry versus the steam engine debate, and though I’m positioning myself on the side of the steel driving man, it would be silly to ignore the benefits brought by the chess engine. There is a value in knowing the right move or moves in any given position, to having a coach that can instantly give feedback to any game, and to getting rapid, accurate reports from the top tournaments. Human insight combined with machine precision has dramatically raised the quality of chess analysis, as demonstrated by the My Great Predecessors series. Kasparov introduces the project by writing,

Awaiting you is a wonderful collection of masterpieces, created by the best chessplayers in the world and studied under the microscope of the latest analytical computer programs; hence — a great number of amazing finds and discoveries. I hope that this work will make it possible to see the colossal evolution of chess during the past one hundred and fifty years, which is fully comparable with scientific and technological progress.

That’s an alluring vision,³ but it’s the view from the vantage point of the steam engine, make no mistake about it. My Great Predecessors appears to be in dialogue with the ghosts of annotators past, but when one person has access to the truth (no matter how nebulous the concept of chess truth may be) we’re talking about corrections from a higher authority, not a conversation among relative equals. Kasparov cites Timman’s work in final volumes of The Art of Chess Analysis, but as arbiter rather than colleague. In this context, Timman’s book feels like a curiosity, one that has enough historical value to avoid consignment to the dust bin, but a text that is increasingly out of place with how chess understanding and improvement is currently marketed. After all, why engage in a debate about a position that the computer can solve instantly?

FM Nate Solon wrote a provocative article a few years ago called “Why Chess Books Don’t Work.” His critique comes down to three main points:

1. Books lack an explicit theory of learning.

2. To the extent that books have an implied theory of learning, it’s that people absorb knowledge simply by reading it transcribed on a page.

3. As a learning strategy, this doesn’t work.

That strikes me as reasonably fair: many chess books present information without any explanation or method for how you are to retain and make use of what you’ve learned.⁴ I would argue that The Art of Chess Analysis is one of the rare books that makes its theory of learning explicit. Timman argues that you have to analyze chess games: your own games, the games of others, it doesn’t matter as long as you engage yourself deeply in the work. Forty years ago, this was the only way to analyze. These days you have to resist the allure of the engine and find others willing to do the same kind of work and engage in the same kinds of conversations as well. Timman is suggesting that his readers do more than read his work: they should write their own books full of their own annotations, warts and all. There is, after all, no way to learn other than making mistakes.

This is a long way of saying what it took John Barth eight words to express: “The key to the treasure is the treasure.” I played an interesting game recently, a rare victory against a grandmaster at classical time controls, which decided I should annotate in the old way, by hand, without the aid of the machine. This is an immensely valuable process, as it teaches persistence and acceptance of uncertainty, a humbling of the ego against the backdrop of the infinity of chess. And yet — it’s been a long time since I analyzed a game this way, and it was more challenging than I though it would be. I found my desire for clarity to be almost overwhelming at times, even though I knew that consulting the engine would derail my own hard work. This article was initially intended to be the introduction to that game, but annotation by hand is a slow process and the introduction kept growing and growing, far beyond the few paragraphs that I initially had in mind. In any event, I’ll publish the game when the annotations are done with the same invitation that Timman gave in The Art of Chess Analysis, for “as much critical comment as possible.” Till then.

1. Emphasis mine.

2. I have a theory that it’s harder to discern the style of most of the top young players because they are all, more or less, trying to emulate the computer — as opposed to the players of the past who had to work out their own approaches and preferences.

3. One that I am certainly not immune to. Longtime readers of this blog may remember when I did the same kind of corrective work with some of my old annotations: “The Pursuit of Chess Truth Parts One and Two.” The insights that came out of this work are interesting, but I didn’t get better at chess by following the computer’s lead.

4. I disagree more with his emphasis on Chessable as a useful tool, since I think that a focus on memorization provides superficial improvement at best.