Review : The secret of chess

A modern book on chess patterns

Recently the Bulgarian chessplayer Lyudmil Tsvetkov has published a book with the intriguing title “The secret of chess”. The title and introduction make clear that this is a very ambitious project, but the author is not a wellknown author or chessplayer. So the first and foremost question may be if Mr. Tsvetkov has the credentials to write a book about chess secrets that have not been uncovered yet. A short journey on the net shows us that his last recorded strength ratingwise is that of a strong clubplayer (candidate master in USCF terminology) and that he has been involved in evaluation methodology for the modern strong chessengines. That in itself is not an affirmation yet of that important question if quality has been guaranteed in this new and supposedly groundbreaking book.

It is not a surprise that controversy broke out as a result of this, with several sharp denials on one side, and the author stating his immense involvement in the last five years and putting that up is his credentials. Both sides have their points but this still leaves the potential reader in the dark.

Until the Australian grandmaster David Smerdon, whom I know as a strong and openminded chessplayer, wrote a review on “The secret of chess”. He has quite a few critical points to make but also writes that the book is ' a kind of work that legitimately has the potential to revolutionize how we think about chess”. That is a very interesting statement and made me buy the book at Amazon.de. This review is a result of collecting my first impressions, because working it through in detail is a herculean job, and will take much more time than I had thus far !

First of all, it is a book about pattern recognition, and we are all aware that this is a effective concept in chess tuition. The world famous “Steps” method is built on pattern recognition, and many tutorials use this concept. In this book, a strong clubplayer, strong enough to understand what chess is all about, not strong enough to face titled players with succes in a practical game, has done five years of research , methodically evaluating a multitude of positions, testing them (with the help of the strongest engines) and classifying the results in patterns. Quite a few patterns are clearly new to me, that is I have never considered them before.

What about the method that Tsvetkov used to research the underlying 'secrets of chess' , does it make sense ? He has used his brain to design a method and to draw up conclusions, and strong engines to help him to verify his evalutions. Values have been added to the pieces and certain characteristics of the patterns he has researched. Besides he refines the system by adding bonusses or penalties , for example a backward pawn in a pawnstructure, or a decentralised knight is given a penalty bonus on the basic value. There is also a sharp division between middlegame and endgame, where certain characteristics may have a different impact on the evaluation of the position, this sounds very logical. One of the conclusions that the author draws is that the bishop is a much stronger piece than the knight – except for blocked positions of course – and he really makes a point here. I remember Mikhail Gurevich after winning a game where he faced a socalled 'good knight' that he agreed with Victor Kortchnoi that “a bishop is 'always' better than a knight”.

A bit tongue in cheek, but most chessplayers may in fact underestimate bishops !

What I conclude about Tsvetkov’s methodology is that is is largely empirical, he is finding results, not inventing them.

What themes have been researched by the author ?

• material / piece values

• mobility

• pawns

• outposts

• imbalances

• king's safety

• general piece activity and coordination

Each theme is illustrated by a number of patterns, and these patterns have been evaluated by attaching values to them. Lyudmil Tsvetkov believes that the knowledge of these patterns will strongly improve the reader's strength in chess. He states that by these five years of intense chess research, he has gained quite a considerable playing strength, compared to what he understood about chess before he took on this task. That is possible, but can not be verified on games with engines alone, as games with humans require different skills. But, judging the examples he used in the book, and the enormous amount of time spent on them, I am ready to believe this has been very substantial to his current chess understanding which must have improved considerably. I remember books like Ziatdinov's “GM-Ram” claiming that there is a certain amount of 'positions' that you should know in order to play well. In this book it is a number of about 1.000.

My general opinion on the book is quite positive but it has it's weaknesses as well. Layout and use of English language are important, but are not major issues in my opinion, and not being a native speaker nor being an expert in layouts I do not feel qualified to comment on that. Could have been better, but not the most important aspects of the book.

What is clearly more important is the question if the goal of this book has been attained. A chess manual like this should inform and teach the reader about the (new) conclusions that the author has drawn up. The introduction stated that the book should be of use to a large range of potential readers, from weaker to stronger players and computer programmers. I clearly agree with Smerdon that the book is written like a mathematics textbook. Which harms it's accessibility for the general reader and makes it quite a tough job to work it through. What I am still missing is a good didactic method to transfer the revolutionary knowledge in this book to the reader in an easily digestable form. In this form I can imagine computer programmers are happy. But the other part of the potential customers would probably have liked to have seen the interesting conclusions in this book

in textual form and generalisations which makes learning a lot easier. It adds to acceptance of this new book to know how the author has worked and see proof of that. For example in tables with values, but I do not want to remember hundreds of them...

Still, I think this book has a lot to recommend it, is a highly modern manual of chess patterns, but there is still a lot of individual work involved to formulate the knowledge to our own purpose (a form that we can memorise easily).

G.Welling, IM