Dvoretsky's Endgame Manual [Kindle Edition]

Before we go further, lets note that this book is not an endgame encyclopedia; it is in fact a `teach yourself the endgame' manual, albeit quite a comprehensive one. This rather basic fact does not seem to be mentioned in most of the reviews I have seen, and the appellation `endgame manual' is rather ambiguous, so some people may be under the mistaken impression that this is a reference work rather than a self tutor. Lets note also that this is not for beginners - I would guess that players rated above Elo 1600 are the target audience. That's about my level, and I find the book quite easy to use.

For your money you get a large, attractive book which clocks in at 384 pages. It opens flat easily. The pages are well laid out. Grandmaster Yusopov writes the foreword, stating that his greatest victories are owed to Dvoretsky's training. Aagaard follows with a gushing preface, commenting that he feels as if Shakespeare has asked him to write a foreword to Hamlet (!). Skipping to the back of the book, there is an index of strategic and tactical techniques (e.g. "driving the king away by vertical checks") - a nice feature, but I'm not sure how useful it is. There is also an interesting bibliography, although Dvoresky notes that most of the material is sourced from his own training files.

The bulk of the book is, of course, dedicated to the various types of endgame the reader should learn. The length of each chapter varies considerably, based on the number of ideas and applicable techniques found in the type of endgame. Unsurprisingly the rook section is the longest, comprising almost a third of the book's length over four chapters.

The idea, writes Dvoretsky, is to present an endgame knowledge system. He states that in order to be a good endgame player the student should study relatively few endgame positions but understand them totally. These positions will then act as mental landmarks when playing. This is coupled with the memorization of important principles relevant to the general type of endgame. The positions and principles which form this core endgame knowledge system are printed in blue print; other material, which expands on the basics, in black print. Supposedly, the irrelevant positions found in other texts have been weeded out, leaving only those of maximum utility for the practical player, chosen on the basis of the authors extensive teaching experience.

However he emphasizes that there is no magic shortcut - memorization of principles cannot replace precise calculation, but can make it a lot easier by showing the right path.

Well, that makes sense I guess, but a couple of points. First, every endgame book I have seen claims to contain only those positions the practical player needs rather than the countless irrelevant positions found in `other books'. Second, I counted (rather quickly, so I might have missed a couple) the number of diagrammed positions in blue print. Two hundred and eight! Now, if I understand Dvoretsky correctly, I need to memorize ALL these `relatively few' positions - memorize them to the extent of being able to play them perfectly should they arise over the board. Yikes. And this is only the minimum, core system - not including all the material in black print.

Am I complaining? Not really. The reality is, I suppose, that there is a lot to the endgame, and thus a lot to study. No doubt this is the reason so many amateur players have excruciatingly bad endgame technique. However it is also encouraging to know that the general standard of endgame play is low, meaning that time spent studying will bring disproportionately large rewards!

I suppose the amateur player should take it on faith that the positions they are required to memorize will be of use to them in practical play and treat the book as an ongoing course of study. I have certainly noticed an improvement in my endgame play, and I have only just started working with the book. Dvoretsky writes in a vivid, engaging style that makes the precepts easy to remember. The memorization of positions is quite hard work but I can see how powerful this technique will become over time, increasing not only one's endgame skill but also general calculation and visualization ability. Imagine settling down to an endgame over the board, flicking through your mental library of memorized winning/drawing positions until you locate the one required, then aiming for that, aided by the general principles that apply to that type of endgame. That is a heck of an improvement over the typical amateur style of confused muddling toward a half remembered Lucena position.

There is enough in this book to keep any dedicated student going for a long time. This is the best general endgame book I have ever seen, and probably the most instructive chess book I have seen as well. Great stuff.

I saw a transcript of some fan questions Garry Kasparov answered several months back, and this is a quote from it:

"I read chess literature. But most modern books are short-lived. That's the difference between them and Bronstein's Zurich 1953! One book maybe not yet in English is Dvoretsky's endgame book. I was impressed with the material. That's not a short-lived book."

That should speak for itself. I would say the main difference between this book and Fundamental Chess Endings by Muller and Lamprecht (and I would recommend both) is that FCE covers slightly more (diverse) material, but DEM gives more explanation for understanding. DEM will perhaps increase your understanding of the endgame more, but an endgame encyclopedia (of which FCE is the best) is also important to give you a greater variety of endings to practice.

Definitely one of the best works on the endgame ever.

As way of introduction, I am a former USCF master who after many years of inactivity has recently taken a renewed interest in the game and has therefore decided to become acquainted with the current chess literature...which brings us to this highly acclaimed Endgame Manual by Dvoretsky.
You may know that Dvoretsky is a famous trainer of world class grandmasters, and has perfected a system of training by which he claims he can bring a 2200 level player to the level of grandmaster. His series of books are in many ways comparable to the famous "Think Like a Grandmaster" series by Kotov (and all are also highly recommended). So, the target audience for this book is, I would think, like all of Dvoretsky's books, the serious student who has already reached a high standard and is searching for a way to improve his game even further.

So, if you buy this book, and master it, will you also become a master of the endgame? You of course will be well on your way, and that alone is a good reason to add this book to your library, but still there are 2 major problems with it. The first is unfortunate, and somebody who has mastered the endgame would not commit it. Dvoretsky attempts to formulate universally general principals of endgame play, and then ignores the counter-examples that show the exceptions to his rules. This is OK for a general endgame text, but we are training future grandmasters here remember, and the promulgation of erroneous rules is not the way to do it. So, example, consider this (wrong) Dvoretsky rule on page 152 given in the section discussing rook and pawn vs. rook endgames, with the pawn on the fourth rank, and you will understand the problem with his approach:

"It is important to remember that in case of a knight pawn, cutting off the king by two files is not sufficient for a win." -- Dvoretsky

To support this rule, he then gives the standard position, with the black king on e6, black rook on b8, white pawn on b4, white rook on d1, white king on b3...It is very true, the rule applies here, but move the black king to either e7 or e4, and white to play wins. The sad and ironic thing here is that on page 147 Dvoretsky berates Dlugy for missing a draw against Alburt in the USA ch, Los Angeles 1991, because his adjourned position was in fact analyzed in the classic work on rook endings by Levenfish and Smyslov, and which Dlugy had at his disposal. Well, if Dvoretsky would care to obtain a copy of the classic Levenfish and Smyslov himself, turn to page 38 in either the algebraic or descriptive notation versions of the book, he will find the positions with the king on e7 or e4, showing his rule to be wrong. And, having gotten a hold of Levenfish and Smsylov, he should read the last chapter, the conclusion, where he would find:

"The reader who has accompanied us along the difficult path from the simplest of rook endings to the complicated ones involving numerous pawns may be expecting ready-made rules or formulae for the correct treatment of endgames. He will not find them. The solving of even simple rook endings depends on delicate nuances in the position, to elucidate which requires a considerable amount of analytical work" -- Levenfish and Smyslov, "Rook Endings"

That is also the reason the authoritative "Secrets of Rook Endings" by Nunn is 352 pages -- it is impossible to reduce the endgame to simple rules...and leads us to the second problem with the book...402 pages is just not enough to treat the subject of the endgame in an accurate fashion. So, if you are using Dvoretsky as your only source, it would be impossible to not have serious gaps in your endgame knowledge, which I will illustrate again with a Dvoretsky rook and pawn vs. rook ending example. Dvoretsky gives a fundamental position (in his exercise 9/1, page 144) with pawn on the seventh where by an ingenious maneuver, white can win. The position is white pawn on f7, white king on e7, white rook on g1, black king on h8, black rook on c7. Normally, the black king in such positions is on h7, but on h8, a new feature is introduced, the white rook can occupy g6. The solution main line then is: 1.Kf6 Rc6+ 2.Ke5 Rc8 3.Rg6!! Kh7 4.Rc6 Ra8 5.Kf6 with Re6-e8 to follow which wins. So what's the problem? Move the black rook in the initial position from the c-file to the b or a-file, and it's a draw. Dvoretsky does not tell you this. I would think a student reading Dvoretsky's exercise very likely would form the opinion that with black king on h8 white always wins such positions because the g6 square is always available to the white rook, but that is not correct.

(As an aside, in the first example, Nunn gives both the e7 and e4 king positions on page 143 of his book. Karsten Muller in his "Fundamental Chess Endings" does not give the e4 position. Muller may have used computer tablebases to check the accuracy of his work, but one cannot check an omission. Both the e4 and e7 positions must be given because the winning method is different in the two cases. In the second example, Nunn also correctly identifies with the rook on the b or a-file, the position is drawn. Muller does not give the second example.)

So, yes...buy this book by all means, but realize that you will be forced eventually to supplement it with either the books by Nunn, or the maybe a combination of the old Averbakh endgame series, and of course the classic Levenfish and Smyslov, together with the computer tablebases.