Eric Schiller's Improve Your Endgame! takes a very original approach in its modest 84 pages. The idea is to take 30 endgame studies, the great majority with reduced material (normally 4 to 7 pieces, counting the kings), and form an instructive book from them. The Introduction is a bit disappointing. Schiller admits right off that "there is no shortage of useful literature" on the endgame (a scandalous notion). But he does say that endgame study "tends to be rather tedious" and proposes "to make the acquisition of endgame essentials easy, fun, and aesthetically pleasing." That's more like it. The back cover states: "Practical endgame play is best studied by examining composed positions," a dubious claim, one challenged by Schiller himself in his book above.

But what about those studies? Even a Composition-Ignoramus like me can recognize names like Reti, Polerio, Botvinnik, Grigoriev, Cheron, Euwe, Weenink, Havasi, and the great Troitsky. As you can therefore imagine, these studies, even the simple ones, tend to be very elegant, and satisfying once you've solved them. I think it's true that a careful study of these compositions could do a lot for your overall skill level (including calculating ability). But I was fooled by the mass-market title: this is a rather advanced work, with only a few moderately-simple studies to start with, and it is well beyond the level of what I have characterized as "elementary" books. Club players who like problems should find this book rewarding, and I know that I had a great time with several of the studies, so it's only fair to categorize this as an "advanced" or "moderately advanced" work.

I did find an error and confused analysis in an early study (#3), which ends up in this standard position: White: K-e2; Q-a8 Black: K-g2; Q-h1; P-c7. (Black is in check). First, the diagram before this position had White's king on e5 instead of e2; and, from the correct position as I listed it, the analysis goes 1...Kg1 2. Qa1+, at which point Schiller says "2.Qg8+ would work just as well," missing 2...Qg2+. A small point, but it's possible that other studies are similarly flawed. In any case, I enjoyed this highly original book, which will mainly appeal to those who enjoy problems.