To judge by a recent trend in publishing, chessplayers are very bad at keeping secrets. "Secrets" seems to be the buzz-word of the moment in book titles, with a number of players rushing to put their secrets into print.

After 60 pages of the first volume of Secrets of the Russian Chess Masters by Lev Alburt and Larry Parr (WW Norton & Co; pounds 18.95 for each of two volumes) we are still on exercises such as "which piece can the white knight not capture?" The second volume, subtitled "Beyond the basics", goes into elementary positional and tactical principles, but calling this elementary set "Secrets of the Russian Chess Masters" suggests that the Russian Official Secrets Act is even more stringent than our own.

John Nunn's Secrets of Practical Chess, pounds 14.99 from the new English specialist chess publisher Gambit, takes up long after the Russian secrets leave off. I was delighted to see that he begins with a criticism of the analytical method advised in Alexander Kotov's Think Like a Grandmaster. Rather than working his way along every branch of Kotov's rigid "Tree of Analysis", Nunn prefers a flexible approach, letting ideas suggested by one variation lead to improvements in other lines. Here's a study he gives as an example, composed by J Gunst in 1922. It is White to play and win.

At first it looks as though the game must end in a draw: after 1.Bxd7 Kc7, or 1.Bb7 Kc7, or 1.Ba6 Kc7 Black wins one of the white pieces at once. Then you notice the idea of 1.Ba6 Kc7 2.Kc5 Kxb8 3.Kd6 Ka8 4.Kc7 and mate with Bb7 next move. But after 1.Ba6 Kc7 2.Kc5, Black plays 2...d6+! 3.Kd5 Kxb8 4.Kd6 Ka8 when 5.Kc7 is stalemate. You find the right answer by jumping back (in a way Kotov would have condemned) to a line already analysed: 1.Bb7! Kc7 2.Ba6! Kxb8 (or 2...d6 3.Nc6) 3.Kd6 Ka8 4.Kc7 d6 5.Bb7 mate. A highly instructive book, if a little expensive.