I like the basic idea - a book in which a stat guy debates uses interviews and stats to decide who's better and who's the best in NBA history.

I like the writing, too, as Kalb reports statements from a number of hall-of-fame coaches and players in ranking the fifty greatest players of all time. The quotes and analogies (Tim Duncan vs. Pete Sampras, Wilt Chamberlian vs. Babe Ruth, Michael Jordan vs. Willie Mays) that the author uses to illuminate his many points make for a pretty great read.

My one issue with the book is that for a book written by a stats guy, it doesn't actually appear to be too statistically-based. Where Bill James's book Win Shares took a statistical approach to first developing a systematic method for evaluating a player's worth (comparing players across positions and eras) and then used the method to compare players, Kalb almost appears to have gone the opposite way. he seems to have ranked players based on his feelings and the survey results he got from great coaches and players who saw many of the players, and then used stats to back up his arguments. He doesn't use the same stats every time, and he doesn't always pick stats thoroughly, but seems rather to choose stats that support his arguments. It's no less interesting a book because of this difference in philosophy, but for a book meant to be the definitive book on the subject, it weakens the author's arguments.

That being said, I'm looking forward to his next book.