Craig Werner's Reviews > The Book: Playing the Percentages in Baseball

For Sabremetric aficionados only. The collaborators are statisticians in the Bill James mold and they do their work well. Usually, that makes for deadening reading--I'll admit to glazing over the tables increasingly in the second half of the book. Most of the take-home messages are predictable for those who follow the field. 1) Most of the sportscaster wisdom about the relationship between hot streaks and past match-ups and future performance is simply illusion. The only real indicator of future performance is the player's ability level. 2) "Conventional sabremetric wisdom" about the counter productivity of sac bunts and steals is right on a general level, but needs to be put in very specific contexts, some of which justify the strategies.

The one chapter I really learned something new from was the one on how to put together a batting order. You do want the high OBP guy leading off, but forget the standard approach to the number two slot--that's where you want one of your two best remaining hitters (the other one goes in the clean-up slot). It was a slight surprise to see that you probably want a stronger hitter at 5 than at 3. Once you pass 5, just place the hitters in descending order of quality, no matter the rest of their profile.

Got a laugh out of the way they handled Earl Weaver's bizarre decision to put Mark Belanger--who you'd basically bat 12th if you could--in the number two slot. Terrible strategy (although Belanger did hit better at 2 than at 8 or 9, he was till awful). But, as they say, Weaver's response is that he can't hear the criticism because he's holding the World Series rings over his ears.