Sporting one of the longest subtitles in recent memory, Dayn Perry's "Winners: How Good Baseball Teams Become Great Ones (and It's Not the Way You Think)", examines the common statistical measures at which playoff teams tend to excel. Perry considers nine areas–things like power hitting, relief pitcher deployment and team defense–and compares all playoff teams from 1980-2003 with their non-playoff counterparts.
For this reason, the subtitle is a tad misleading. Perry isn't interested (in this book anyway) in things like player development or offseason maneuvering, factors that one might ordinarily attribute to teams making the leap. His conclusions are drawn from ranking playoff teams in several statistical areas to discern which might be most important to winning. But in many cases, the statistical evidence is not overwhelming, and just a slight majority of playoff teams finished above the pack.
Perry leans heavily on Baseball Prospectus-derived statistics in a few chapters. Thankfully, explanations in clear terms are provided for metrics like support-neutral lineup-adjusted value above replacement (SNLAVAR) and leverage score, reinforcing their significance to his study. For example, in the chapter The Ace, he concludes that very few playoff teams have more than one ace-caliber starting pitcher, as measured by SNLAVAR, and that keeping runners off base and the ball out of play are paramount.
Perhaps most interesting, Perry presents strong evidence in The Deadline Game that it's not so much who general managers trade for as when they trade for them. He dots the chapter with profiles of the most active GMs, detailing their best and worst moves. And that pattern holds true for other chapters, where Perry has included brief bios of several prominent players–Pedro Guerrero, Phil Niekro and Dwight Gooden, included.
Be warned: If you pick up Winners you will be subjected to a boatload of statistical analysis. Refreshingly, what you won't find is blind adherence to past findings. As such, Winners is a worthwhile read for anyone interested in roster construction, e.g. for simulation league purposes, or in finding which performance metrics tend to underlie successful baseball teams.