Recently I read Behind-the-Scenes Baseball: Real-Life Applications of Statistical Analysis Actually Used by Major League Teams...and Other Stories, and I came to a realization while going through it. I really have no idea what those with the position of statistician do for a major league franchise. I understand that they provide analysis (no kiddin') but I never really thought about the day-to-day things, or how exactly the information was presented, and how the manager reacted to it. Doug Decatur's book certainly helped to answer a few of those questions.
Decatur worked as a statistical consultant for the Reds, Brewers, Cubs and Astros, as well as for player agent Myles Shoda. The first section of the book details some of his stories from his time with those clubs, from particular analysis he gave that worked out -- such as sugggesting Greg Vaughn be put in the Brewers starting lineup -- to various other ideas, like dealing for Ron Gant (Braves minor leaguer) and Cecil Fielder (playing in Japan) and putting them on the 1990 Reds, who would win the World Series even without the help of those two players.
What I found most intriguing is how well the Decatur and his managers got along. In the third section of the book, Decatur shows how Astros manager Phil Garner applied statistical analysis to the 'Stros daily in order to make up for lost ground. Using Brad Lidge for more than an inning, rearranging the Astros batting order in a "bunched" manner, and calling up Brandon Backe were all Decatur-inspired moves, and all were essential to the Astros sneaking into the playoffs in 2004.
Decatur also gives off the vibe of being a very passionate fan, which helps the readability of the book. For those of you on the fence about statistical analysis -- although I don't think anyone who reads this particular blog has any such problems -- Decatur helps to show that yes, lovers of sabermetrics are capable of enjoying a baseball game and being passionate fans.
The center of the book is made up of the GM IQ test, which I found fascinating. The questions are made up from various publications and research projects, from some of Bill James Abstract work to more recent Baseball Prospectus studies. For your viewing pleasure, and in the hopes that you will pick up a copy of the book, here are a few sample questions for you to try out:
1. Which team would be expected to win the most games?
(A) a team which scores 1000 runs and allows 900 runs
(B) a team which scores 800 runs and allows 700 runs
(C) a team which scores 600 runs and allows 500 runs
(D) all teams will win the same number of games
2. True or False: Almost all good young pitchers with strikeout rates below 4.00 per game disappear quickly
3. True or False: On the average, closers pitch worse in "non-save" situations than they do in save situations.
4. Which free-agent strategy works best:
(A) sign to "fill a need"
(B) sign the "best free agent" on the market
(C) "whole-scale" signing
(D) let all your players become free agents and then blame your 100 loss season on the size of your market
5. Late August 2004. The Reds decide to dump Barry Larkin. They have two choices for the replacement shortstop: Felipe Lopez or Andy Machado. Should the Reds immediately make a decision on who should play shortstop, or should they alternate each player, giving them 20 starts each the rest of the season and make a decision based on those 20 starts?
I will provide the answers at a later time, but I'd like people to comment first. Astros manager Phil Garner scored a 97 on this 100 question test. I must say that this book, combined with my respect for Garner's use of Lidge in the 2004 playoffs, has changed my opinion of him as a manager. I'm very curious as to how many other managers are as open-minded about statistical analysis as he has shown himself to be.
This book was a quick and worthwhile read, and it has been published as a paperback, which means your wallet isn't going to take a heavy hit either. It is certainly worth a purchase and your time if you are at all interested in the goings on inside of a major league front office with a statistical consultant.