Customer Reviews

Statistical Thinking in Sports

5 star:		(2)
4 star:		(0)
3 star:		(0)
2 star:		(0)
1 star:		(1)

 

 

Jim Albert is a professor of statistics and an avid baseball fan. In addition to being one of the leaders of the Section on Statistics in Sports, he has also written the book "Curve Ball" with Jay Bennett that covers the use of statistical methods in baseball. Another one of his books was written for teachers of elementary statistics courses. It teaches them how to make an introductory statistics course interesting by having all the examples come from baseball.

This book is a collection of papers edited by Albert and Koning that covers a wide variety of sports, all with the same theme of showing how serious statistical methods can be used to learn important things about the sport through formal statistical inference and predictions. Albert and Koning wrote an introductory chapter and Albert has also written a chapter in this book about clutch hitting in baseball. Hal Stern, the chairman of the Department of Statistics at UC Irvine and another avid baseball fan contributes a chapter on pitcher-hitter match-ups in baseball.

In 2002 at the Joint Statistical Meetings in New York, I was the program chair for the Section on Statistics in Sport, I am also a professional statistician with a great interest in sports and baseball in particular. Jan Magnus from the Netherlands gave an invited talk about the importance of the second serve in tennis at the invited session that I organized. His paper about myths in tennis is an outgrowth of that research. In addition to baseball, American professional and college football, tennis, European and English league soccer, ice hockey and basketball, these articles cover such sports as Australian football and Scottish football, the Olympic events and track.

Of all the topics covered, baseball is the dominant one. This book makes for interesting reading.

While many of the topics are of general interest, this book is not geared for casual fans. The level of mathematics and analysis will exceed your average reader, and excite math/sports geeks. To fully absorb all the articles, you will need somewhat more than an undergraduate degree in mathematics. The difficulty level varies substantially though, and most articles are explained well enough to understand the conclusions, even if the math is beyond you. There are a lot of original concepts analyzed here. From the perspective of a sports modeler, there are some great ideas that serve as starting points to achieving a greater understanding of sports. The ideas and manner of presentation are extremely appealing to sports bettors that approach problems quantitatively.

Some of the original ideas that translate directly into sports gambling include
Statistical analysis of FIFA World rankings (and approaches to handicapping soccer)
Handicapping approaches in Scottish league football, including the use of a bi-variate Poisson distribution for scoring. This one article is worth the price of the book
A peculiar analysis of momentum in MLB following a double play (which I expected to be worthless, but proved very interesting)
Myths in tennis (which had some brilliant analysis, giving a starting point to a player tennis model)

I cannot remember the last time I was so excited reading a new sports stat book.

The concept is promising, but the actual essays offer nothing of value. European soccer, long-distance running, and olympic events, all of which have almost no stats worth analyzing, are featured heavily. The one promising essay -- on home field advantage across various sports -- is already sorely outdated, as it includes only data from before 2006 and references sources from the 1990s.

In short, simply not worth the purchase or the read for a serious sports analysis enthusiast.