What do you think?
Rate this book
Percentage baseball
As a result of a lunchtime conversation with Professor Wendell Garner concerning the productiveness of the sacrifice bunt, Earnshaw Cook took on the three-year task of presenting a formal analysis of baseball. His analysis, explained in terms perfectly clear to anyone with college freshman level mathematics, suggests that no one has ever known the true percentages, and if anyone did know them he could manage almost any team into the top ranks of major league baseball. Among other theories that Cook attacks with irrefutable mathematical findings are the benefits of the sacrifice bunt, the use of relief pitchers, the traditional batting order, the hit and run play, and the standardization of baseball itself. As with almost any serious innovation, the first edition of this book met with bitter controversy and criticism from some baseball fans, team managers, and sportswriters. James Gallagher in Sporting News wrote, "I do not understand how the Baltimore mathematicians reached their controversial conclusions, but in my book any generalizations about baseball have to be wrong." Yet in 1964 this "Baltimore mathematician," using his scoring index, K.2 factors, base-scoring equations, etc., predicted that the hometown Baltimore Orioles would finish in fourth place behind, in order, New York, Chicago, and Minnesota -- with perfect accuracy!
345 pages, Hardcover
First published January 1, 1966
1 person is currently reading
26 people want to read
Ratings & Reviews
Friends & Following
Create a free account to discover what your friends think of this book!
Community Reviews
Displaying 1 - 2 of 2 reviews
May 6, 2026
Researcher Earnshaw Cook investigated player and team statistics of baseball, not with computers but with pencil, paper and slide rule. This book, published before I was born, is the result of that research. The statistics fall a little short of reality, though.
The math here is solid, though a little dense. What Cook works with are the results of baseball seasons. He tests his theories on this data, which provides an imperfect measurement of what might happen. For instance, his thoughts on the sacrifice bunt have been proven today; his thoughts on reducing pitcher batting have not been tried but, with universal DH, will probably never be. Then there's Ohtani...
Cook consulted on the Manhattan project, but studied the stats more than the game of baseball. He may have been swayed by conclusions, as he wanted to prove Cobb was a better player than Ruth and that home runs are not that useful.
I was drawn to this book for love of baseball and math, and may apply some of the math here to tabletop baseball games (dice are more predictable than people). These pages are math dense, and I wouldn't recommend this book to just anyone. Apparently John Thorn and Pete Palmer simulated some of Cook's proposals in The Hidden Game of Baseball, which may be the better book to read.
Still and all, I am impressed with the measurements taken and the tools used. For me, this book was close to 4 stars out of 5.
The math here is solid, though a little dense. What Cook works with are the results of baseball seasons. He tests his theories on this data, which provides an imperfect measurement of what might happen. For instance, his thoughts on the sacrifice bunt have been proven today; his thoughts on reducing pitcher batting have not been tried but, with universal DH, will probably never be. Then there's Ohtani...
Cook consulted on the Manhattan project, but studied the stats more than the game of baseball. He may have been swayed by conclusions, as he wanted to prove Cobb was a better player than Ruth and that home runs are not that useful.
I was drawn to this book for love of baseball and math, and may apply some of the math here to tabletop baseball games (dice are more predictable than people). These pages are math dense, and I wouldn't recommend this book to just anyone. Apparently John Thorn and Pete Palmer simulated some of Cook's proposals in The Hidden Game of Baseball, which may be the better book to read.
Still and all, I am impressed with the measurements taken and the tools used. For me, this book was close to 4 stars out of 5.
May 9, 2020
More interesting and thought-provoking as an analysis of an incredibly bright man's statistical disaster than a triumph over the conventional strategies of baseball, this book should only be read by those wanting to grow in their application of statistical process, and not those looking to be enlightened in their understanding of baseball.
Earnshaw Cook was a genius, he worked on the Manhatten Project, had graduated from Princeton, and capped an illustrious professional career with a role as an educator for decades at an elite East Coast institution.
However, by starting from a failed premise (Cook hoped to prove that Ty Cobb was a more impactful player than Babe Ruth) and working backward, this incredibly bright man threw away all the intellectual gifts at his disposal, and instead wrote a confusing, fast-paced, and dense manuscript of analytical insanity. Cook assumes that his readers will be lock-in-step with his mathematical analysis at every turn, but even I, who have worked for years in basketball analytics, struggled mightily to keep up with the massive logical and mathematical leaps that seemed to come around every other paragraph. His metrics are confusing, and his process is even more so. Somehow, Cook manages to divorce the reaction of runs entirely from hitting the long ball that he so despised, the only action in the game of baseball that is guaranteed to generate at least one run.
Another interesting note that can be taken away from this book for any analytics hopeful is the importance of underselling your abilities within these fields, purely as a survival tactic in fields with incredible amounts of variance.
Cook believed that any team could gain 250 runs over a then 154-game season from embracing his strategies. To put this number into context, in 1960 the New York Yankees scored a league-leading 746 runs on the way to a league-leading 97-57 mark before losing the World Series in 7 games. If they had embraced Cook's strategies (and no team at that time did, as they involved starting relief pitchers and pinch-hitting for them at their first at-bat, for example), they could've increased their run out-put by more than 33%, and their expected win/loss record would've been one of the best records of the modern age, a mark of 110-44.
Then, in 1981, John Thorn and Pete Palmer ran simulations in their far superior and very recommendable book, The Hidden Game of Baseball, and found that Cook's strategies would've actually resulted in a net loss of runs.
Whoops.
I will treasure my copy forever as a memento to what can happen when good intentions come before the mathematical process in the field of sports, and to remind myself that considering I will never work on anything as complex as an atomic bomb or graduate from as prestigious a university as Princeton, I should learn from Cook's mistakes instead of repeating them.
5 stars for effort, background, and intentions.
1 star for execution, writing prose (almost academic to a point of insecurity), and hubris.
Earnshaw Cook was a genius, he worked on the Manhatten Project, had graduated from Princeton, and capped an illustrious professional career with a role as an educator for decades at an elite East Coast institution.
However, by starting from a failed premise (Cook hoped to prove that Ty Cobb was a more impactful player than Babe Ruth) and working backward, this incredibly bright man threw away all the intellectual gifts at his disposal, and instead wrote a confusing, fast-paced, and dense manuscript of analytical insanity. Cook assumes that his readers will be lock-in-step with his mathematical analysis at every turn, but even I, who have worked for years in basketball analytics, struggled mightily to keep up with the massive logical and mathematical leaps that seemed to come around every other paragraph. His metrics are confusing, and his process is even more so. Somehow, Cook manages to divorce the reaction of runs entirely from hitting the long ball that he so despised, the only action in the game of baseball that is guaranteed to generate at least one run.
Another interesting note that can be taken away from this book for any analytics hopeful is the importance of underselling your abilities within these fields, purely as a survival tactic in fields with incredible amounts of variance.
Cook believed that any team could gain 250 runs over a then 154-game season from embracing his strategies. To put this number into context, in 1960 the New York Yankees scored a league-leading 746 runs on the way to a league-leading 97-57 mark before losing the World Series in 7 games. If they had embraced Cook's strategies (and no team at that time did, as they involved starting relief pitchers and pinch-hitting for them at their first at-bat, for example), they could've increased their run out-put by more than 33%, and their expected win/loss record would've been one of the best records of the modern age, a mark of 110-44.
Then, in 1981, John Thorn and Pete Palmer ran simulations in their far superior and very recommendable book, The Hidden Game of Baseball, and found that Cook's strategies would've actually resulted in a net loss of runs.
Whoops.
I will treasure my copy forever as a memento to what can happen when good intentions come before the mathematical process in the field of sports, and to remind myself that considering I will never work on anything as complex as an atomic bomb or graduate from as prestigious a university as Princeton, I should learn from Cook's mistakes instead of repeating them.
5 stars for effort, background, and intentions.
1 star for execution, writing prose (almost academic to a point of insecurity), and hubris.
Displaying 1 - 2 of 2 reviews