ADVERTISEMENT
Published: October 7, 2007
'Super Crunchers,' by Ian Ayres (Bantam, $25)
This book makes a strong argument that super crunching, the use of huge amounts of statistical data to predict an outcome, is challenging the ages-old dependence on human experience and expertise.
Ayres takes what could be dry information about data usage and makes it interesting by looking at super crunching research in topics as varied as United States Supreme Court decisions to credit card interest and consumer purchasing.
It starts with wine. Every year, experts judge new wines, predicting which are going to be exceptional. Princeton economist Orley Ashenfelter thought he could do better, using decades of weather data from France's Bordeaux region. Crunching all this data, he found that 1) low levels of rain and 2) above-average summer temperatures end up making 3) exceptional Bordeaux wines. Experts mock him, but Ashenfelter has been right more times than not.
Ayres goes into other detailed examples of super crunching, from predicting how Supreme Court justices will vote (with a 75 percent accuracy rate) to credit card companies super crunching to determine how high they can set an interest rate and still get people to sign up for a new card.
The most controversial chapter is the one on health care, in which Ayres quotes studies showing doctors would make better diagnoses if they played the percentages, as determined by super crunching.
Statistical studies have found the same thing is true in education: Teachers who follow a script get better results from students.
Mary Patrick of Tampa is a freelance writer.
ADVERTISEMENT
Advertisement
TBO.com - Tampa Bay Online ©2009 Media General Communications Holdings, LLC. A Media General company. Member Agreement | Privacy Statement | Work With Us
| * To: | |
| Your Name: | |
| Your Email Address: | |
| Personal Message [optional]: | |