The numbers game: can insurers learn anything useful from baseball? Absolutely.

Michael Lewis' bestseller, "Moneyball," explains in fascinating detail how the Oakland A's became one of the most successful teams in professional baseball despite a payroll substantially smaller than nearly all of their less successful rivals. Not only is Lewis' book an exciting read, but it has potential implications for insurance as well as for baseball.

A key player in the drama is Billy Beane, the A's general manager, whose background and circumstances made him willing to abandon conventional scouting wisdom in selecting draft picks for his team. In its place, he substituted the analysis and advice of statisticians immersed in the vast quantities of data now available on amateur and professional players and games. Substituting statistics for scouting was heretical, but Beane recognized that innovation of some sort was essential to overcome bis financial competitive disadvantage. His success in exploiting statistical analysis has since inspired the Boston Red Sox and the Toronto Blue Jays to adopt a similar approach to recruitment. Perhaps as a consequence, both Oakland and Boston came close to winning pennants this past season.

Does any of this have a bearing on insurance? Yes, for several reasons.

First, both baseball and insurance rely heavily on traditional statistics. In baseball, the box score was invented in 1845 and improved in 1859 by British-born journalist Henry Chadwick. In insurance, the statistics that insurers, analysts and rating agencies pay attention to are those based on the numbers that regulators require to be reported. In both cases, what doesn't get reported is invisible, and has no impact on management.

Second, these traditional statistics can be misleading. For example, statistical analysis conclusively demonstrates that getting on base, whether by hits or walks, is far more important to a team's win-loss record than RBIs (runs batted in), the traditionally reported number. In insurance, a firm's premium-to-surplus ratio is typically treated as a measure of underwriting risk despite the fact that an increase in the ratio utterly fails to distinguish between firms that are increasing their exposures and firms that are increasing their pricing while maintaining the same exposures and, therefore, the same risk.

Third, rather than producing more adequate statistics, managers in both baseball and insurance rely heavily on experience and on accepted stereotypes. The scout with 30 years experience is hard to ignore, as is the underwriter with comparable seniority. Both are reluctant to recommend actions that defy conventional wisdom, whether by recommending draft picks who don't look like consummate athletes, of by taking on risks that "everyone knows" are unacceptable. The personal lines pricing actuary at a major firm once told me that his credibility would vanish if be recommended anything other than high rates for all motorcyclists, no matter what the data indicated.

Experience, however, is simply not adequate for making critical distinctions. In baseball, the difference between a .300 hitter and a .275 hitter is one hit every two weeks, a difference not readily visible to even the most avid fan or scout. Likewise in insurance, significant differences in loss frequency or severity can be extremely difficult to discern at the level of individual risks, especially since insurance typically focuses on rare events.

The success of the Oakland A's owes as much to the recent availability of abundant data--and data of the right kind--as it does to the willingness of Beane to challenge conventional wisdom. Nonetheless, the fact remains that other managers ignored the new data, whereas Beane and his statisticians not only welcomed it, but sliced and diced it in any way that could help them to better understand the game and how they could win it. Lewis began to appreciate just how far this had taken them one afternoon when he joined them in watching a game on TV. After discovering that his reactions to both spectacular and embarrassing plays differed substantially from theirs, Lewis gradually realized that he was "not only watching the game differently but ... watching a different game."

Given the similarities between baseball and insurance, there might--just might--be an opportunity for someone in our own industry to similarly combine managerial vision with analytic expertise and so break through to a whole new level of understanding. If it can happen in baseball, then surely it's possible in insurance.

William H. Panning, a Best's Review columnist, is executive vice president and managing director at Willis Re Inc. He can be reached at [e-mail].