The owner of a firm using predictive modeling to locate baseball talent told a recent meeting of insurance actuaries that a key challenge they face is getting their management to invest in and support modeling techniques.
John Dewan, owner of Baseball Info Solutions and actuary, made his remarks Oct. 4 at the Casualty Actuarial Society Seminar on Predictive Modeling in Boston.
Mr. Dewan, according to a report of his remarks by CAS, said new actuarial techniques such as data mining and predictive modeling are increasingly being used by Major League Baseball teams to help them gain a competitive edge on the field.
He urged his audience to search out the unexpected and the new when collecting data. "The data has to be interesting to get the attention of management you are trying to impress or win over to your side," he remarked
He pointed out that as for sports statisticians, one of the key challenges facing actuaries is to get management to invest in and support the predictive modeling.
"Management in insurance doesn't always want to use the data or spend the money to collect the data. That's the challenge," he said.
Mr. Dewan said when baseball teams combine the output from his models with traditional scouting, baseball teams have the potential to achieve better results.
"There are more things you can do with data than you imagine. Our job is to find the nuggets of data that are interesting and relevant, and that reveal tendencies," he remarked.
Mr. Dewan explained that a number of teams now employ statisticians in addition to scouts to help determine the best prospects who can turn around a team's performance.
"You think of a sport where numbers mean something and it's baseball, but we're tracking things that in the past no one ever tracked," he related.
His company collects data on every play of every MLB game and enters it into a computer, Mr. Dewan said. The data collected is then analyzed in great depth. For example, each pitch is charted by type, location, and velocity.
Such a granular analysis using actuarial approaches can yield strong correlations. Mr. Dewan cited the example of comparing fastball velocity with earned run average. His analysis showed that pitchers who throw harder (based on their average fastball speed) perform better.
He noted that a velocity of 94 mph or more yielded an ERA of 3.84 in the regular season, compared to an ERA of 4.58 for velocity of less than 90 mph. The correlation was even stronger in the playoffs, with ERA improving to 3.42 for those pitchers who throw harder (velocity of more than 94 mph).
The Casualty Actuarial Society provides education and research for over 4,000 members to give them expertise in the evaluation of hazard risk and the integration of hazard risk with strategic, financial and operational risk.
© Arc, All Rights Reserved. Request academic re-use from www.copyright.com. All other uses, submit a request to TMSalesOperations@arc-network.com. For more information visit Asset & Logo Licensing.