Sign In

Communications of the ACM

ACM TechNews

Digital Eyes Will Chart Baseball's ­nseen Skills

View as: Print Mobile App Share:
Minnesota Twins Outfielder Carlos Gomez

A new camera and software system will collect data for statistical analysis of players' defensive performance.

Credit: Minnesota Twins

The game of baseball could be significantly affected by a new camera and software system capable of recording the precise speed and location of the ball and every player on the field, which will generate statistics that will grade players with greater accuracy. The data gathered by this method, which is in final testing, will be revealed to a group of baseball executives, statisticians, and academics on July 11.

The system's software and artificial intelligence algorithms, which will analyze what will likely be petabytes of raw data, are still under development. "We've gotten so much data for offense, but defensive objective analysis has been the most challenging area to get any meaningful handle on," says Cleveland Indians general manager Mark Shapiro. The testing and refinement of the camera system has been carried out by Sportvision, which is collaborating with Major League Baseball Advanced Media.

In the San Francisco's Giants ballpark, a quartet of high-resolution cameras sit on light towers, capturing everything that happens on the field in three dimensions and transmitting it to a control room. Software tools link movements to balls, runners, and fielders, and more than 2 million meaningful location points are recorded in each game.

"[The system] can be another tool to help you improve in areas of the game," says Toronto Blue Jays center fielder Vernon Wells. "People will learn about playing defense, which has gone by the wayside as people have cared so much about offense and hitting the ball out of the ballpark."

Major League Baseball Advanced Media's Bob Bowman says he would like the data to be available in some form to statistically minded fans and academics.

From The New York Times
View Full Article


Abstracts Copyright © 2009 Information Inc., Bethesda, Maryland, USA


No entries found

Sign In for Full Access
» Forgot Password? » Create an ACM Web Account