Salford Systems Data Mines Baseball

March 10, 2011 SAN DIEGO--(EON: Enhanced Online News)--Data mining technology allows sports teams to find new indicators to measure player performance while helping them gain insight into athletes' future success, asserted Mikhail Golovnya, Salford Systems' senior scientist, during his presentation at the MIT Sloan Sports Analytics Conference in Boston last week.

According to Golovnya, data mining for sports analytics is revolutionary and could change the face of sport. For example, a typical data mining problem might focus on Major League Baseball's 2010 regular season performance, although it could be applied to other sports as well. In this case, each team has access to a player's statistics for the entire season including performance data, injury data, physical data and behavioral data.

A key question might then be: Which player statistics are associated with a team winning or losing the division? Data mining analysis highlights some unexpected findings based on the 2010 regular season. For example, teams with a home run-centered approach negatively contributed to the odds of winning the division titles because of larger rate of strikeouts, while wild pitches worked positively for the defense. In short, data mining results show a home run-driven strategy did not work in 2010.

To request the slideshow presentation from the conference, contact Heather Hinman, Salford Systems, hhinman@salford-systems.com.

For more information, visit www.salford-systems.com