In 2006 MLB started installing a system of high-speed cameras trained on the area between the pitcher and catcher in every park, the Pitchf/x system. These cameras capture about 30 pictures of each pitch as it travels between the pitcher and catcher, an algorithm locates the ball in each picture, and the equations of motion are fit to the 30 locations. The parameters of that fit give a wealth of information about the pitch: its release point, speed at release, the location where it crossed the plate, the amount it ‘broke’ or ‘curved’.
With this information about every pitch throw, we can look baseball questions like never before: Where are the best and worst places to pitch in the strike zone? Who are the pitches who routinely hit these locations? What is the optimal speed difference between a fastball and a change up? Is the size and shape of the strike zone constant or does it change based on the state of the game? …
Here are PDFs of slides from baseball-related invited talks I have given.
Creating contour and ‘heat map’ graphs to display PITCHf/x data. PITCHf/x Summit, San Francisco, CA, July 11, 2009.
Using PITCHf/x to measure pitch success by location. SABR 40, New Technologies in Baseball Panel, Atlanta, GA, August 7, 2010.
Using FIELDf/x to assess fielders’ routes to fly balls. PITCHf/x Summit, San Francisco, CA, August 28, 2010.
Disentangling success in baseball with ball-tracking technologies. Skill vs. luck: disentangling success in complex systems, Ann Arbor, MI, February 11, 2011.