At The Sybase Open (Page 4)

In the first year John and Nasif filmed from this initial position, which gave a fascinating view of the progress of the ball flight as it progressed from player to player. One thing was clear from viewing the video--it was visible to the naked eye that the distance the ball traveled on each video frame egan to decline almost immediately, and decreased noticeably again after the bounce.

Clearly the ball was slowing down, but how much? By making some simple measurements with a ruler on the television screen, John estimated that the ball might be losing half its speed between hits. But how to measure this accurately and present the data in a scientific fashion? That spring and summer Nasif puzzled in his spare time over the best possible methodology for calculating this from the video, and in the fall of 1996 began writing an analysis program.