Project Description:

Pitch tipping in baseball occurs when a pitcher unintentionally reveals which pitch is to be thrown (e.g. fastball or off-speed) through subtle mechanical differences during windup. If an opposing team identifies these distinguishing behaviors, their batters can adjust their stances and swings, having potentially game-altering effects.

Despite its impact, systematic detection of pitch tips largely relies on manual frame-by-frame analysis of pitch videos by coaches, a subjective, time-consuming, often inconsistent endeavor.

We introduce PitchVision – a dashboard which processes pitch videos and mathematically identifies mechanical differences most effective at discriminating between pitch classes.

Project Photo:

Baseball pitchers often subconciously signal what kind of pitch is to come in the milliseconds before they begin to throw. We hope to identify these signals.