Mr Bacon Believe

Using positional data to evaluate offensive line and defensive line success

San Francisco


Technologies Used


Project Team

Ivy Chen

This team is looking for

Product Manager Investor


"Football games are won and lost in the trenches" -- football adage attributed to John McKay (legendary coach of USC) Sacks, hurries, and pressures are partially subjective and are not a continuous metric -- instead they're binary. With player tracking, you have the ability to look at protection in a continuous way. The offensive line has generally 2 responsibilities, creating gaps and pushing forward on run plays and protecting the QB on passes. We look at QB protection by tracking the distances from defensive players to the QB. We look at all the distances to the QB on passing plays and observe how close is the nearest defender. We visualize this with the tracking data, and then we show that it's a meaningful metric to understand who is winning at the line. We show heightened success on plays with good protection, as measured by distance to QB, and heightened failure on plays without it. Use cases: GM -- Do I need to sure up my offensive line, how much do I need to? Do I need to sure up my defense, and in particular, my defensive line? Now I have a metric to know pressure that's continuous, and the individual units effectiveness at generating pressure can be measured. Coach -- Do I create enough pressure (if not, should I blitz more or some other creative schemes)? On the offensive side, should I screen more or keep an RB in the backfield? Fan (of fantasy football) -- We don't have offensive lines in fantasy (we now have a stat for the collective performance, you could imagine another position, OL) Fan (of a team) -- Now you don't just need to be angry at Ted Ginn for dropping footballs, you can measure offensive line players getting beat (aside from just sacks).