Springs benefits from mistakes

One of my goals in researching and writing Scientific Football 2005 was to find better ways to track the performance of individual defensive players in the passing game. We currently have ways of tracking defensive unit performance, but these stats are just the flip side of the offensive unit statistics. What I wanted was some of the same kinds of individual statistics that we have for offensive players.

The first statistic I started tracking was completion percentage. There are measurements of completion percentages for quarterbacks, and you can even find completion percentages for receivers (which is also something I track), but there has never been an attempt to determine the completion percentage against specific defensive players.

There are many reasons given why this isn't tracked. The most prevalent is that it is difficult to determine who was responsible for the coverage on a certain play. This argument implies that coverage responsibility can't be measured, but this clearly isn't the case.