The model that I am using though is trying to capture all the plays during the season to determine how productive teams are relative to the other 119 teams in the NCAA football bowl subdivision, and while this is a large event, it gets mingled with the other 446 plays by Wisconsin's defense and the other 491 plays by Michigan State's offense.
My question is: should one play effect an entire season? I am not sure that this is all that efficient in terms of measuring productivity, and hence believe that this one play over-emphasizes the measure of each teams performance on the field.
For those interested - here is the play and the resulting decision - which I agree was the correct call. OK, it is fairly dramatic.