Injuries per snap is a consistent measure across

all of football. I don't know what to tell you. I'll be frank and tell you that you're reading the statistics the wrong way - if the average player has a 0.1% chance of being injured on every snap, then his chance of not being injured in a game is (1-0.001)^#SNAPS. His chance of not being injured over the course of the season are Y% given X%. Why is the aggregate injury rate going upwards all season? Furthermore, the risk of injury is the same snap to snap and game to game, but the cumulative injury risk is higher. If I roll a dice 100 times, I don't expect it to come up on 6 50% of the time. College football isn't different than pro in this matter, and the only real difference is that the #snaps per game doesn't vary much in the pros because of small rosters (hence almost no variation game to game).

This is all disregarding Kelly or the particulars of ND's coaching. This is college-football wide. It is obvious that the starting 22 gets fewer snaps in blow outs than it does in semi-struggles against decent Tier 2 opponents (take Northwestern last year as a good example).

Navy is something of a special exception, but even then, we haven't had a lot of injuries against them in at least five years or so.