Regarding the Average Usage row near the bottom of the table, I am interested in how much the team used a player when that player actually played. For example, as to Keke, I ignored the five games he missed. Strictly speaking, he played 393 snaps out of a possible 1079, which is 36.4% of possible snaps. However, I have listed him at 50.5% because that is how much he played when he was healthy enough to play at all. That number is skewed a little low because he suffered a concussion mid-game in week 8 and played just 15% of the possible snaps in that game, but he almost certainly would have played more snaps but for the injury.