Dan Mullen and was asked about quarterback accuracy at one of his press availabilities last week. In his answer, he made a point that is worth exploring: completion percentage does not equal accuracy.
To a certain extent, completion percentage is a useful metric. There aren’t many things that give real solid guidance on whether a quarterback can make it from the college level to the pro level successfully, but completion percentage is one of them. There have been a few different attempts to determine the threshold, and an inability to hit 60% (or maybe it’s 58.5%) and the player’s final year is typically a bad sign. Being on the north side of 60 does not guarantee anything, but being below it probably means the quarterback is not accurate enough to play at the highest level.
Beyond that, large gaps in completion rates do contain real information. A 66% passer probably is more accurate than a 54% passer. He probably plays in a very different offense too. If the latter runs a triple option, throwing 11 to 13 times a game with several of those being deep play action bombs, that might even be a wholly acceptable completion percentage.
That said, the simple completion rate has some real drawbacks. One of the biggest is a lack of context for throws. An interception counts exactly the same as hitting a receiver a half inch out of bounds. A circus catch on a tip drill counts the same as a perfect back shoulder throw.
There have been attempts to fix some of that problem. Pro Football Focus came up with an adjusted completion percentage, which is the normal stat but counting drops as catches and throwing out balls batted at the line, passes that happened while the quarterback was being hit, and throwaways. It also tosses out spikes, which aren’t counted in a quarterback’s stats at the college level.
Spikes aren’t the only difference in the official stats, which is why you don’t see a similar adjusted completion percentage used in college much, if at all. NFL stats have gotten more detailed over the years, but college ones are stuck in the past.
There is no official NCAA stat keeping on drops, batted balls, and passers being hit while throwing. It would be possible to come up with adjusted completion percentage for a college passer, but it’d require going through video of every incompletion.
Even that stat isn’t foolproof, since some of these can be subjective. For instance, where exactly is the line between a drop and not being able to come up with a tough catch? Plus, it can be tough to tell when a quarterback is being hit while throwing versus getting the ball out just before a defender hits. I’ve gone over Feleipe Franks’s red zone interception against LSU several times, and even going frame-by-frame I go back and forth on whether the defender bearing down on him hits him before the release of the ball or not.
This gets back to the point Mullen was making last week. He said not only that completion percentage doesn’t measure accuracy, but that accuracy is about ball placement.
Ball placement is surely a hard thing to quantify. There are some places like Pro Football Focus that pour over the film and try to assess every throw for accuracy in a way that accounts for ball placement. A quarterback can get a bad score on completions, even touchdowns, on a throw where the receiver bails out a bad toss from the quarterback.
I did something along these lines for Franks after the Florida State game. I used a scale from -2 to +2 for not just passes but any play in which he had to make a choice of some kind in order to rate his decision making. He was great in that contest, and I only gave him a single -2 rating. It came on a play where an inaccurate throw saved him from an interception after he stared down C’yontai Lewis on a short out.
In Franks’s completion percentage, that bad throw counts the same as a throwaway on the following drive or an accurate deep pass on the drive after that hit Tyrie Cleveland in the hands but he couldn’t quite come up with it. I’m sure when reviewing the game, Mullen had a much more nuanced view to share with his signal caller. It probably had nothing to do with ball placement, as I would bet it was more about reading defenders and not even attempting that pass in the first place.
As you watch the Gators this fall, look for how well Franks does with ball placement. Is he putting it between the receivers’ numbers? Is he hitting guys in stride or are they having to adjust to his passes? How often does he throw someone open or put it where a target can get it but a defender can’t? If you can keep track of all that, you’ll be doing better than the folks who just look at which side of 60% completion percentage he ends up on.