QBs and completion percentages: Frequently Asked Questions

The response to last week's piece about quarterbacks, adjusted completion percentage, etc., was a bit overwhelming, both because of the volume of responses (comments, Twitter responses, emails) and the fact that I was traveling when the responses rolled in. So I figured I'd just compile responses into one piece. Seemed easier.

Below are questions I received from any number of sources.

Alabama has a surprising number of short passes, do you have any sense how many of those are checkdowns/screens vs. designed short passes?

We don't have a perfect way to answer this question ("checkdown" isn't an option in the dropdown, though perhaps it should be), but here are some details regarding the 122 A.J. McCarron passes that traveled fewer than five yards:

• 79 came out of a shotgun formation (64.8%), 35 were from under center (28.7%), and seven were from the pistol (5.7%).
• 25 targeted Eddie Lacey (20.5%), 19 targeted Amari Cooper (15.6%), 14 targeted Christion Jones (11.5%), 14 targeted Michael Williams (11.5%), 11 targeted Kevin Norwood (9.0%), 10 targeted T.J. Yeldon (8.2%), nine targeted Cyrus Jones (7.4%), and five targeted Kenny Bell (4.1%). That's balance.
• 18 were halfback screens (14.8%), 14 were receiver screens (11.5%), 12 were bootlegs/rollouts (9.8%), eight were bubble screens (6.6%), and two were screen passes (1.6%).
• Three were throwaways (2.5%).
• Two were tipped at the line (1.6%).

Yeah, Johnny Manziel's completion percentages in our sample were average or below average on all throws beyond 20 yards (30% from 20-24 yards, 46% from 25-29, 17% from 30-34, 40% from 35-39, 0% from 40+). If he can borrow Geno Smith's deep ball, his numbers would be unfathomable.

Can't dispute the truth. I withdraw any comments I made about Collin Klein being underrated.

Well, I'd say "ruins" is a bit overstated, yes? The comparison would be better with more stats, but we did what we could with two interns. I'm taking volunteers for next year; we could do amazing things with broader data.

As a Husker fan, should I be more encouraged or less encouraged by this data...

Yes.

(Seriously, this reaffirms how well Tim Beck did at play-calling. With a quarterback who is woefully inaccurate after 10 yards, he crafted one of the most effective standard downs offenses in the country around easy, short passes and a lovely running game. Opponents couldn't completely gang up on the run because Martinez would complete an easy 8-yarder on first down.)

Yeah, you don't need stats to back up your impressions of that one.

Tevin Washington. Where the crap is he?

Did he even throw a pass last year?

Chelf was Ok St’s THIRD option at QB. That system has to be the definition of plug-and-play offense.

Right???

Here's a breakout:

• <5 yards: 72% from under center, 72% from shotgun
• 5-14 yards: 63% from under center, 59% from shotgun
• 15-24 yards: 63% from under center, 50% from shotgun (tiny sample)
• 25+ yards: 47% from under center, 0% from shotgun (tiny sample)

Pretty inconclusive overall. Barkley was just slightly worse from the shotgun on passes under 15 yards, and we didn't have enough charted shotgun passes to get a sample of passes over 15 yards.

Though Ryan Nassib and Tyler Bray also looked pretty awful in my sample, Barkley is probably the quarterback whose stats and perceptions were the most disparate (at least in a negative sense). I actually struggle with this a bit because I've always enjoyed watching Barkley. For a golden-boy, glamour quarterback, he has always seemed to have a really good head on his shoulders, he's great in interviews, his mechanics are pretty and old-school, and he really seems to want to be a leader.

And for what it's worth, his numbers on the longer throws were somewhere between average and downright strong. But a) you have to make shorter throws, too (otherwise Jeff George would be in the NFL Hall of Fame), and b) while it seems like he's a pretty good leader, it's really hard for me to ignore that this year's USC team appeared completely lacking in leadership in 2012. Obviously Barkley didn't play defense, so we cannot pin the Trojans' occasional struggles on that side of the ball on him. But even the offense spent good chunks of the season underachieving, just like it did in 2011 and in every season in which he was the starting quarterback. The ceiling for Barkley, with the composure and the longer throws, is probably pretty high. But the floor is low. Really low. And picking him too high would place expectations on him that I don't think he'd have much of a chance to meet. Maybe I'm wrong, but the numbers backed up what I perceived when I watched him play in 2012. With the best receiving corps on the West Coast and a (theoretical) offensive wiz calling plays, Barkley and the USC offense were only occasionally strong. They finished the year ranked 33rd in Off. F/+, and while part of that can be ascribed to the egg the Trojans laid in the bowl game without Barkley ... only part of it can. South Carolina ranked 28th, for goodness' sakes, and they had two quarterbacks and spent half the year without their star running back.

What games did you chart?

I thought I had shared that, but I had not. Here's the list of 109 games. I recommend a Ctrl-F search to find your team's games.

Stats, LLC, does great things with it, and I'm really hoping that it catches on at this level. It just provides such wonderful depth and context. With SBN's help, I will be pushing to expand our charting reach (with help from volunteers) this coming fall. And yeah, if it were to somehow catch on at the high school level ... man ... you want to talk about improved prospect scouting? It would be fantastic and provide so much more context than watching studs run drills or embarrass overwhelmed opponents.

This is some really exciting stuff- Bill, is there any plan to crowdsource/share/open-source all of this data?

That's certainly the goal at some point. Stay tuned.