Writers of Pro Football Prospectus 2008

22 May 2014

2013 Offensive Personnel Analysis

by Aaron Schatz

A couple of days ago, we got a question over Twitter regarding personnel groupings during the 2013 season. Usually that information shows up in the book, in the "Performance Based on Personnel Groups" tables in each team chapter. But there's no reason to hold it all the way until then. We'll save just how well specific teams did in certain personnel groups for Football Outsiders Almanac 2014 and for more offseason articles like this one from last summer, but today I'm going to share some interesting things you see going through personnel grouping data from 2013.

Before we get into things, a quick note: this is personnel data, not formation data. When C.J. Spiller goes out wide, he's still a running back, and when Cordarrelle Patterson lines up in the backfield, he still counts as a wide receiver. We're using the standard numerical system where the first number is the number of backs and the second number is the number of tight ends. We count formations with six or more offensive linemen separately, rather than counting the offensive linemen as tight ends. A formation with two backs, one tight end, one wideout, and six offensive linemen is marked as 621 and not 22. No "wildcat" formations with a non-quarterback taking a shotgun snap are considered in this analysis.

We've noted it in the past, but the trend accelerated over the past two years: three-wide is the default formation in the NFL right now, and the second tight end or (hey, remember this guy) the fullback are the part-time players. We've written before that the majority of plays in the NFL have three or more wide receivers, but for last year you can even just delete the words "or more." Teams came out in 11 personnel on a majority of plays: 51.2 percent. They came out with three or more wide receivers on 58.8 percent of plays. We've tracked personnel data separate from formations for four seasons now, and look at the way 11 personnel has grown over those four seasons:

Eleven is One Louder: Offensive Personnel Groupings 2010-2013
Personnel 2010 2011 2012 2013 x DVOA
in 2013
11 39.5% 40.4% 45.7% 51.2% x 8.1%
12 19.5% 20.9% 22.0% 20.6% x 0.4%
21 19.5% 16.8% 15.2% 12.6% x -1.0%
22 7.2% 7.0% 5.5% 4.4% x -4.4%
13 1.8% 2.3% 2.0% 2.3% x -4.3%
20 2.9% 2.3% 1.2% 1.8% x -1.4%
10 3.0% 3.2% 2.1% 1.3% x -7.1%
6+ OL 2.8% 3.3% 3.4% 3.5% x 7.0%

"6+ OL" adds together all personnel groups with at least six offensive linemen; the most common are 621 and 611, each with about 1.0 percent of offensive plays last season.

Use of 11 personnel skyrocketed over the past couple seasons, and one reason it has skyrocketed is that it works. Look at how more efficient offenses were with 11 personnel last year. This isn't a one-year fluke. In 2012, the leaguewide DVOA with 11 personnel was also 8.1%. In 2011, it was 5.5%, and in 2010, it was 9.1%. Right now, the most efficient way to play offense in the NFL is to put three wide receivers, one running back, and one tight end on the field with your quarterback in shotgun for a majority of snaps. Not all of them, you have to switch it up of course, but most of them.

(It also paid to get those six-lineman sets on the field last year, but unlike with 11 personnel, that may not be a trend. The year before, the combined DVOA of all personnel groupings with six or more linemen was -7.2%.)

11 personnel was the most common personnel group for 29 of the 32 NFL offenses last year. The exceptions:

  • Cincinnati had 12 personnel on 43 percent of plays, 11 personnel on 40 percent.
  • Oakland had 21 personnel on 36 percent of plays, 11 personnel on 34 percent. Hey, it's the Marcel Reece Effect!
  • San Francisco had 21 personnel on 26 percent of plays, 22 personnel on another 26 percent of plays, and 11 personnel on 21 percent of plays.

The 49ers are really in their own world as far as personnel groups are concerened. They were the only offense to use 11 personnel less than 30 percent of the time. Add in 12 personnel, which they used on 15 percent of plays, and for the second straight season they were the only offense to have four different personnel groups that were used on at least 15 percent of all plays. The 49ers were also the only team to use seven offensive linemen on a regular basis, with 721 personnel on 3.4 percent of plays.

The team most devoted to 11 personnel? Why, that would be Baltimore, believe it or not. So much for Vonta Leach as the best blocking fullback in the NFL. The Ravens used 11 personnel on 75 percent of all their offensive plays. That's a big shift from the year before, when Baltimore used 11 personnel on 43 percent of plays, slightly less than the NFL average. Two other teams used 11 personnel over 70 percent of the time in 2013, Philadelphia and Denver.

Thanks to Dennis Pitta being out for most of the season, the Ravens were the least likely team to put 12 personnel on the field, using it on just 5.9 percent of plays. Oakland and Tampa Bay were the other teams below 10 percent. (That's definitely going to change in Tampa this year.) Cincinnati was the only team to use 12 personnel over 40 percent of the time, but four teams used it over 30 percent of the time: St. Louis, Cleveland, Dallas, and San Diego. The Rams are this high in part because we've categorized Cory Harkey as a tight end rather than a fullback. The Bengals also like to use tight ends as fullbacks but rotated their tight ends through that spot -- it was a lot of Alex Smith, but Jermaine Gresham and Tyler Eifert would also show up there, plus Orson Charles on the rare occasion he got offensive snaps.

At this point, "conventional" 21 personnel should be called something else, like "old school personnel" or "70s personnel" or maybe "John Facenda personnel." The Oakland Reeces used it the most, as noted above, and the 49ers used it a lot. Three other teams used it on more than 20 percent of plays: Houston, Tampa Bay, and New Orleans. The Patriots used it on 19 percent of plays after not using it at all the year before, which is a great stat to quote the day before Brown Reunion Weekend. As long as Develin is playing this much, both me and Chris Berman will be completely unsufferable. Although the Patriots were tossing 21 personnel out there a lot in 2013, three teams used it on fewer than five plays. As noted above, Cincinnati and St. Louis don't get marked with 21 personnel because of how we marked certain tight end/fullback hybrids like Cory Harkey and Alex Smith. The third team is Denver, because Peyton Manning doesn't need your stinkin' fullbacks.

Every year it seems like one team uses six linemen much more than any other team, and this year that team was Chicago, which used a sixth lineman (usually Eben Britton) on 15.9 percent of plays. Six other teams used six or more linemen more than five percent of the time, in descending order of frequency: Indianapolis, Pittsburgh, New Orleans, San Francisco, Oakland, and Tampa Bay.

A few other notes:

  • San Francisco used 22 personnel more than any other team; Minnesota at 13 percent was the only other offense to use it more than ten percent of the time.
  • Arizona used 10 personnel on 11 percent of plays. New England and Seattle were the only other teams above five percent, and they were just barely above. Kansas City was the other team other than Arizona that went without a tight end more than ten percent of the time, but they were often using 20 personnel instead of 10.
  • You'll often hear a television commentator talk about a team going "five-wide," but it rarely happens with five actual wide receivers. Minnesota was the only team to run more than five plays with five wideouts, as long as we count Joe Webb as a wideout. Only five of the 23 plays had Webb or Patterson in the backfield.
  • The Jets used a formation without a standard quarterback on 3.8 percent of plays; no other team was above two percent.

Posted by: Aaron Schatz on 22 May 2014

12 comments, Last at 29 May 2014, 6:00pm by samuellogue

Comments

1
by theslothook :: Thu, 05/22/2014 - 3:11pm

I wonder if you can credit PM for this revolution. After all, his hatred for all things fullback goes back a while.

7
by Roch Bear :: Thu, 05/22/2014 - 5:17pm

Depending upon how the arithmetic was done, maybe a lot. It would be very interesting to see how DEN did in 11 vs 21 groupings. It is less interesting if the great offenses in the league tend to use 11 groupings and the bad ones 21 (for example).

I'd like to see the calculation done with each team's groupings then do an unweighted average across the 32 teams. Is that what was done?

2
by Perfundle :: Thu, 05/22/2014 - 4:19pm

What was the DVOA breakdown on SF's 22, 21 and 11 formations? Their run game really did not look good for large swathes of the season, and indeed their top three running backs each had negative DVOA.

3
by Karl Cuba :: Thu, 05/22/2014 - 4:34pm

On the other hand the passing game was still effective even with only one decent wide receiver for most of the year, which suggests that defenses were more concerned with the run game and that's why it suffered.

4
by Perfundle :: Thu, 05/22/2014 - 4:48pm

Which is why I wondered about the efficiency when they were in 22 and 21 personnel. If they primarily ran from those two formations, that would help to explain why defenses were stacking the box, so perhaps they should cut down on 2-RB sets to begin with. As for the passing game, though the depth was certainly pretty horrible when Crabtree was out, and hit rock bottom when Davis was injured for a couple of games, Davis and Boldin make quite a good 1-2 combo.

5
by Karl Cuba :: Thu, 05/22/2014 - 4:51pm

Everything you expect to see you will see. They run a ton out of 22 and a lot out of 21, with predictable results. It's a predictable offense now that most teams have caught up to the trap blocking.

6
by Perfundle :: Thu, 05/22/2014 - 5:06pm

Interestingly, Hamilton and Indianapolis was quite successful at that before they got saddled with Richardson (and continued to be very good with Brown on the rare occasions when they swallowed their pride and benched Trent for a bit). They completely dominated San Francisco at their own game in week 3.

http://www.sbnation.com/nfl/2013/10/4/4800856/indianapolis-colts-2013-nf...

I simply would've expected more 12-personnel out of San Francisco if they're worried about depth at WR. It just seems to be a more flexible grouping when you have Davis.

8
by Karl Cuba :: Thu, 05/22/2014 - 5:19pm

Well the Colts have Andrew Luck, so your first priority as a defensive coordinator will be stopping him. Give the niners an unloaded front and they'll run all day, they just don't see many.

As for the 12 group, it becomes a lot less flexible when your rookie second tight end was a bit slow learning the offense (the blocking assignments for the H-back role are pretty complicated) and you still need two receivers out of that package, enter Kyle Williams!

9
by Perfundle :: Thu, 05/22/2014 - 5:44pm

I have a suspicion it was less about stopping Luck the inconsistent but potential superstar QB, and more about stopping Luck the former QB under Harbaugh. I think Harbaugh went a bit overboard making sure that his college QB didn't beat him, and Hamilton was ready for it. It was Luck's lowest passing yardage and the Colts' highest rushing yardage in the last two years. Something similar happened with New Orleans in the first game against Seattle, when they seemed hell-bent on preventing Beast Mode from happening again and Wilson tore them to shreds.

12
by samuellogue :: Thu, 05/29/2014 - 6:00pm

Going to respectfully disagree with just about all of this.

1) The passing game was without Vernon and Crabtree. It was not a very pretty snap count chart at WR for that game. Hard to keep your defense off the field when Kyle Williams is literally the second best option in the passing game.

2) 1st downs off penalties: IND 5 SF 0. IND with one flag all game. (It happens, just hope it happens to everyone equally)

3) They had just lost their starting DT the week before. When it's week 3 maybe that explains some run d troubles?

Patrick Willis left the game with an injury, and it was only a six point game at the time. That was also the day after Aldon got his DUI which was a weird situation. Also that was the week after playing at Seattle.

I don't think Hamilton was "ready for it" in the sense I think you're using. SF was obviously badly undermanned that game and IND went with the obvious play. Combined with the luck they received that explains a 27-7 game.

And lastly, when NO plays at SEA do you think scheme has anything to do with it? ;)

10
by Dr. Mooch :: Fri, 05/23/2014 - 12:31am

Arguably, the Jets used a formation without a standard quarterback on 100% of plays.

11
by FrontRunningPhinsFan :: Fri, 05/23/2014 - 1:41pm

Zing!