I Introduce To You The P.E.P.A.T Project.
P.E.P.A.T stands for ‘Power-eligible Playcalling and Tendencies.’
This initiative was borne of a desire to determine which Detroit Lions running back was most effective in short yardage situations in 2016. After a short amount of time poking around the internet, I stumbled across the Football Outsiders’ offensive line stats page.
Here is the first full article in the series, take a look after reading this introduction.
The good folks over there track a statistic that they call ‘power success,’ which is quite simply, a team’s success in converting power situations. They define power situations thusly:
- 3rd/4th & 2≥ from anywhere on the field, and
- Any down that takes place on the opponent’s two-yard line or closer.
So, I began scrubbing all sixteen of Detroit’s game logs, and found 55 such instances.
Initially, I wanted only to see which running backs converted on runs in these situations, and therefore which running back was the most useful in such situations. It dawned on me that this would fail to account for several variables. For example, I found that Zach Zenner converted only one of six possible times, but deeming Zenner as having less value in these situations because of this would have been erroneous. Why?
For argument’s sake, what if Zenner took just as many snaps on pass plays, and he was targeted six times for six first down catches? That would mean he converted 7/12 times (58.3%) in such situations, rather than 1/6 times (16.67%). That’s a big difference.
So down the rabbit hole, I went. I began charting personnel groupings and alignments on all power-eligible plays, not just the runs. Then I started to diagram the plays themselves and identify frequently-used concepts and formations. I noted the direction of the runs; whether they were to the left, right, or middle; to the strong side or the weakside. I charted the margin of the score for each play call.
This article will be the first in a series that will breakdown these findings. I’ll talk about things like their run game, the passing concepts they like to utilize, their tendencies based on score, and much, much more.
But, first – a primer. The Lions called 22 run plays in such situations, or 40% of the time. Those plays resulted in a first down or a touchdown 50% of the time. They dialed up a pass the other 33 plays, or 60% of the time, and converted 16 of those, which would have been good for a 48.5% rate had an Eric Ebron touchdown reception not been called back on a phantom offensive pass interference in Week two.
Per TeamRankings, the Lions called a pass on 64.6% of all their plays in 2016. And according to Sharp Football, they ran 1031 plays. This means that the Lions called a pass 667 times and a run 364 times. Arithmetic isn’t my strong suit, but my math indicates that the Lions called a run play 35.1% of the time outside of all power situations. That figure could include quarterback scrambles which means the rate of called runs could be even lower.
Therefore, the Lions were at least 4.9% more likely to attempt a rush in power situations than under any other circumstance.
However, being that Stafford only attempted a pass 594 times in 2016, and was the only Detroit quarterback to attempt one, I believe that this data includes all snaps. This would mean that plays negated or offset by penalties would be included.
If that makes your brain hurt, I’ve included this convenient table for your reading pleasure below.
| Situation | Pass% | Run% |
| OVERALL | 64.6% (667) | 35.4% (364) |
| ALL NON-POWER-ELIGIBLE | 64.9% (634) | 35.1% (342) |
| ALL POWER-ELIGIBLE | 60% (33) | 22 (40%) |
Speaking of tables, the examination of personnel groupings indicated the following backfield snap shares:
| Player | Snap count |
| Theo Riddick | 24 (43.6%) |
| Zach Zenner | 17 (30.9%) |
| Dwayne Washington | 12 (27.3%) |
| Justin Forsett | 2 (3.6%) |
| Ameer Abdullah (missed 14 games) | 1 (1.8%) |
Note that the percentage will total more than 100%, because of two-back looks.
This is just a small sampling of the data that will be released and analyzed over the course of this series. The subsequent installment will follow shortly and I will be conducting a Twitter poll to determine what component to write about next. I’ll also be taking to Twitter to do GIF threads as a companion resource – you can find me @btrossler. Stay tuned for this and more great content to come.
