Some of the phrases you hear often in player evaluations, scouting profiles or camp reports are things like
- "He seems to be around the ball a lot."
- "He has a nose for the ball."
- "He has a knack for making plays."
- "He just jumps out at you on tape."
I've often wondered what exactly these phrases mean, because I've always thought that if somebody makes a lot of plays or is around the ball a lot, there should be a way to measure that. While I was doing research on a recent four-part series of posts on the Cowboys' pass rush, I came over this quote from defensive line coach Brian Baker who said that BTB-favorite Sean Lissemore was one of the team's more productive players, despite his limited snaps:
"Actually, he was our most productive player, per play last year on defense," defensive line coach Brian Baker said. "When I look back, he's probably a guy I should've played a little more."
Okay, so the Cowboys are measuring productivity on a per play level. "Fascinating", as Spock would say. Then on Thursday, I ran across this quote by secondary coach Jerome Henderson on safety Barry Church:
"When you watch him on tape, he just jumps out at you," Henderson said of the 6-2, 220-pound Church. "You say, 'Wow, we've got to find ways to play this guy more.' He just shows up making plays. All the time. Always around the ball. You look at that, you say you want to give him more opportunities. Now he's getting his opportunities. We'll see how it plays out."
"Interesting". So I figured, if the Cowboys can measure a guy's productivity, let me see if I can come up with a model that shows both Sean Lissemore and Barry Church as the most productive players on the roster. Find out after the break if I was successful.
Before we start, some clarification as to what I mean by productivity. Basically, I'm looking for "guys who make plays", or how often a defender impacts the game positively. This means that I won't be making a distinction between the quality of plays. I'm simply going to count the plays on which the defender made a positive defensive contribution, and divide it by the number of his defensive snaps. This should give me a close approximation of what Coach Baker described and what I'll call Play-by-Play Productivity.
To avoid any confusion, these are the positive defensive plays I'll be looking at.
Sacks: A QB is tackled for a loss or no gain before he can throw a pass or while in the pocket. I'm using PFF numbers here, not the official NFL stats, because what the NFL calls half a sack is counted as a full sack by PFF. As a result, PFF sack totals can differ slightly from the official numbers.
QB Hits: A hit is when a QB is knocked down but not sacked, again as counted by PFF.
Stops: Stops are plays on which the defense holds the offense to less than 40% of necessary yardage on first down, less than 60% on second down and less than 100% on third and fourth downs. Again, this number is taken from PFF, but I've separated it into run stops and coverage stops, and have deducted sacks from the totals, as technically each sack is also a stop.
Interceptions and Passes Defended: Official NFL stats, but INTs are deducted from the PD Total to avoid double counting
Forced Fumbles and Fumble Recoveries: Official NFL stats.
So every time a Cowboys defender recorded one of these stats, I counted it as positive play with which I calculated his Play-by-Play Productivity. Here are the results for the defensive line:
Player | Total Snaps | Pass Rush | Stops | Coverage | Others | Play-by-Play Productivity | ||||
Sacks | Hits | Run Stops | Coverage Stops | INT | PD | FF | REC | |||
Sean Lissemore | 283 | 2 | 3 | 14 | 0 | 0 | 0 | 0 | 0 | 6.7% |
Kenyon Coleman | 425 | 1 | 1 | 21 | 2 | 0 | 3 | 0 | 0 | 6.6% |
Jason Hatcher | 428 | 5 | 6 | 11 | 0 | 1 | 1 | 1 | 0 | 5.8% |
Jay Ratliff | 750 | 2 | 6 | 20 | 2 | 0 | 3 | 0 | 1 | 4.5% |
Marcus R. Spears | 400 | 1 | 1 | 10 | 2 | 0 | 2 | 0 | 0 | 4.0% |
Josh Brent | 137 | 0 | 0 | 3 | 0 | 0 | 0 | 0 | 0 | 2.2% |
Using this approach, Sean Lissemore tops the list of Cowboys defensive linemen. Lissemore recorded 19 positive defensive plays on 283 defensive snaps, to make him not only the most productive lineman in 2012, but also the most productive player on a per-play basis on the Cowboys defense, narrowly beating out Sean Lee as we'll see later.
The surprise on this list is undoubtedly Kenyon Coleman. His 21 run stops place him second on this list, and those 21 stops also give Coleman the third highest stop rate on running snaps of all 3-4 defensive ends. This makes Coleman the classic 'fencepost' DE. He's nowhere near Jason Hatcher as a pass rusher, but he was very productive against the run in 2011. Ratliff and Brent are closer to the bottom of the list, a clear sign that there simply aren't that many plays to be made in the middle. Spears' numbers must be considered a little disappointing.
Next up, linebackers:
Player | Total Snaps | Pass Rush | Stops | Coverage | Others | Play-by-Play Productivity | ||||
Sacks | Hits | Run Stops | Coverage Stops | INT | PD | FF | REC | |||
Sean Lee | 868 | 0 | 3 | 30 | 16 | 4 | 3 | 0 | 2 | 6.7% |
Victor Butler | 233 | 3 | 4 | 6 | 1 | 0 | 1 | 0 | 0 | 6.4% |
DeMarcus Ware | 913 | 20 | 8 | 15 | 7 | 0 | 2 | 2 | 1 | 6.0% |
Anthony Spencer | 939 | 6 | 9 | 27 | 6 | 0 | 1 | 4 | 0 | 5.6% |
Bradie James | 415 | 0 | 1 | 12 | 5 | 0 | 1 | 1 | 1 | 5.1% |
Keith Brooking | 408 | 0 | 1 | 13 | 2 | 0 | 2 | 0 | 0 | 4.4% |
Nobody touches Sean Lee in terms of overall positive defensive plays. His 58 positive plays lead DeMarcus Ware (55) and Spencer (53), and because Lee played fewer snaps, he has a better overall play-by-play productivity.
It's a little unexpected to see Victor Butler with a higher Productivity percentage than both Ware and Spencer, but that's how he played last year, and if linebacker coach Eberflus is looking at similar stats, he'll try to find more opportunities for Butler this season.
And now for the secondary:
Player | Total Snaps | Pass Rush | Stops | Coverage | Others | Play-by-Play Productivity | ||||
Sacks | Hits | Run Stops | Coverage Stops | INT | PD | FF | REC | |||
Barry Church | 172 | 0 | 2 | 4 | 2 | 0 | 0 | 0 | 0 | 4.7% |
Orlando Scandrick | 680 | 2 | 2 | 5 | 9 | 1 | 4 | 0 | 0 | 3.4% |
Frank Walker | 331 | 0 | 1 | 1 | 3 | 1 | 2 | 0 | 0 | 2.4% |
Terence Newman | 829 | 0 | 0 | 4 | 5 | 4 | 4 | 0 | 1 | 2.2% |
Mike Jenkins | 607 | 0 | 0 | 0 | 5 | 1 | 7 | 0 | 0 | 2.1% |
Alan Ball | 499 | 0 | 1 | 1 | 3 | 1 | 5 | 0 | 0 | 2.2% |
Gerald Sensabaugh | 1,004 | 0 | 0 | 6 | 4 | 2 | 1 | 2 | 2 | 1.7% |
Abram Elam | 1,050 | 0 | 0 | 10 | 6 | 0 | 0 | 1 | 1 | 1.7% |
As expected, Barry Church leads the list of defensive backs. As a rule, the DBs play a little further away from the ball, so it's no surprise to see lower overall productivity numbers. Still, it's disappointing to see the two starting safeties at the bottom of this list. In 2011, Sensabaugh and Elam were not exactly what you would call playmaking safeties, and perhaps this entire exercise does the safeties (and others) a disservice by not including tackles - Sensabaugh and Elam ranked no. 2 and 3 behind Sean Lee - but it is what it is.
And yes, Church played a lot closer to the line of scrimmage than Sensabaugh and Elam, making him perhaps more of a linebacker than a safety, but he still came through on limited opportunities.
Overall though, this is an interesting exercise to somewhat mirror what the coaches are looking at as they evaluate their players. We know that the Cowboys' internal stats that they take from the Coaches Film differ from the official NFL numbers, but the fact that we've been able to model Lissemore to the top of the list and Church to the top of the DB list (with room to spare) suggests that we're close enough.
And it's comforting to see that each positional ranking is topped by a young guy with a lot of upside left to his game. Young players can be mercurial. They can improve suddenly. They can improve in leaps and bounds. Lissemore, Butler, Church will challenge hard for starting spots this year, and while Lee is already a starter, we'll also look to him to elevate his game further.