### Pricing the 4-3 Defensive End

One of the things that I like to do here is come up with different ways to value positional players. Today on Twitter Joel Corry mentioned Jared Allen and how difficult it will be to get a \$10 million a year contract because of his age. I pretty much think its impossible but Joel brought up a great point when I compared him to John Abraham in that Allen plays well over 90% of the snaps while Abraham is a situational player. So of course that got me to thinking about putting something together on the 43 Defensive End position. As always the raw data comes from Pro Football Focus but the actual analytical work is my own.

Defensive Ends primarily have two responsibilities- rushing the passer and stopping the run. Some drop off into coverage but it’s a small part of the positional responsibilities so Ill avoid that here. I thought maybe an interesting way to look at things would be strictly from the point of view of increased failures on a play. How do we define a failure?

People who read my work know I focus on something called Pass Russ Performance (PRP) where I determine the expected result of a pass play if there was no pressure vs that of pressure. That difference for each player equals the added benefit of a player. Using Allen as an example he rushed the QB on 638 snaps and registered 64 pressures and 11 sacks. If the QB was not pressured he would have completed 413.8 passes.  Factoring in the effect of Allen’s pressures and 11 sacks he only completes 394.7. So he adds 19.06 negative plays to the Vikings defense via his pass rush.

To value run defense I wanted to look at PFF’s category of stops since a stop would constitute a “loss” for the offense similar to an incompletion. To value this I wanted to compare the player to the average performance of DE’s in the NFL.  The average DE generates a stop 6.1% of the time he is on the field for a run possession. So to grade this we calculate the number of times a player would simply be stopped by an average player and subtract that from the number of stops the player was actually credited. Using Allen as an example we can calculate that a player should fail 21.5 times based on Allen’s 349 run snaps.  Allen was credited with 28 stops, so we can say he was responsible for 6.5 additional failures on his run play.

We can use those figures to determine a number of benefits about a player. Adding them together we can say that the Vikings are paying Allen for 25.6 negative plays. You can look at it as negative plays as a percentage of on field opportunities. You can also look at it from the perspective of percent of increased failures. There can be a big difference between the 3 numbers. Allen ranks 2nd just based on total number of negative plays but falls to 13th and 14th among players with at least 200 snaps in the other categories, putting his totals more as a byproduct of snaps played than overall impact.

I always find this a hard item to reconcile. A player in on every down is hurt by the latter metrics because we are not taking situations into account, such as that of a screen pass having no chance of leading to a pressure.  If a player comes in only on 3rd and 6 or more he can just tee off on a QB. Those same players often suffer when they are forced to play more snaps. So it can be hard to put a numeric value on the player.

One method I thought of using was to use percentage increase in failed plays and then adjust downward based on downs played. In other words if a player is worth \$10 million but only plays 300 snaps we need to reduce that salary by what it costs for an average players that plays somewhere between 600 and 700 snaps to make up for the snaps. After some thought that seemed to be a lot of extra work so instead I looked at the median snaps for the top 32 players (807) and adjusted the players who had above that level downward based on their negative plays per snap. While not perfect I felt that this provided a fair estimate of what the player would do in a normal role with another team.

To  revalue the position I wanted to look at the performance of the top 50 players and determine how much above or below the average a player  performed and then use that to determine the players salary. The average performance worked out to be 12.23 negative plays and the average salary was \$4.77 million.

Not surprisingly the results indicate a gross overpricing of the market, which is the result of a combination of free agency and likely the overvaluing of the sack statistic. Cameron Wake, who graded out as the best player, should be the top of the market at just over \$10 million a year. Based on the current marketplace there are 7 players who make more than that. Greg Hardy of the Panthers and Jason Pierre Paul of the Giants ranked 2nd and 3rd, which shows the importance Pierre-Paul has to that team and how devastating his injury can be to New York. Allen ranked 5th and should be worth around \$8.2 million a year.

The biggest upside players would be Brandon Graham of the Eagles and William Hayes of the Rams. Neither played 400 snaps but both still racked up enough negative plays to rank in the top 11. Their performance per play was off the charts with over 20% increases in failures. Top of the market is under 14% for a full time guy.  Other high upside players would be Junior Gallette and Austen Lane.

The most overpaid is clearly Mario Williams at \$16 million. Williams only ranked 19th last season and was worth barely over \$5 million. The signing of Williams, which was way outside of any logical parameters even when signed, shows the problems with many philosophies in pricing free agents. In fact the 7 players who are paid in the double digits in APY only ranked 19, 14, 10, 5, 9, 21, and 17. Six of the top eight players are on rookie contracts and with some lower cost options like Hayes, Ron Ninkovich, and Kroy Biermann in the top 16 it should signal something to NFL front offices.

If your options are Williams at \$16 million or drafting a rookie you should be drafting a rookie. There is more upside and far less cost involved. You also need to set positional allocations and decide how best to fill those voids. The Bills spent nearly \$21 million on two free agents and they combined for only an additional 15.7 negative plays. The Giants got 34 on around 1/3 of the cost with a rookie and a timely contract with Osi Umenyiora. The Panthers got around 41 with 1 high cost player and a rookie. Their spending was around \$13 million. So if you must sign a high priced player or re-sign one of your own he must be paired with a low cost player. The dual high priced approach is doomed to failure.

More teams seem to realizing this as teams paid record low dollars for free agents and the older veterans had trouble even finding jobs. John Abraham, who ranked 18th last season can’t even find a job. Getting back to the original question surrounding Allen it would be hard pressed for him to reach the \$10 million mark next year. Even as arguably the 2nd best player last year he would not be worth that kind of money. My guess is the market drops further in the near future. Both Williams and Julius Peppers will likely see their contracts vanish by 2014 or 2015. Chris Long can be renegotiated at any time as he has little protection in his contract. Michael Johnson is a free agent next season as is Allen. That leaves Charles Johnson as the lone player who will be left earning over \$10 million a year. We will need to wait and see how the market turns in the future but it should be closer to this chart than the current chart as it exists.

RankNameTeamFailuresTotal Snaps% Increase
in Failures
% Fails/SnapNew APY
1Cameron WakeDolphins26.8984013.1%3.2%\$10,077,999
2Greg HardyPanthers23.6070713.8%3.3%\$9,203,097
3Jason Pierre-PaulGiants23.0283711.6%2.8%\$8,657,166
4Lamarr HoustonRaiders21.6880112.1%2.7%\$8,456,413
5Jared AllenVikings25.6198710.4%2.6%\$8,168,386
6Carlos DunlapBengals19.9454914.0%3.6%\$7,775,941
7Brandon GrahamEagles19.7039923.5%4.9%\$7,683,581
8Derrick MorganTitans20.5287010.2%2.4%\$7,422,457
9Trent ColeEagles17.6069111.1%2.5%\$6,864,512
10Charles JohnsonPanthers17.507849.2%2.2%\$6,826,909
11William HayesRams17.3535320.5%4.9%\$6,767,174
12Kroy BiermannFalcons16.2760011.2%2.7%\$6,345,278
13Cameron JordanSaints19.079768.5%2.0%\$6,150,484
14Julius PeppersBears15.717468.4%2.1%\$6,129,148
15Michael BennettBuccaneers16.838717.7%1.9%\$6,082,862
16Rob NinkovichPatriots14.517108.6%2.0%\$5,659,520
17Michael D. JohnsonBengals14.447967.4%1.8%\$5,630,695
18John AbrahamFalcons14.266748.8%2.1%\$5,563,636
19Mario WilliamsBills15.048807.6%1.7%\$5,381,078
20Juqua ParkerBrowns13.7848210.2%2.9%\$5,375,992
21Chris LongRams14.158456.8%1.7%\$5,271,106
22Israel IdonijeBears13.0448210.6%2.7%\$5,087,483
23Everson GriffenVikings12.865408.7%2.4%\$5,016,584
24Elvis DumervilBroncos12.088256.3%1.5%\$4,608,474
25Osi UmenyioraGiants10.996077.0%1.8%\$4,287,123
26Junior GaletteSaints10.8527114.3%4.0%\$4,230,377
27Jason BabinEagles10.6340710.5%2.6%\$4,147,199
28Bruce IrvinSeahawks10.594038.9%2.6%\$4,132,221
29Justin TuckGiants10.436097.1%1.7%\$4,068,802
30Jabaal SheardBrowns11.029035.3%1.2%\$3,842,367
31Brian RobisonVikings9.598134.8%1.2%\$3,712,745
32Austen LaneJaguars8.9335513.4%2.5%\$3,481,578
33Chris ClemonsSeahawks8.207804.3%1.1%\$3,196,767
34Olivier VernonDolphins8.023578.0%2.2%\$3,126,219
35Chandler JonesPatriots7.756834.7%1.1%\$3,021,599
36Derek WolfeBroncos7.608323.9%0.9%\$2,874,585
37Robert QuinnRams7.327993.7%0.9%\$2,856,697
38Frostee RuckerBrowns7.315396.8%1.4%\$2,850,665
39Corey WoottonBears7.205525.5%1.3%\$2,808,349
40Jermaine CunninghamPatriots7.014066.5%1.7%\$2,734,306
41Lawrence JacksonLions6.953677.9%1.9%\$2,711,119
42Jared OdrickDolphins6.968673.4%0.8%\$2,526,017
43Robert AyersBroncos6.292888.1%2.2%\$2,455,106
44Will SmithSaints6.969283.2%0.8%\$2,361,275
45Da'Quan BowersBuccaneers5.872468.5%2.4%\$2,288,554
46Cliff AvrilLions5.806573.8%0.9%\$2,261,095
47Wallace GilberryBengals5.652987.9%1.9%\$2,202,624
48Daniel Te'o-NesheimBuccaneers5.416703.4%0.8%\$2,109,439
49Martez WilsonSaints5.171639.9%3.2%\$2,017,658
50Shea McClellinBears5.003355.3%1.5%\$1,949,363