BGA: The Double Team Project – Part 4 (Nose Tackles)

Traditional statistics and most modern analytics measure how effective a player is without necessarily taking into account assignments or degree of difficulty. While I’ve made passing reference in my game analysis to how often certain players have been double teamed, nobody tracks this, so I’ve been keen to figure out whether the reality matches up to our expectations and my recollection from watching the film.

In part one of this series, I set out my methodology for charting every snap from the 2013 season and went through some of the things I learned from undertaking this task. In part two, I started to share the data in respect of those plays when the Jets were rushing the passer. Most recently, I started to share the data in respect of those plays when the Jets were defending the run.

After the jump, we’re going to make the first steps towards drawing some concrete conclusions about the Jets players we have investigated so far, by comparing their numbers to those of some similar players. We start with the nose tackle position.

Recapping the numbers so far

According to data compiled from re-watching the coaches film for every single defensive snap of the 2013 season, Damon Harrison was doubled 39 times and peeled 99 times in 284 run defense snaps. (For a definition of what constitutes a double or a peel, re-read part one of this series, which is linked above). Kenrick Ellis was doubled 31 times and peeled 29 times in 120 snaps.

At the time, we wondered if the fact Ellis saw double team attention on a higher percentage of plays and that he was able to occupy a double team on a higher percentage of those plays had anything to do with the fact he plays less and is therefore fresher when he enters the game, so we must ensure we compare his numbers with players in a similar role to try and test this theory.

Since it’s not their primary role, we will not be doing any more work on the pass rushing numbers for Ellis and Harrison. However, just to recap those, Harrison had 53 double teams and 84 chips in 226 pass rush snaps, while Ellis had 34 doubles and 29 chips in 90 snaps.

Methodology

In order to add context to these numbers, we need to make a comparison with other top run stopping defensive tackles around the league. Of course the best way to do this would be to chart every single snap of every game, but that would take far too long, so (unless and until such data actually does become available) I propose the following:

1. Identify relevant players to compare, either based on their role or their production (or both);
2. Chart those games for which the player’s team faced a team the Jets also faced in 2013;
3. Use the comparison across common opponents to add context to the production achieved.

In the case of the two nose tackles, we’ve already determined that they were both right at the top of the league in terms of their production. Therefore, if we can establish that the amount of double teams each had to deal with was in line with their peers, then we can more confidently state that their production was elite. However, if we were to find that some nose tackles were being doubled a lot more often, then that might be an explanation for why those players were unable to match the production Ellis and Harrison achieved.

The wisdom behind charting common opponents is two-fold. First of all it allows us to chart more players in the same timeframe, as opposed to the more time consuming process of collecting a whole season’s worth of data. Secondly, as we’ve noted, different teams have different approaches, so if you simply compared a season’s worth of data, you could end up comparing one player who faced more teams that habitually double teamed that position a lot and this could artificially skew the data.

For each set of data, we will compare the data against those common opponents with the data for Ellis and Harrison against those same opponents and where relevant will also state each player’s averages per 100 snaps (for comparison with the following: Ellis 26 doubles, 24 peels and Harrison 14 doubles, 35 peels).

Backup run stopping specialists

First of all, we’re going to consider players that had a similar role to Kenrick Ellis. Ellis played 210 snaps in 16 games, 57% or 120 of which were running plays.

The best three comparisons (all of whom graded positively against the run per PFF) were as follows:

1. Jonathan Hankins, a 320-pounder who – while he played in a basic 4-3 system – was the closest to Ellis in terms of per-snap production (195 snaps, 113 runs);

2. Ishmai’ily Kitchen, a 330-pounder who played primarily as a nose tackle in a 3-4 system (191 snaps, 116 runs); and

3. Dan Williams, another 330-pound 3-4 nose tackle, in his fourth year in the league (291 snaps, 142 runs).

Unfortunately, we can’t learn much from Hankins, because the Jets and Giants only had one common opponent last year (Oakland). He’s probably the least similar to Ellis is terms of being a space-stuffing block-occupyer anyway. For the record, Hankins played seven run snaps against Oakland and was peeled twice with no doubles. Ellis was doubled once and peeled twice in five snaps.

Williams is a bit more useful, given that the Cardinals and Jets faced three common opponents (Carolina, Tennessee and Atlanta). However, he only played six run snaps each against these three opponents. In those 18 snaps, he was doubled once and peeled nine times. The sample size is perhaps too small to draw any concrete conclusions, but that’s a high peel rate and a low doubles rate based on Ellis’ (and Harrison’s) numbers.

Against these particular opponents, however, Ellis was doubled five times and peeled seven times in 30 snaps. Again Williams had a low rate in terms of full doubles, but a higher rate in terms of overall plays where he was doubled or peeled. Enough to justify how much Ellis out-produced him on a per-snap basis? Probably not, especially considering the fact Ellis was doubled more.

For the record, Harrison was doubled nine times and peeled 15 times against these three opponents in 62 snaps, far exceeding Williams’ numbers in terms of doubles, but with a lower overall attention rate.

Kitchen, while he had the lowest run defense grade of the three, does give us a decent comparison because we have six common opponents (Miami, Baltimore, Cincinnati, Buffalo, Pittsburgh and New England). Where a common opponent was faced twice due to being a divisional rival, both games are included in the numbers. Against those opponents, Kitchen had six doubles and 23 peels in 63 snaps. Against the same opponents, Ellis had 23 doubles and 12 peels in 67 snaps. Here we can clearly see that Ellis was doubled or peeled more often with a significantly higher percentage of those plays being full doubles. That his per-snap production was so much higher than Kitchen’s despite this gives some indication of what an upgrade he would theoretically provide for that team.

For the record, Harrison was doubled 22 times and peeled 62 times in 169 snaps against these teams. His peel rate is effectively the same as Kitchen’s but his rate of doubles was higher against these opponents (13.6 per 100 snaps, which is in line with his average, compared with 9.5 for Kitchen). That’s of course in addition to Harrison taking on a starter’s workload.

This seems to clearly show that Ellis was the best backup nose in the NFL last year. While some would attribute that to the fact that his teammates around him are so good, the data in terms of how much double-team attention does not support any suggestion that this could be the sole reason for his success. In fact, if anything, it could suggest he played more of a role in the success of the run defense than he has received credit for.

Looking at the contracts for a player with a similar role to Ellis might be instructive in terms of the likely deal he can expect to get if as his contract is set to expire at the end of the season. However, each of these three players are still on their rookie deals, so this doesn’t help us. However, Williams – a first round pick from the year before Ellis was drafted – is also a pending UFA, so it might be worth looking out for him getting a deal which could set the market. Generally, an older player would need to either be playing a more prominent role, otherwise teams would be reluctant to pay them much more than the minimum, especially if they felt they could obtain equivalent production for less money from a younger player.

Starting run stopping specialists

We now move on to find some close comparatives for Damon Harrison to get some idea of how his numbers stack up against some of his peers. Harrison played 510 snaps and 57% or 284 of these were running plays. These were the three players I chose to compare him to:

1. Brandon Mebane is only 311 pounds but was the closest defensive tackle to Harrison in terms of his run defense grade and did play a comparable number of snaps (540) and run snaps (277).

2. Paul Soliai is a very close comparison in respect of size (344 pounds), total snaps (526) and run snaps (290). He did grade positively against the run.

3. Earl Mitchell actually graded poorly against the run last year, but maybe that was attributable to him getting more doubles than people realized. He’s a good comparison for Harrison in terms of his role (3-4 nose, comes out of the game on passing downs), although he is listed at less than 300 pounds. He’s of particular interest to Jets fans because he’s Soliai’s replacement in Miami. He played 325 run snaps in 2013.

Mebane faced five common opponents in 2013 (Carolina, Tennessee, Atlanta, Tampa Bay and New Orleans twice). In those games, he was doubled 15 times and peeled 31 times in 92 snaps (equivalent to 16 doubles and 34 peels per 100 snaps, which is pretty similar to Harrison’s averages). Against those five opponents, Harrison was double teamed 13 times and peeled 22 times in 85 snaps. Here we can see that the common opponents comparison is useful and that Mebane did receive more attention than Harrison overall.

For the record, Ellis was doubled six times and peeled 15 times in 42 snaps against this group of teams (equivalent to 14 doubles and 36 peels per 100 snaps).

Soliai faced nine common opponents (Cleveland, New Orleans, Baltimore, Buffalo twice, New England twice, Cincinnati, Tampa, Carolina and Pittsburgh). In those games, he had 214 run snaps and was doubled 27 times and peeled 79 times (equivalent to 13 doubles and 37 peels per 100 snaps, again pretty similar to Harrison’s averages). Against those opponents, Harrison played 192 snaps and was doubled 27 times and peeled 72 times. On this occasion the double team rate is slightly higher (14) and the peel rate is about the same. Certainly we can say that these two players received a comparable amount of attention.

For the record, Ellis played 88 run snaps against this collection of teams and was doubled 25 times and peeled 19 times, emphasizing just how often he was the recipient of a full double team.

Mitchell played a total of 112 run snaps against four common opponents (Tennessee twice, Baltimore, Oakland, New England). In those games, he was doubled 17 times, but peeled only 28 (equivalent to 15 doubles and 25 peels per 100 run snaps). The double rate is actually pretty high, but the overall amount of attention he saw is lower than we’ve seen from any nose tackle so far. From watching the film I can also state that most of the double teams he did draw came on plays where he was initially single blocked and another blocker came over to double at the end of the play. I’d also say that not many of these were on plays where he had to be doubled because he was getting good penetration, just where another defender had backed off, fallen down or avoided their initial blocker. I suspect the Texans are a special case anyway because of JJ Watt, who will feature prominently in a later instalment of this series.

Against these teams, Harrison played a total of 79 run snaps and was doubled eight times and peeled 24 times (equivalent to 10 doubles and 30 peels per 100 snaps). In that regard, Harrison’s numbers were actually less impressive than Mitchell’s against these teams, even though they were well below his overall averages. Also, as noted, the nature of double team that Mitchell typically received was arguably less impressive.

For the record, Ellis only played 28 snaps against this group of teams and was doubled four times and peeled nine times (equivalent to 14 doubles and 32 peels per 100 snaps). Again, this is pretty similar to Mitchell and Harrison, but perhaps too small of a sample size to get too excited over.

In terms of contracts, we can get some idea from these players at the sort of figures the Jets and Harrison’s agent will be discussing if they look to extend the pending restricted free agent’s deal. Mebane is signed for five years, $25m. That was signed back in 2011 though. Mitchell signed this offseason for four years, $16m. Soliai signed a five year, $33m deal with the Falcons this offseason, which followed his two-year, $12m deal with Miami. That five year deal could easily prove to be a three year, $19m deal for Soliai, who is already in his thirties.

The “Elite”

What about the so-called elite defensive tackles around the league? Many of these get their reputation by virtue of being good pass rushers in addition to being solid against the run. However, while we’d anticipate the amount of attention they get to be high, how would the numbers for the running game only stack up against a run stopping specialist like Harrison?

Here are the players I selected to make a comparison with:

1. Geno Atkins, whose specialism is most certainly pass rushing, but who is without question a disruptive force in the running game too. However, being 293 pounds, he’s not really the type of space stuffer you’d associate Harrison with being. In fact, that’s a role the likes of Domata Peko would typically have with the Bengals, but do the numbers reflect this?

2. Dontari Poe, a pro bowler who is perhaps the closest comparison to Harrison and Ellis in terms of his size, but impressively was on the field for over 1,000 snaps in 2013.

3. Ndamukong Suh, another pro bowler who again is more of a 4-3 tackle but this time slightly bigger at 307 pounds. Do elite players like these typically draw as much attention as a classic nose tackle type?

I’m also going to look at Vince Wilfork, just out of interest, because that’s another guy still considered in some circles to be elite.

With Atkins, whose season ended prematurely due to injury, there are five common opponents (Pittsburgh, Cleveland, New England, Buffalo and Miami) against whom he was doubled eight times and peeled 29 times in 78 snaps (equivalent to 10 doubles and 37 peels per 100 snaps, so slightly below Harrison’s averages). Against those same opponents, Harrison played 139 run snaps and was doubled 25 times and doubled 51 times (equivalent to 18 doubles and 37 peels, so above average for him). It seems apparent that as disruptive as Atkins is, he doesn’t have to deal with as many blockers as a specialist run stopper like Mebane or Harrison (or even Ellis).

In terms of Ellis’ numbers against those same opponents, he played 55 run snaps and was doubled 15 times and peeled 14 times.

Poe played 128 run snaps against a total of four common opponents (Tennessee, Oakland twice, Cleveland and Buffalo) and was doubled 18 times and peeled 51 times. That’s equivalent to 14 doubles and 40 peels per 100 snaps.

Against these opponents, Harrison played 109 run snaps and was doubled 15 times and peeled 36 times (equivalent to 14 doubles and 33 peels per 100 snaps). In this example, they were double teamed an equivalent amount of times, but Poe received more temporary doubles. Would this additional occupying of blockers be enough to account for Harrison’s superior statistical productivity? Against the same opposition, Ellis played just 37 run snaps and was doubled eight times and peeled nine times, a higher rate of doubles and a lower rate of overall attention.

Suh played 101 run snaps against a total of five common opponents (Tampa and the AFC North). In those games he was doubled 10 times and peeled 44 times.

Against these opponents, Harrison played 85 run snaps and was doubled eight times and peeled 39 times (equivalent to nine doubles and 46 peels per 100 snaps). Against the same opposition, Ellis played just 44 run snaps and was doubled 19 times and peeled just five times. Clearly this is a group of opponents that treated Harrison and Ellis very differently.

Finally, Wilfork had a total of 48 run snaps against three common opponents (Atlanta, Tampa Bay and Buffalo). He was peeled 25 times, but never doubled. As we can see from all the data so far, you expect peels plus doubles to work out at about half the amount of snaps. However, to not have any full double teams is surprising, especially when against those same three opponents, Harrison was doubled 14 times in 77 snaps (and peeled 27 times) and Ellis was doubled six times in 29 snaps (and peeled six times).

Re-visiting 2012

To get a better context of Harrison’s role within the 2013 defense, I went back through the entire 2012 season and charted the doubles and peels for Sione Po’uha. 2012 was not Po’uha’s best season as he was suffering with a bad back all year. However, he still played 201 run snaps, primarily in the same nose tackle role as Harrison, started in 10 of his 12 appearances and had an essentially neutral run grade.

In this instance, I feel it is better to go to the trouble of charting the entire season. Charting common opponents across different seasons is a flawed science because the players on each team will be different as may schemes, coaches or strategic approaches.

Surprisingly, Po’uha was doubled more frequently than Harrison. He was doubled 34 times and peeled 75 times in 201 run snaps, which translates to an impressive 17 doubles and 37 peels per 100 run snaps. It’s worth reiterating that this wasn’t Po’uha at his most disruptive either. In fact, I can recall remarking during a BGA early in the season that I was surprised at how often the Dolphins left Mike Pouncey to block him one-on-one, which would suggest that I was used to seeing him doubled even more than he was throughout that game and perhaps all season.

There is one key difference between the two data sets, although this only impacted on a handful of plays. After the 2012 season, it was no longer permitted for an offensive lineman to go low on a player who was engaged with another blocker. On a few occasions, Po’uha was doubled in this fashion (and it would, according to my stated method in part one, constitute a double and not a peel, unless that player had time to get up and try to block someone else before the ball carrier was tackled or broke through to the second level).

While this led to Po’uha getting credit for being doubled on some plays that would not have happened to Harrison, it’s worth re-considering what we’re trying to measure here and that’s the degree of difficulty for these defensive linemen. Therefore, the fact you can no longer do this arguably makes the defensive lineman’s job easier as that blocker might no longer be able to make contact and the defensive player would no longer need to be alert to the possibility.

Conclusions

We seem to have enough evidence to suggest that Harrison’s high level of production was not as a direct result of him benefiting from not being double teamed that often, since his numbers are for the most part comparable with those of his peers. In the end, his double team numbers were on a par with the likes of Paul Soliai and Ndamukong Suh. However, there were some players – Dontari Poe and Brandon Mebane – who were almost as productive against the run while being doubled more often. It’s difficult to assign any kind of value to those additional blockers being occupied but it does need to be accounted for when placing his grades and statistics for 2013 in the correct context. Earl Mitchell also occupied a slightly higher number of blockers based on the data from common opponents, but he was less productive than Harrison and graded out poorly. However, that could at least mean that maybe he isn’t as bad as his grades would suggest.

It’s also interesting to note that Harrison was doubled a lot more than Geno Atkins – a disruptive force but more of a pass rush specialist and not really a space stuffer – and the fact that Vince Wilfork wasn’t doubled at all in the three games against common opponents is also pretty surprising.

The fact that Po’uha was occupying more blockers in 2012 than Harrison did in 2013, even though this wasn’t one of Po’uha’s better seasons, does once again shed light on how under-appreciated his contributions were at that time as he battled through back issues. We’ll be investigating whether this meant other players were being doubled less as the series continues.

For Kenrick Ellis, the numbers would seem to suggest that the fact he was doubled at a much higher rate than any other player had less to do with the fact that he had a lighter workload than Harrison than it had to do with the way teams opted to approach the running game with him in there. The fact he was so productive and the run defense performed better with him in there than without him is a testament to how well he handled this.

In part five, we’re going to try and add context to Muhammad Wilkerson’s 2013 season by comparing with 2012 to see how the amount of double teams he saw compared and also try and look at some other elite defensive linemen to see how this compares with them.




25 comments
Disgruntled Jets Fan
Disgruntled Jets Fan

wow, this has turned into a pretty epic study!  sports need a peer-reviewed academic journal for this sort of thing.

bklyndude
bklyndude

Wow,  obviously a very time consuming project,  thank you for providing us with your findings.  Ellis/Harrison is a nice problem to have to worry about.

I guess we know we have some studs on the line and have to prioritize which ones we sign long term to the big contracts in the near future.  It would seem Ellis is going to be the player that is allowed to go as a FA and become a starter for somebody else and command a nice contract.

Are the Jets are going to use this season to find out what Kenrick Ellis can do with more playing time,  or do they just continue with the same rotation ?

NYCPEinGermany
NYCPEinGermany

Fantastic stuff, thanks for the in-depth analysis 

bob
bob

Sounds to me like Wilkerson and Richarsons numbers will go up with Ellis in. Ellis would need to stay healthy however.

Brendan
Brendan

Excellent work, Bent. The Po'uha stuff is interesting. Makes you wonder how often it happened when he was at his peak. 



And seeing Earl Mitchell pop up in there made me laugh. 

kniff
kniff

With our considerable and very strong depth at the Defensive line, I wonder if it would it be reasonable to tap one or two of them on the shoulder, perhaps once a week, and discreetly (or not) get them doing reps on the Offensive side of the ball, at the Guard or Center positions for example, to make sure we are prepared for anything and everything (injuries, poor performance, ect.)


Sorry bros, just thinkin' outside the box because of my continued concerns for our O line....



harold
harold

Based on your numbers do you feel Harrison is better than Ellis?

Do think Ellis could have done more in the same full time role?

a57se
a57se

Good stuff Bent.



Can't wait for the conclusion!

jetlife
jetlife

@kniff Respectfully disagree. I appreciate the open-mindedness though.  You have to consider that for almost the entire time training camp and beyond takes place, offense and defense are separated.  Game days is when they come together. Does anyone know how many successful both-sides-of-the-ball linemen there are/have been?

Bent
Bent moderator

@harold I don't think this data answers that question.  He did actually perform slightly better and the run defense was better with him in there on a per-snap basis (and all this data confirms for us is that the reason for this was not that one was doubled more than the other).  However, that was affected by the situations he was used in.  For Harrison to play over twice as many snaps at an equivalent level is a much higher value contribution to the team and a bigger achievement.  They were both great though.


You hit the nail on the head with your second question (which again is impossible to answer with anything other than speculation).  Who knows if he had the motor to continue to perform at such a high level with a bigger workload?


However, he wasn't 100% throughout the first half of the year, so he'd be quick to point out that perhaps he could have achieved more over the course of the year.


I think the team needs to try and figure out the answer to that question before they decide what to do with their contracts, if possible.  Unfortunately, that might not be practical.

Bent
Bent moderator

@a57se This might get to double figures by the time all is said and done.

Disgruntled Jets Fan
Disgruntled Jets Fan

@Bent @harold This sounds bad, but they may need Harrison to get injured for a couple of weeks (something with no long term impact) sso they can see how Ellis handles a full load.

Bent
Bent moderator

@Hanknaples Yes both have been awesome.


The World Cup is not over yet.  Hopefully the final lives up to the rest of the tournament.

Bent
Bent moderator

@Disgruntled Jets Fan @Bent @harold I just hope they don't keep Ellis' role the same in the hopes that this keeps his price down and enhances their chances of keeping him.

harold
harold

@Bent @Disgruntled Jets Fan

First thanks for answering my questions.

I lean toward Ellis as the potential star of the two.  While I think Harrison is and can be a productive player for years to come.

I think Ellis has the potential to be a star.  Because he can hold up against the run and has shown much more potential rushing the passer.

Hopefully, they split the reps morte evenly this season so we can as you stated determine what Ellis truly is/can be.

Bent
Bent moderator

@Hanknaples I think that game made this world cup MORE memorable.  But if the final is a drab 1-0 win or decided on penalties, that's going to be a letdown.

Bent
Bent moderator

@harold You could make an argument that Harrison is already a star.

harold
harold

@Bent

I don't think I would go as far as a star.  One very good season does not make him a star, imo.  Especially since he has been a one dimensional player (although a very good one) so far. 

I think Ellis has the potential to be a 3 down player.  I dont see the potential in Harrison yet, although to be fair he is young.

harold
harold

@Hanknaples

Respected yes, a star I don't agree with that.

Also was down in Naples at the Waldorf Astoria a couple months back.  Nice area really enjoyed it.

harold
harold

@Hanknaples

Next time I will let you know so you can show me the town.