Writers of Pro Football Prospectus 2008

27 Jan 2010

Stat of the Day: DL Stop Rate vs. Run

As part of our ongoing Stat of the Day series, we're digging deep into our spreadsheets to run a new stat every weekday until Super Bowl XLIV.

Today, we'll look at which defensive linemen led the league in Stop Rate against the run. This measures Plays against the run (any tackle or assist, according to official play-by-play) which counts as successful under FO terms: preventing the offense from getting 45 percent of yards on first down, 60 percent on second down, or 100 percent on third and fourth down. Along with Stop Rate for each player, I'm running the average yardage gain when the player made a Play.

This year's leader is a big surprise.

Best Run Stop Rate for Defensive Linemen, 2009
(min. 16 run tackles)

Player

Team

Pos

Run Plays

Stop Rate

Avg Yds
95-S.Cody HOU DT 18 100% 1.4
77-C.Jenkins GB DE 19 95% 1.7
95-J.Babineaux ATL DT 35 94% 0.5
75-J.Parker PHI DE 16 94% 1.6
92-A.Franklin SF DT 30 93% 1.2
96-M.Spears DAL DE 22 91% 1.9
92-C.Davis ATL DE 19 89% 1.7
99-M.Wright NE DT 28 89% 2.6
92-S.Ellis NYJ DE 42 88% 2.0
94-J.Smith SF DE 42 88% 1.8
96-P.Soliai MIA DT 25 88% 2.1
90-M.Williams HOU DE 32 88% 1.4

Shaun Cody was just an average tackle for Detroit from 2005-2008, but he was a great run-stuffer for Houston after they added him to the starting lineup in Week 4. He was just one reason why the Texans were fifth in ALY after being 30th in 2008.

Dwight Freeney also had a 100 percent Stop Rate. Of course, he only made 8 Run Plays all year.

Here's the other side, with the defensive linemen who struggled to stop the run. The two Saints high on this list help demonstrate why run defense is their big weakness, although Charles Grant hasn't played in the postseason due to injury.

Worst Run Stop Rate for Defensive Linemen, 2009
(min. 16 run tackles)

Player

Team

Pos

Run Plays

Stop Rate

Avg Yds
95-M.Stanley JAC DT 17 35% 3.1
92-R.Denney BUF DE 21 43% 3.8
75-T.McBride DET DE 16 44% 3.2
96-T.Crowder TB DE 37 46% 5.4
74-J.Cesaire SD DE 23 52% 4.4
98-S.Ellis NO DT 30 53% 3.5
94-C.Grant NO DE 38 55% 2.9
99-G.Ellis OAK DE 18 56% 4.9
93-R.Miller TB DT 30 57% 3.0
94-T.Jackson KC DE 35 57% 3.0
90-I.Sopoaga SF DE 27 59% 3.3
94-M.Harrison CHI DT 20 60% 1.9
92-C.Avril DET DE 25 60% 3.6
79-R.Brock IND DE/DT 20 60% 2.6

Posted by: Aaron Schatz on 27 Jan 2010

37 comments, Last at 30 Jan 2010, 11:47am by Felton

Comments

1
by FooBarFooFoo (not verified) :: Wed, 01/27/2010 - 6:13pm

Mmmh, how do you compile this? How do you know which player is responsible for which gap/assignments?

2
by FourteenDays :: Wed, 01/27/2010 - 6:38pm

This table doesn't take that kind of thing into account. It's simply the number of tackles DLmen made on run plays, and the percentage of those tackles that resulted in "stops" (less than 45% of the needed yardage on 1st down, 60% on 2nd down, or 100% on 3rd or 4th down). That's all.

3
by FooBarFooFoo (not verified) :: Wed, 01/27/2010 - 6:42pm

Well but how do you count the "run plays" a given player took part in then? Sure you take look at the tackles the guy actually did make. But what about the plays he did not make? You cannot just take them out of your statistics, or, you just told me that the stat is useless.

5
by Dean :: Wed, 01/27/2010 - 6:49pm

Unless you're the coach, you're not going to know whose assignement a particular gap was, so you do the best with the information you have.

6
by Vincent Verhei :: Wed, 01/27/2010 - 6:51pm

From the first paragraph of this story:

Today, we'll look at which defensive linemen led the league in Stop Rate against the run. This measures Plays against the run (any tackle or assist, according to official play-by-play) which counts as successful under FO terms: preventing the offense from getting 45 percent of yards on first down, 60 percent on second down, or 100 percent on third and fourth down. Along with Stop Rate for each player, I'm running the average yardage gain when the player made a Play.

Personally, I find total Stops to be more telling than Stop Rate. Shaun Cody only made good plays ... but he only made one, maybe two plays per game.

10
by FooBarFooFoo (not verified) :: Wed, 01/27/2010 - 7:52pm

I was honestly wondering how one could compare a player with 18 plays to a player with 40+ plays. But that's not the issue. This stat tells very little to nothing. You can't even compare two players with a similar number of plays. Sure, a stat is never ever going to be completely objective and telling, but I have a gut feeling this one is extremely useless.

On a completely unrelated note: How do Vince Wilfork's numbers look like and will he be worth it?

The second question (completely unrelated and in no way judgemental) is how you get the official play by play to your db. Is that a manual process, or do you have some fancy parsing algo or do you get the play by play in structured form?

26
by Scott C :: Thu, 01/28/2010 - 5:25am

Its not "completely useless".

Certainly, its a 'noisy' stat and a context sensitive one. But it does tell you a lot about the difference between players that are far apart on the chart.

Missed tackles would certainly tell us more, as would "blocked out of the play" numbers and who blocked them (double team? single team? fullback?). But not having that doesn't make it useless, it just has to be interpreted for what it is.

31
by Rich Conley (not verified) :: Thu, 01/28/2010 - 12:58pm

I'm convinced that in the this case, stop rate is completely useless.

Stops per play would be a much better metric.

Any metric where a guy is better off whiffing at a guy then stopping him for a 5 yard gain has its problems. The difference between a guy like Cody (18 stops in 18 run tackles) and a guy like Sean Ellis (15 stops in 30 run tackles) may just be that Cody only gets a tackle when someone runs directly at him, and Ellis is making tackles that other guys have missed.

you're essentially punishing players for making tackles someone else should have made.

36
by tgt2 (not verified) :: Thu, 01/28/2010 - 1:59pm

Exactly my thoughts. Take the Raven's Ngata. He makes some run stops at and behind the line, but he also has been known to chase down RBs from behind 10 yards down field on sweeps.

The noise overwhelms any usefulness.

27
by Scott C :: Thu, 01/28/2010 - 5:30am

On your last question, all plays are charted. For passing plays, the player in coverage, "overthrown", "underthrown", depth of throw, middle or side, etc are charted.
For running plays, middle (either side of center), left/right guard (inside tackles), and left/right tackle (outside the tackle) runs are noted.

Number of WR's, RB's, FB's, and TE's are also noted, as is the number of pass rushers.

4
by Dean :: Wed, 01/27/2010 - 6:48pm

Juqua Parker's name jumps out at me.

Only 16 run stops, but he was strictly a part-time/rotational player. And when you add in the fact that he had 8 (?) sacks, it shows production against the run and the pass.

I wonder about Raheem Brock down at the bottom of the list, as well. I always thought of him as being solid against the run. I want to say Indy moved him inside to DT, but I could be wrong. If so, it looks like he's getting overpowered and maybe isn't big enough to play inside.

7
by Marcumzilla :: Wed, 01/27/2010 - 7:02pm

I'm just working off memory, but I believe he's played outside. He may have lined up as LDE when Mathis was out / Freeney was out and Mathis moved to RDE.

I'm guessing they moved him inside due to losing so many players to the DT curse.

13
by turbohappy (not verified) :: Wed, 01/27/2010 - 8:21pm

Brock has played all over the place this year. He's been the guy who drops into coverage on a zone blitz type scheme, for example. He hasn't had a hand in the dirt at the snap all the much this year, he's been seen standing up near the line a lot.

24
by Bobman :: Thu, 01/28/2010 - 4:29am

And if he gets dinged for allowing Ray Rice's 20 yard run, he should at least get credit for causing the fumble on the tackle 20 yards downfield. I'll take that every day, except maybe in the red zone.

Brock's value to the Colts is probably immeasurable. Like a middle-reliever who can go for 5-6 innings, or set-up/close if need be, he's the do it all man.

8
by Who Dat (not verified) :: Wed, 01/27/2010 - 7:07pm

The Saints are a d-lineman or two away from being incredibly dominant, IMO. Considering that we've spent half of our last 8 first round picks on d-lineman, that's kind of frustrating.

9
by AnonymousA (not verified) :: Wed, 01/27/2010 - 7:46pm

Two observations:

1) There are a lot of Ellis' here. 2 on one list, one on the other. That's weird.

2) As a pats fan, my suspicion is confirmed: Wilfork not present. He's not as good as people think...

28
by Felton (not verified) :: Thu, 01/28/2010 - 8:20am

Isn't Willfork's role to occupy two offensive linemen and keep the LBs free? If so, he would not be on the list because he is not tackling, but allowing others to tackle.

11
by Joe M. (not verified) :: Wed, 01/27/2010 - 7:54pm

I don't have the slightest clue how this could be a useful stat. A D-lineman who chases down runners five to ten yards downfield would be someone you'd want to reward, wouldn't you? And a D-lineman who either stops a guy in his tracks or lets him go past him and doesn't make much effort to tackle him if he gets by would be a bad thing, wouldn't it?

12
by ammek :: Wed, 01/27/2010 - 8:15pm

Stats don't have to be useful. Sometimes they're just interesting.

14
by BadgerDave :: Wed, 01/27/2010 - 8:26pm

I wonder if you could use these numbers for "lineman downfield hustle" by looking at plays that weren't "stops" but started off away from the lineman.

15
by parttimemovieguy :: Wed, 01/27/2010 - 9:23pm

I'm surprised that Dallas' Jay Ratliff didn't make the top ten. By all accounts, he's one of the best DT's this season. What's his rank this year?

Based on 2008's numbers, he was #8 with 31 stops and 84%.

16
by Mr Shush :: Wed, 01/27/2010 - 9:59pm

I don't agree with those who think these stats show nothing, but they do ignore plays on which a guy made a valuable contribution without making a tackle. Top nose tackles, like Ratliff, frequently eat two blockers without being moved back, which means both that it's easier for someone else to make the play and that the running back essentially can't go up the middle. If they penetrate the backfield, the back may be forced to divert to an outside lane where his blocking is not designed to open a gap, perhaps forcing a tackle for a loss by a corner, end or OLB. Nose tackles are damn near impossible to evaluate statistically, because they can excel at their jobs without ever making a play that shows up on the stat sheet. That's probably never going to change.

17
by alexbond :: Wed, 01/27/2010 - 10:03pm

What the hell is going on in San Fran? Two guys near the top of the NFL and the third DL near the bottom?

John Babineaux looks like an animal from these stats, do any ATL fans have comments on his play?

18
by Brendan Scolari :: Wed, 01/27/2010 - 11:18pm

Justin Smith, Aubrayo Franklin, and Patrick Willis are really, really good. Isaac Sopoaga is not that good (although I'd still say he's a decent run defender, but he wasn't very impressive this year).

20
by Karl Cuba :: Wed, 01/27/2010 - 11:53pm

Sopoaga also plays the strong side and so is more likely to have to deal with the tight end chipping him.

As has been pointed out by several other posters, this isn't a perfect stat and if you look, Sopoaga still made plenty of plays: 30, the same number as Franklin; but his stop rate was lower. However, there is only a 28% difference between the two charts and this is a rather unreliable staistic with sample size issues.

21
by Peregrine :: Wed, 01/27/2010 - 11:54pm

Babineaux was absolutely our best DL this season, as Abraham tailed off from 2008. He's more quick than powerful, but he makes plays and usually ranks high in tackles for loss.

I just noticed the other day that the Falcons, much to my astonishment, finished #2 in the DVOA run defense rankings for the regular season. Reason #1, I'm convinced, is that we let Keith Brooking - the Six-Yard Sheriff - take his special brand of "leadership" and whining to Dallas. Reasons #2 and #3 might have been Babineaux and the growth of Curtis Lofton. Of course, good coaching helps too. Mike Smith knows what he's doing.

19
by tuluse :: Wed, 01/27/2010 - 11:20pm

Is Marcus Harrison's average yards right? How do make your average tackle after 1.9 yards and only have a 60% stop rate?

23
by Brian Davis (not verified) :: Thu, 01/28/2010 - 1:43am

Negative plays maybe?

25
by Bobman :: Thu, 01/28/2010 - 4:31am

Or his only tackles are on 3rd and 1 and 3rd and 2 situations...?

22
by Temo :: Thu, 01/28/2010 - 1:28am

According to my participation stats, Cody played in about 1/3 of his team's snaps. I wouldn't attribute that increase in ALY to him, or at least him alone.

I think he's a good example of what can happen when you surround a "meh" player with real talent like Mario Williams, DeMeco Ryans, and Brian Cushing.

30
by Mr Shush :: Thu, 01/28/2010 - 9:19am

I'm guessing here, but I suspect that the snaps Cody missed were mostly either early in the season, before he entered the starting line-up, or on passing downs, where he generally came out and Antonio Smith moved inside. I'm not saying Cody's some kind of superstar, and nose tackle is definitely a position I'd like to see the Texans upgrade this off-season (Hampton? Franklin?), but the guy, along with the addition of Pollard and the ongoing development of Cushing (who missed most of camp and pre-season through injury) was a significant factor in the improved run defense. The Texans now have a pretty good young front seven; it's a pity the secondary still kinda sucks.

29
by Felton (not verified) :: Thu, 01/28/2010 - 8:29am

Aaron - interesting stuff. Is it possible that a poor DL stop rate is a function of LB play? If so, an excellent DL stop rate tells us something about a player while a poor DL stop rate tells us more about the defense overall. Then there are the DL who occupy blockers.

32
by Rich Conley (not verified) :: Thu, 01/28/2010 - 1:02pm

":If so, an excellent DL stop rate tells us something about a player"

What exactly does it tell us? It doesn't tell us that the guy is making every tackle. It just tells us that all the tackles hes making are near the line of scrimmage. That isn't necessarily a good thing.

37
by Felton (not verified) :: Sat, 01/30/2010 - 11:47am

Rich, if a guy is making a tackle 1 yard upfield, that's good in most situations. Further upfield is better, but a tackle at 1 yard tells me the guy is beating his man and making a play. A tackle 4 yards downfield could tell me the defender was beaten or it could tell me that the d-lineman got off his block and had to make a tackle because someone else did not get the job done. Am I wrong?

33
by Rich Conley (not verified) :: Thu, 01/28/2010 - 1:03pm

Any chance we can get a "Total Stops" leader list?

34
by Huh, what? (not verified) :: Thu, 01/28/2010 - 1:12pm

I don't see how stop rate tells us anything. Guy gets blocked at the point of attack and somebody ELSE cleans up his failure, makes the 2nd guy look bad does it not?

35
by big_jgke :: Thu, 01/28/2010 - 1:23pm

No real insight here, but Buffalo is terrible. They could have like five first round picks this year, and it still wouldn't fill the holes at: DE, DT, TE, QB, OT, OG, both S and LB. The talent deficit on that team is striking.