Writers of Pro Football Prospectus 2008

21 Jan 2010

Stat of the Day: Pass Tackles for Cornerbacks

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.

We know that not all tackles are created equal. There's a big difference between a cornerback in run support and a cornerback who allows a big completion and then gets credit for a tackle. Football Outsiders separates defensive Plays into run tackles, pass tackles, and sacks. When we say "Plays," we're referring to any time the defender's name appears in the standard play-by-play, whether it is a solo tackle or an assist. To be clear, these stats are not from our game charting project.

First, here are the cornerbacks who made the most tackles after receptions. (Of course, that doesn't necessarily mean a reception by the guys they were covering.)

Top 10 CB Making Plays on Complete Passes, 2009
Player Team Plays
20-N.Harper TEN 62
27-J.Lacey IND 61
31-R.Marshall CAR 59
22-W.Gay PIT 58
33-C.Tillman CHI 57
41-W.James DET 56
25-B.McFadden ARI 54
22-J.Joseph CIN 54
24-C.Bailey DEN 52
35-Z.Bowman CHI 51

Of course, not all tackles are equal. In zone coverage, it's perfectly fine to let your man catch a short pass as long as you keep him from a first down. We count a play as successful (a "Stop") if it prevents 40 percent of needed yards on first down, 60 percent on second down, or 100 percent on third/fourth down. Here are the top cornerbacks in pass Stops in 2009.

Top 10 CB Making Stops on Complete Passes, 2009
Player Team Stops
31-R.Marshall CAR 20
32-O.Scandrick DAL 19
33-C.Tillman CHI 18
20-N.Harper TEN 16
23-C.Griffin MIN 16
24-T.Thomas NYG 16
24-D.Foxworth BAL 15
24-S.Brown PHI 15
21-C.Woodson GB 15
Five tied with 14

So now we know who had a lot of Plays and Stops, but what about a cornerback who played well but was generally avoided by opposing quarterbacks? Let's take a look at which corners had the highest Stop Rate when making a Play after a complete pass. You'll notice one team stands out when it comes to cornerbacks preventing big gains (although, of course, this doesn't factor in when a cornerback gets burned).

Top 10 CB Stop Rate on Complete Passes, 2009
(min. 15 tackles after receptions)
Player Team Plays Stops Stop Rate
22-C.Jackson ATL 19 9 47%
21-N.Asomugha OAK 15 7 47%
32-O.Scandrick DAL 42 19 45%
21-J.Hanson PHI 31 14 45%
22-A.Samuel PHI 30 13 43%
21-C.Woodson GB 37 15 41%
24-S.Brown PHI 39 15 38%
41-C.Munnerlyn CAR 30 11 37%
24-D.Foxworth BAL 41 15 37%
24-T.Thomas NYG 44 16 36%

If we flip things around to look at the worst Stop Rate on pass tackles, you'll see a very surprising name near the bottom of the list.

Bottom 10 CB Stop Rate on Complete Passes, 2009
(min. 15 tackles after receptions)
Player Team Plays Stops Stop Rate
22-C.Rogers WAS 28 2 7%
35-J.Reeves HOU 21 2 10%
28-D.Butler NE 29 3 10%
23-Q.Jammer SD 38 4 11%
27-M.Jenkins NO 37 4 11%
21-C.Owens ATL 16 2 13%
25-B.McFadden ARI 54 7 13%
26-S.Routt OAK 23 3 13%
24-C.Bailey DEN 52 7 13%
24-E.Wright CLE 44 6 14%

Another surprising name that almost made this Bottom 10: Michael Jenkins of Dallas, who had 40 Plays on complete passes but only six Stops (15 percent).

General caveat applies: This is just one stat that helps us look at how cornerbacks played in 2009, both in terms of quality and general style. It is not an ultimate all-in-one metric that tells us which cornerbacks were good and bad.

For those wondering, Darrelle Revis had 34 Plays on complete passes with eight Stops (24 percent).

I'll hit this stat for safeties tomorrow and linebackers sometime next week.

Posted by: Aaron Schatz on 21 Jan 2010

33 comments, Last at 24 Jan 2010, 8:31pm by RickD

Comments

1
by Eddo :: Thu, 01/21/2010 - 2:32pm

Nice to see Charles Tillman near the top of that second chart (since the first one, like noted, doesn't really give much information).

It would be nice to see how many of the tackles were on running backs, as well. It seems like many cornerback tackles are on passes to the flats.

2
by raffy (not verified) :: Thu, 01/21/2010 - 2:36pm

Did I miss the CB which had fewest number of attempts his way? Or is that only found in the charting data? Thank you.

4
by tgt2 (not verified) :: Thu, 01/21/2010 - 2:46pm

uncompleted pass defenders are not listed in the play-by-play. That information along with stops can get us the real total stop %.

3
by Liberal Lion (not verified) :: Thu, 01/21/2010 - 2:45pm

Not surprised to see Darius Butler on that list...he can't tackle for shit.

11
by Nathan :: Thu, 01/21/2010 - 3:29pm

Is there a stat that measures running step for step with the receiver but whiffing on the deflection attempt? Cause I bet Jonathan Wilhite lead the league in that stat.

Or maybe it was Butler, I can't remember cause I drank too much during games this year (go figure) but one of the rookie corners, hot damn, I've never seen a CB be in the right position so many times and yet somehow fail to make the play. He needs some bifocals or something.

25
by ABW (not verified) :: Fri, 01/22/2010 - 12:38pm

Pretty sure that's Jonathan Wilhite. Great wheels and terrible, almost comically bad skills at positioning and deflecting.

27
by Rich Conley (not verified) :: Fri, 01/22/2010 - 3:50pm

EH, you'll see a lot of that when you have a good corner and terrible pass rush.

Those are the sorts of plays where if the QB is pressured, and the ball is a little off, it gets picked. If the QB isn't pressured, and is good, you see a lot of close completions.

29
by Nathan :: Fri, 01/22/2010 - 4:32pm

Go back and watch the 2nd game vs the Dolphins and all you'll see is this over and over:

http://cachemediasrv.patriots.com/ImgDyn.cfm?s=KDN_6628A.JPG&c=1&w=500&c...

30
by Rich Conley (not verified) :: Fri, 01/22/2010 - 8:47pm

uMM, That's kind of making my point. That ball is perfectly placed.

The patriots don't have ANY pass rush, and without a pass rush, cornerbacks simply don't matter.

5
by Nathan :: Thu, 01/21/2010 - 2:50pm

I've had a sneaking suspicion that Champ has been seriously slipping the last 2-3 years... I've asked a couple of times and always heard that he's not... but this seems to confirm my impression that he's getting thrown at and receivers are making completions against him.

9
by The Guy You Don't Want to Hear (not verified) :: Thu, 01/21/2010 - 3:17pm

I didn't see him a lot this year, but as a Bronco fan he seriously slipped the last two years. From what I did see this year (and I was actually lucky enough to see them three weeks in a row early in the year), he actually looked much better than the last two seasons. He also just wasn't very healthy in '07 and '08 and played through most of it, so that could be much of the problem.

In other words, I think your suspicion was correct with a couple of possible caveats: his health in '07-'08 and his level of play in '09.

28
by ATLeagle (not verified) :: Fri, 01/22/2010 - 4:23pm

I only watched him in the games vs the Eagles, but Champ was one on one against DeSean Jackson for much of the game and seemed to do fairly well. I think incompletions not being included is altering these stats, and actual defensive scheme would show that he is actually being put in more high-risk situations than a lot of other CBs, and his stats may not compare fairly.

6
by jklps :: Thu, 01/21/2010 - 2:55pm

Not shocked that Rodgers is #1 ... if he actually tackles somebody he's effective, but he misses.

7
by dk240t :: Thu, 01/21/2010 - 3:08pm

Aaron, I really think you need to update the article to remind readers how widely varying these stats are from year to year. I seem to remember Champ Bailey and other big-name CBs bouncing from the top list to the bottom list from year to year, as well as numerous other really bad corners having fluky good years based on these types of metrics (Houston's Fred Bennett comes to mind). Some stats are predictive, and some stats are practically useless due to small sample sizes and noise; this stat is the latter.

10
by Pat (filler) (not verified) :: Thu, 01/21/2010 - 3:21pm

Some stats are predictive, and some stats are

descriptive. Here, I fixed that for you. A non-predictive stat doesn't mean it's useless if you're trying to figure out exactly how a defense played.

18
by Brendan Scolari :: Thu, 01/21/2010 - 7:43pm

It also doesn't mean it's useful. When the sample size is very small a couple fluke plays can change the numbers drastically, and don't necessarily describe accurately how well the players played. That's not even mentioning how much scheme can influence these numbers. If one corner has safety help over the top on every play and another guy never does, the numbers will not be useful at all. When players often show up on the top of these lists for one year and then are never anywhere near the top again, it's unlikely that the played really, really well for one year and then never returned close to that level. More likely is that the stats failed to accurately describe how well the player played that year because of all the confounding factors.

22
by Pat (filler) (not verified) :: Thu, 01/21/2010 - 11:17pm

accurately how well the players played.

No, but it describes how they played.

If one corner has safety help over the top on every play and another guy never does, the numbers will not be useful at all.

Sure it will. If you know the schemes each of the teams played, you could look at how schemes affect the number of tackles made, correct for it, etc. Within a given team, if you know roughly how each corner was used, you can gauge how well that corner was doing their job, etc.

You're just thinking the numbers are useless on their face, but almost all stats require some degree of thought.

When players often show up on the top of these lists for one year and then are never anywhere near the top again, it's unlikely that the played really, really well for one year and then never returned close to that level.

Yes, that's what's meant when you say it's 'descriptive' rather than 'predictive.' It doesn't describe anything necessarily intrinsic about the player, but it does tell you what happened. It's a lot like sacks - QB hurries+sacks is a lot more predictive than sacks alone, but sacks have a higher effect on the play. So a defense that seems like it's down one year in the number of sacks, but gets a lot of QB hurries, nothing's really changed, it's just statistical noise ("bad luck").

For instance, if you combine Eagles CBs leading the league in tackles for stops, with a lot of interceptions, with a good overall pass defense, with relatively high Open Field yards, you'd make the conclusion that the Eagles CBs are very frequently in position, but tackle poorly. And based on the missed tackle stats from Pro Football Focus, that appears to be correct.

It also depends on how the stat correlates (if at all) to defensive DVOA, points, or wins, for instance. If you've got a non-predictive stat which correlates to wins, that'd be a strong candidate for a rebound the next year for outliers.

It's very hard for a stat to be completely useless.

33
by RickD :: Sun, 01/24/2010 - 8:31pm

Thank you!

I get so tired of people calling statistics "useless" and whining about sample size.

This is football. It's not a science. We're not in a controlled setting, measuring quantities that are fixed in time and space.

13
by Dean :: Thu, 01/21/2010 - 3:47pm

That has been a standard caveat for this entire series of stats. They're presented as interesting conversation-fodder, not as evidence of anything.

8
by ammek :: Thu, 01/21/2010 - 3:13pm

And the moral is: choose a jersey number than ends in 1, 2, 3, or 4.

12
by Dean :: Thu, 01/21/2010 - 3:46pm

Aaron, just wanted to say that this whole series has been great. I know that's hardly the most insightful comment ever, but every day has been interesting. Well done.

14
by commissionerleaf :: Thu, 01/21/2010 - 3:58pm

Cornerback is a really, really hard position to evaluate statistically because it is at one or two removes from the actual ball, and because cornerbacks have responsibilities other than (and superior to) preventing completions.

It's worth noting that Nnamdi was on the field every play and still A. led the league in stop rate and B. Barely got thrown at enough to make the list.

15
by King Caleb (not verified) :: Thu, 01/21/2010 - 4:15pm

Fascinating stuff. Very awesome to see the level of detail here - breaking out pass tackles vs. run tackles is brilliant. It would also be interesting to see tackles by play length - e.g. best/worst tacklers on plays under 10 yards or more than 30 yards.

They don't have the same level of detail, but I've been using Playerfilter to check out stat leaders throughout the season and they have a sortable list of cornerbacks by total tackles here:
http://playerfilter.com/position#pos==CB,u,h;g,r,u;tot,desc;st;at;dsack;...

16
by NHPatsFan (not verified) :: Thu, 01/21/2010 - 5:04pm

To provide context, wouldn't you want to look at both total balls thrown at the corner as well as total ball NOT thrown at that corner? A Play and a Stop, the way you've defined them, are the last and least forms of successful pass defending. Far better to prevent a reception one way or another - actually breaking up the reception in some way, disrupting timing between QB and receiver in one of several ways, even a pick. Best of all, be so good at your job that the other team avoids you.

I'm not sure how you can extract these data - if at all - from your database, but without it, these stats are kinda free-floating.

17
by Brian :: Thu, 01/21/2010 - 6:19pm

As a Bears fan, not surprising to see Tillman and Bowman high up on the first list.

Anyone have an opinion on Zack Bowman, by the way? He led the Bears in picks and I recall some dropped ones as well, but I'm wondering how he measured up overall.

31
by c_f (not verified) :: Sat, 01/23/2010 - 2:22am

The Bears pass rush did not give the DBs a lot of help. Bowman got picked on a fair amount, lots of short stuff. He just got turned around a lot. To his credit he made some nice plays on the ball. Not enough to make up for getting picked on, but it was something.

They decided Vasher had lost it, so Bowman was just going to have to endure a trial by fire, since Corey Graham has regressed badly and DJ Moore couldn't see the field.

19
by JetfanMike (not verified) :: Thu, 01/21/2010 - 7:44pm

These stat-of-the-day things are pretty cool. Thanks, guys.

20
by HostileGospel :: Thu, 01/21/2010 - 9:50pm

Asante Samuel is credited with 13 stops after pass plays. I'm assuming that in all of those, the receiver either fell down or went out of bounds, because the alternative is that Asante actually tackled someone, and I can't believe that.

Seriously, any metric that involves tackles made and says anything good about Asante Samuel is wonderful proof that CB stats (simple or complex) need a lot of context to be evaluated properly. Neat stuff, though.

On a related note, Derek at Igglesblog did some analysis on the PFF numbers for the Eagles secondary this week, which is a good read.
--
Overall, I'd be kind of embarrassed to critique something when I didn't know what the hell I was talking about, but then, oh yeah, my NAME is on what I write, isn't it?

-Les Bowen

23
by bubqr :: Fri, 01/22/2010 - 5:57am

Agree on the Samuel part, and on the IgglesBlog article : it's a good one (I sent some mailbag suggestions to FO about some of the articles they made, surprised me to never see it linked in any Extra Points. they make some really good numbers articles).

3 PHI CBs in the top 10 : Samuel is a joke (see above), sheldon doesn't surprise me, Hanson got lucky he never made the tackle when getting burned bad by Crayton in the last 2 games.

24
by Hurt Bones :: Fri, 01/22/2010 - 9:41am

I had the same sort of thought about Dominique Foxworth. Maybe they were counting the "I'm going to desperately hold onto the receiver's jersey until my teammates arrive and clean up the tackle" occurences.

21
by Dales :: Thu, 01/21/2010 - 11:09pm

I have to tell you- Terrell Thomas must have had a hell of a year. He was up there on your last deep-dive into DB stats.

26
by Nathan :: Fri, 01/22/2010 - 12:55pm

This article made PFT...

32
by Dice :: Sat, 01/23/2010 - 7:54pm

No DeAngelo Hall? Shocked.