Writers of Pro Football Prospectus 2008

Most Recent FO Features

GurleyTod16.jpg

» OFI: SEC Surprises

In an opening week where even the elite teams in college football looked mortal, the SEC had two big surprises in Texas A&M and Georgia defeating their South Carolinian opponents by big scores.

06 May 2014

Best Cornerback Charting Stats 2013

by Aaron Schatz

Let's break out the cornerback charting stats for 2013. These stats come courtesy of the Football Outsiders Game Charting Project. There are two main stats we track. The first, Yards per Pass, is simply the average yardage gained on every pass where we list this cornerback in coverage. The other stat, Success Rate, tells you how often the cornerback prevents opposing receivers from what we consider a successful gain: 45 percent of needed yards on first down, 60 percent on second down, and 100 percent on third down. Note that nickelbacks will often rank much better in Yards per Pass than in Success Rate, while starting corners who blew a couple of huge plays over the course of the season will rank much better in Success Rate than in Yards per Pass.

As usual, the typical caveats about the game charting apply: this is imperfect data charted by a group of volunteers plus a handful of FO staff members. Sometimes a cornerback will benefit because he happens to be in coverage when a quarterback throws a bad pass, even if he wasn't covering close. Sometimes a cornerback will benefit from a better pass rush, because it's easier to cover when you don't need to cover for six seconds. The cornerback charting stats have been very volatile, bouncing up and down year-by-year for a lot of players, and the best cornerbacks will occasionally rank lower in these stats because quarterbacks only throw in their direction when they make a mistake. As we always say, these stats should not be seen as absolute statements on player value. They're just part of the story.

As we do with the cornerback charting stats in the book and on the player pages, I've removed passes marked as Hail Mary, Hit in Motion, Tipped at Line, or Thrown Away. I've also removed wide receiver screens, which aren't really a good way to measure cornerback coverage because a cornerback in man coverage is going to (or at least, is supposed to) immediately get blocked out of the play by another wide receiver. Once again this year, we're bringing you the most accurate numbers by waiting to post cornerback stats until they've been double-checked. These stats have also been analyzed to give half-credit on plays where we list two defenders in double coverage, or where we listed DEFENDER1 as "Hole in Zone" but also list a specific defender as responsible for that zone. We also have both actual numbers and the metrics adjusted for the quality of receivers each cornerback had to cover. (Two metrics, average pass distance faced by each cornerback and yards after catch, are not adjusted for opponent.)

Pass interference is included, although no other defensive penalties are included. With defensive pass interference, the defender flagged is almost always the player who was in coverage on the intended receiver; with illegal contact or defensive holding, the flag often comes far away from where an actual pass may be thrown.

This year compared to past years, there were about a dozen more cornerbacks listed in coverage on at least 40 passes. So for this year's tables, I've raised the minimum to include players who either are listed in coverage on 50 passes or had eight games started. I decided to count Tyrann Mathieu as a cornerback since he played the nickelback role whenever the Cardinals had a sub package on the field, but I'm not including Aaron Williams of Buffalo who played safety most of the year but started at cornerback in Weeks 4-7. Eighty-eight cornerbacks qualified to be ranked in these stats, and I've got a dozen guys on each table here. To give more information about the cornerbacks that saw the field the most, each table also lists the next three cornerbacks who would be ranked if we limited the minimum to only the 36 cornerbacks who started at least 14 games.

Oh, and as an added bonus, three of this season's best cornerbacks will be special Football Outsiders elite stars in Madden 25 Ultimate Team this upcoming weekend. More on that below.

We'll start with the cornerbacks who allowed the fewest Adjusted Yards per Pass according to our game charting. Of course, we could start with Adjusted Success Rate, and we would have the same guy on top. And boy, is it not who you expect. 

Top Cornerbacks in Adjusted Yards/Pass, 2013
Player Team Charted
Targets
Adj
Yd/Pass
Rk Adj
Suc Rate
Rk Avg. Pass
Distance
YAC Rk
38-T.McBride NYG 72 4.2 1 69% 1 13.3 2.2 3
22-W.Gay PIT 82 5.8 2 58% 14 11.9 2.6 11
22-B.Skrine CLE 105 5.9 3 54% 39 11.9 3.3 28
23-M.White CAR 57 5.9 4 55% 33 9.4 3.1 19
29-X.Rhodes MIN 79 5.9 5 53% 42 12.1 2.7 13
29-D.Florence CAR 50 6.1 6 63% 3 12.7 5.2 75
23-J.Haden CLE 88 6.1 7 58% 15 14.6 2.5 8
21-D.Trufant ATL 81 6.2 8 56% 24 13.1 3.4 30
21-B.McCain HOU 59 6.2 9 51% 54 10.8 3.8 41
22-J.Smith BAL 90 6.3 10 59% 12 12.5 3.1 18
24-D.Revis TB 50 6.4 11 57% 22 14.2 3.3 24
21-L.McKelvin BUF 101 6.4 12 55% 32 12.5 3.0 17
21-P.Peterson ARI 84 6.4 13 57% 23 12.4 4.5 63
32-O.Scandrick DAL 91 6.6 16 60% 11 10.9 3.2 22
30-J.McCourty TEN 86 6.6 17 56% 30 12.3 2.4 7

We spent a lot of last year trying to figure out how the Giants ranked eighth in pass defense DVOA even though their once-strong pass rush had fallen to 28th in Adjusted Sack Rate. Well, here's one big reason. Trumaine McBride's charting stats for 2013 are completely surreal. Look at how far ahead of everyone else he is in Adjusted Yards per Pass. We're talking here about a player who was almost entirely a special teamer prior to last season and played only one game over the past two years. He shows up in New York this year, makes the Giants roster, enters the starting lineup for good around Week 7, and proceeds to have an incredible year of stopping opposing receivers. We went back and re-watched a bunch of Giants plays to try to figure out if we had somehow missed long gains that should have been blamed on McBride. We couldn't find any. We list McBride in coverage on only three receptions over 16 yards last season: two for 56 and 21 yards by DeSean Jackson in Week 5, and a 52-yard completion from Scott Tolzien to Jarrett Boykin in Week 11. We only list him giving up two touchdowns: a 12-yarder by Doug Baldwin in Week 15 and a six-yard gain by Danny Woodhead in Week 14 where he was probably in zone coverage and maybe shouldn't even be blamed. Despite those two touchdowns coming late in the year, McBride actually got better as the year went along; he missed Week 13 with an injury but in the final four weeks after that he improved to 3.0 Adjusted Yards per Pass with a 71 percent Adjusted Success Rate.

These stats scream "nickelback with a small sample size who you can't take seriously," except McBride was a starter on the outside for two-thirds of the year, not a nickelback who barely slid above our minimums to be ranked in the charting stats. He did cover opposition No. 3 receivers in the last couple weeks of the season, but they were No. 3 receivers who generally play outside and run deep routes, such as Jermaine Kearse and Aldrick Robinson, not undersized chain-moving slot guys. You could explain away these stats as being partly the product of the Giants playing a lot of Cover-2, where the cornerback isn't necessarily the guy "in coverage" when a receiver goes deep downfield. Except McBride himself sees the Giants defense of 2013 as the complete opposite. I was able to e-mail McBride because he had actually contacted me a few years ago when he was one of our top guys in special teams tackles. I asked him for his personal viewpoint on why he was able to have by far the best season of his career in 2013, and he responded:

There are numerous reasons why I feel last year went so much better. Coaching staff, defensive scheme, technique, offseason training,etc. Overall the Giants organization does a great job of getting the best out of players... Coach Fewell's defensive scheme is great and it fits my style of play. I'm more of a man corner vs zone. The majority of my career I was in a Cover-2 scheme... Coach Fewell has a Cover- 2 background but I wouldn't say we are a true Cover-2 defense. Perry has his own unique style.

I must admit I didn't notice big changes in Fewell's scheme from watching the Giants on television this year, but of course nobody pays much attention to coverage schemes when watching a game on live television unless they're specifically looking for that. If you want to know if the Giants were playing more man or zone concepts this season, it's hard to beat asking one of the guys who was actually playing those concepts on the field.

William Gay of the Steelers is another unheralded corner who had a great season by our charting stats last year; his fine stats are even more interesting when you consider that on the other side of the field, Ike Taylor's skills completely crumbled. Jimmy Smith really broke out this year as Baltimore's top cornerback and did very well in both stats. So did other well-regarded veterans including Patrick Peterson, Darrelle Revis, and Joe Haden.

In Atlanta, Desmond Trufant had an excellent year for a rookie cornerback. Top-drafted rookie cornerbacks almost always struggle in these stats in their first year. Revis was just average, for example, and Peterson was terrible. The good news for the Falcons is that Trufant was a glaring exception to the rule. The bad news is that the last first-round rookie to have charting stats this impressive was Devin McCourty, and whatever he was doing his rookie season didn't last. Within two years he had switched positions to safety.

Drayton Florence is the patron saint of cornerback stat oscillation: near the top of the league in 2009, near the bottom of the league in 2010 and 2011, mediocre in 2012, back to having great numbers again in 2013. He's right above the minimum to be ranked, so don't go crazy buying Drayton Florence jerseys after you read this.

Next, we'll look at the top cornerbacks in Success Rate. Success Rate, to remind everyone, is the percentage of passes that don't manage to get at least 45 percent of needed yards on first down, 60 percent of needed yards on second down, or 100 percent of needed yards on third down.

Top Cornerbacks in Adjusted Success Rate, 2013
Player Team Charted
Targets
Adj
Yd/Pass
Rk Adj
Suc Rate
Rk Avg. Pass
Distance
YAC Rk
38-T.McBride NYG 72 4.2 1 69% 1 13.3 2.2 3
37-N.Robey BUF 52 6.9 24 68% 2 10.8 3.7 37
29-D.Florence CAR 50 6.1 6 63% 3 12.7 5.2 75
45-D.Rodgers-Cromartie DEN 66 7.8 52 63% 4 14.0 7.3 86
24-J.Joseph HOU 79 6.9 25 63% 5 13.3 2.7 14
25-C.Harris DEN 81 7.3 38 62% 6 11.7 3.2 21
23-V.Davis IND 76 7.1 33 62% 7 13.6 5.1 73
20-A.Verner TEN 72 7.2 36 61% 8 13.0 3.5 31
23-A.Ball JAC 69 6.9 23 60% 9 14.2 4.3 59
41-C.Munnerlyn CAR 70 7.8 48 60% 10 12.6 3.2 20
32-O.Scandrick DAL 91 6.6 16 60% 11 10.9 3.2 22
22-J.Smith BAL 90 6.3 10 59% 12 12.5 3.1 18
23-T.Porter OAK 78 7.1 31 59% 13 9.4 4.8 69
23-J.Haden CLE 88 6.1 7 58% 15 14.6 2.5 8
24-A.Jones CIN 89 7.6 46 58% 16 12.9 5.7 81

Many of the names on this list are the same as our las t list, but we add some corners who overall played very steady but did give up a few big plays to raise their Adjusted Yards per Pass averages. Dominique Rodgers-Cromartie, for example, is getting blamed here for a 73-yard touchdown by Denarius Moore where DRC and Duke Ihenacho collided trying to tackle Moore and he ran unimpeded into the end zone. As an added bonus, we've got him down for a 79-yard touchdown by Dez Bryant which also featured a blown tackle by Ihenacho.

The really interesting listing here is one of the most appropriately named players in the NFL, Nickell Robey. Yes, he was Buffalo's nickelback, but his stats reverse the usual nickelback pattern and he ranks higher in Adjusted Success Rate than in Adjusted Yards per Pass. He did well in both, though.

At this point, most of you are probably asking one question: where the hell is Richard Sherman? Well, this is where we get into the problem I noted at the beginning of this article, where sometimes great corners have mediocre numbers because they only get thrown at when they make mistakes. Sherman was 29th with 56 percent Adjusted Success Rate and 47th with 7.7 Adjusted Yards per Pass. But before you go crazy thinking Sherman is overrated, note that we only listed him in coverage on 65 passes. That ranked 58th among corners even though Sherman started every single game for a team whose opponents were often stuck passing to try to catch up in losses. Honestly, a better way to rate a cornerback is something like what Cian Fahey does with his "The Numbers, The Tape, The Verdict" series, analyzing every play where a cornerback is in man coverage, even if the ball isn't thrown at his receiver. Unfortunately, doing this for every cornerback in the league would take way too much time to be remotely possible.

This next table is more for conversation than for any kind of evidence of which players were or were not very good this year. Allowing YAC often has more to do with the defensive scheme than with a cornerback's skill set.

Top Cornerbacks in YAC Allowed, 2013
Player Team Charted
Targets
Yd/Pass Rk Success
Rate
Rk Avg. Pass
Distance
YAC Rk
33-J.Greer NO 42 7.0 27 57% 21 13.9 1.8 1
39-B.Browner SEA 41 7.1 32 57% 17 13.7 1.8 2
38-T.McBride NYG 72 4.2 1 69% 1 13.3 2.2 3
28-K.Lewis NO 61 6.9 26 51% 50 14.2 2.2 4
25-K.Jackson HOU 76 8.5 67 42% 82 11.9 2.3 5
25-T.Brown SF 73 6.7 18 50% 59 12.2 2.4 6
30-J.McCourty TEN 86 6.6 17 56% 30 12.3 2.4 7
23-J.Haden CLE 88 6.1 7 58% 15 14.6 2.5 8
22-A.Samuel ATL 33 11.5 87 31% 88 14.5 2.6 9
32-T.Mathieu ARI 60 7.3 40 47% 71 11.0 2.6 10
22-W.Gay PIT 82 5.8 2 58% 14 11.9 2.6 11
24-B.Fletcher PHI 96 7.3 39 47% 72 13.1 2.7 12
24-J.Joseph HOU 79 6.9 25 63% 5 13.3 2.7 14
20-P.Amukamara NYG 90 6.9 22 48% 68 11.0 2.9 16
21-L.McKelvin BUF 101 6.4 12 55% 32 12.5 3.0 17

After years as one of the top corners in the league, especially by our charting stats, Asante Samuel completely imploded this year. Yet he somehow made the YAC allowed top ten despite being awful at tackling. Your guess is as good as mine.

During the 2013 season, we partnered with EA Sports to bring special Football Outsiders-branded items to Madden 25 Ultimate Team. This weekend, starting Friday, we'll be celebrating three of the top cornerbacks of the year according to our charting stats with these players available in packs:

We'll revisit these stats in a few days, looking at the players with the worst numbers in 2013.

Posted by: Aaron Schatz on 06 May 2014

11 comments, Last at 14 May 2014, 3:28pm by jpestrak

Comments

1
by JasonK :: Tue, 05/06/2014 - 5:43pm

And the Giants followed that sterling performance by McBride by... signing 2 major free agent CBs to push him to the bench as their likely 4th CB. Sure, they re-signed McBride to a more-than-minimum 2-year contract, but his $1.4M annual average pales beside Walter Thurmond's $3M and DRC's $7M. Oh, and they also exercised the 5th-year option on Prince Amukamara's contract.

I do suspect that there is a strong schematic element to these results. Maybe not specifically "cover-2," but I'd be shocked if the majority of the deep-safety help wasn't shifted to his side of the field. He also was on the bench for the beginning of the year when the Giants' back-seven was at its most shaky (i.e., before the Beason trade and Will Hill's move from suspension into the starting lineup).

2
by jfsh :: Tue, 05/06/2014 - 6:08pm

Is it possible to use snap count data to get success rate or yards allowed per snap? That may be a way to show the value of a cornerback who rarely gets targeted.

3
by theslothook :: Tue, 05/06/2014 - 7:39pm

We all know safety play matters a ton, but I would really like to see a scouting or charting report about how its changed. Anecdotally, we all know Earl Thomas is a great deep safety, but what does the interaction of chancellor and Thomas together mean for the overall defense. What does having them as a tandem allow?

7
by Karl Cuba :: Wed, 05/07/2014 - 8:46pm

It allows Chancellor to play closer to the line, which he does at an all-star level, whereas if he was being asked to play half the field in a cover 2 zone he'd get burned quite a bit because he's so big. He probably wouldn't be wretched, he'd make some big hits but he'd get exposed by fast receivers and give up more big plays than you'd like. In the box though, he's a beast.

4
by Vince Verhei :: Wed, 05/07/2014 - 3:09pm

Some snap count-style stats (minimum 50 targets or eight starts):

TARGETS PER SNAP

Highest:
1. D.Milliner, NYJ: 13.3%
2. D.Slay, DET: 13.3%
3. M.Cooper, KC: 13.0%
4. B.Boykin, PHI: 12.5%
5. T.McBride, NYG: 12.0% (Yes, he had the best rate stats last year despite being picked on a lot.)
6. X.Rhodes, MIN: 11.7%
7. K.Webster, DEN: 11.7%
8. D.House, GB: 11.0%
9. L.McKelvin, BUF: 10.9%
10. T.Brock, SF: 10.9%

Lowest:
86. D.Revis, TB: 5.1%
85. J.Wilson, WAS: 6.1%
84. C.Munnerlyn, CAR: 6.3%
83. A.Williams, BUF: 6.4%
82. T.Williams, GB: 6.7%
81. R.Sherman, SEA: 6.7%
80. K.Lewis, NO: 6.8%
79. A.Ball, JAC: 6.9%
78. J.Robinson, MIN: 6.9%
77. J.Banks, TB: 6.9%

***

ADJUSTED FAILURES (DPIs or completions that gained successful yardage) PER SNAP

Highest:
1. D.Slay, DET: 6.9%
2. S.Wright, SD: 6.5%
3. M.Cooper, KC: 6.0%
4. D.Milliner, NYJ: 6.0%
5. I.Taylor, PIT: 5.9%
6. R.McClain, ATL: 5.8%
7. K.Jackson, HOU: 5.8%
8. D.House, GB: 5.7%
9. D.Cox, SD: 5.7%
10. X.Rhodes, MIN: 5.6%

Lowest:
86. D.Revis, TB: 2.2%
85. C.Munnerlyn, CAR: 2.4%
84. A.Ball, JAC: 2.6%
83. A.Verner, TEN: 2.6%
82. T.Williams, GB: 2.7%
81. N.Robey, BUF: 2.8%
80. C.Harris, DEN: 2.9%
79. D.Florence, CAR: 3.0%
78. A.Williams, BUF: 3.0%
77. R.Sherman, SEA: 3.1%

***

ADJUSTED YARDS ALLOWED PER SNAP

Highest:
1. D.Slay, DET: 1.337
2. M.Cooper, KC: 1.185
3. C.Houston, DET: 1.135
4. S.Wright, SD: 1.063
5. R.McClain, ATL: 1.041
6. K.Webster, DEN: 0.993
7. D.Cox, SD: 0.993
8. B.Boykin, PHI: 0.990
9. D.Milliner, NYJ: 0.980
10. I.Taylor, PIT: 0.957

Lowest (listing 20 players because so many interesting names just missed the top 10):
86. D.Revis, TB: 0.323
85. A.Ball, JAC: 0.456
84. M.White, CAR: 0.468
83. C.Owens, CLE: 0.471
82. W.Gay, PIT: 0.474
81. A.Verner, TEN: 0.478
80. C.Munnerlyn, CAR: 0.484
79. T.McBride, NYG: 0.489 (One of the ten highest rates of targets per snap, one of the ten lowest in yards allowed per snap. Insane.)
78. A.Williams, BUF: 0.489
77. D.Florence, CAR: 0.489
76. K.Lewis, NO: 0.491
75. P.Peterson, ARI: 0.498
74. R.Sherman, SEA: 0.516
73. J.Haden, CLE: 0.520
72. D.Trufant, ATL: 0.521
71. J.Smith, BAL: 0.523
70. T.Mathieu, ARI: 0.532
69. L.Ryan, NE: 0.534
68. J.McCourty, TEN: 0.538
67. O.Scandrick, DAL: 0.539

These still aren't perfect stats. They don't control for how often opposing offenses ran or passed, pass rush, or scheme. For example, as someone who charted a lot of Carolina games, I can tell you that their CBs ranked so high in part because they played zone coverage behind a monstrous pass rush that forced a lot of dumpoffs and checkdowns to uncovered receivers.

5
by Karl Cuba :: Wed, 05/07/2014 - 3:39pm

Thanks for doing this Vince.

8
by ChrisS :: Thu, 05/08/2014 - 12:49pm

I agree thanks, this makes the numbers a bit more intuitive (a Lion is at the top of the bad list). As you say doing it on pass plays only and pass plays with various numbers of players rushing would be awesome.

9
by zlionsfan :: Thu, 05/08/2014 - 1:27pm

OTOH, as someone who charted a lot of Lions games, I can say that Slay and Houston made these lists because when your secondary isn't that good to begin with, if you have enough injuries, whoever's left is going to be worse ... or not close to 100% but pressed into service because someone's got to play.

And in a league that thrives on passing, if you aren't good enough for whatever reason, you'll be thrown at again and again and again. Couple that with a toothless pass rush, and the only thing stopping other teams from throwing at Slay and Houston and whoever else was starting even more often was the end zone.

6
by Pen :: Wed, 05/07/2014 - 8:10pm

Theslothook, to answer your question, Earl's ability to reach either end of the field as a center fielder allows Chancellor to play up closer to the line and fill any open gaps. This means the Seahawks can play sets that other teams simply couldn't get away with. If the play is a run, Chancellor can plug the gap, contributing to the Hawks great run defense, if the play is a pass, Kam falls back into coverage and delivers those brutal hits to anyone coming over the middle. Meanwhile, the Hawks can play their version of cover 3 which isn't, but kinda is, what they play. Earl's primary task is to back up Maxwell and leave Sherman on an island. Earl takes away the deep routes, but has the speed and ball skills to even come up and stuff a screen pass! The LBers and Kam fall into a zone, the CB's play tight man and ET reads the play and reacts.

10
by Mike B. In Va :: Thu, 05/08/2014 - 1:46pm

As the resident Bills fan, I'm very happy to see the numbers back up what my eyes told me about McKelvin. He really turned a corner (HA!) this year.

11
by jpestrak :: Wed, 05/14/2014 - 3:28pm

Does anyone have a cached copy of the 'The Numbers, The Tape, The Verdict' URL linked above?

I receive a 'temporarily unavailable' error when I try to load the page.