DVOA by Route: Cracks in the Steel Curtain

Tennessee Titans WR A.J. Brown
Tennessee Titans WR A.J. Brown
Photo: USA Today Sports Images

The Pittsburgh Steelers boasted the league's best pass defense last season. With Pro Bowlers at interior lineman (Cameron Heyward), edge rusher (T.J. Watt), and safety (Minkah Fitzpatrick), they gave opponents fits all year. None of those Pro Bowlers played cornerback, however, and on the rare occasions when Pittsburgh did struggle, it was usually on deeper throws to the outside.

Here's a look at Pittsburgh's performance against the most common routes across the NFL last season. The table is sorted by DVOA, and rankings go from low to high—the Steelers rank first in flat targets because they only saw 16 of them all year, the lowest total in the league.

PIT Defense vs. Most Common Routes, 2020
Route DVOA Rk Tgt Rk Avg. Rk
WR/TE screen -53.6% 3 24 4 3.7 1
Flat -45.6% 3 16 1 3.8 4
Dig -35.0% 8 38 22 5.8 1
Drag -35.0% 9 22 11 5.8 10
Out -34.0% 4 76 31 5.8 8
Broken Play -30.3% 7 28 16 8.4 26
Deep Cross -22.2% 5 6 2 6.0 4
Curl -22.1% 1 83 22 6.5 7
Fade 0.6% 16 9 7 8.2 14
Post 5.2% 13 22 31 9.2 10
Slant 30.6% 28 44 21 10.3 31
Go/Fly 40.6% 17 21 30 11.7 10
Corner 91.3% 30 13 14 14.5 28

The Steelers ranked in the top 10 in DVOA against eight of the most common routes in the league—there aren't a lot of weaknesses here. They were about average against fades and posts, though it's worth noting that opponents threw 22 post routes against Pittsburgh, more than anyone except Atlanta. It didn't always work, but it's telling that the Baker Mayfields and Lamar Jacksons of the world repeatedly tested the deep middle of Pittsburgh's secondary.

The real problems for Pittsburgh, though, came on throws to outside receivers—they had a DVOA of 30.0% or worse against slant, go/fly, and corner routes. The corner, in particular, was a bugaboo. Opponents completed seven of 11 corner passes against the Steelers for 153 yards; all seven of those completions picked up first downs, including two touchdowns. Two other corner throws produced DPIs of 9 and 26 yards. What's most shocking is how some of the league's worst quarterbacks burned Pittsburgh on corners. The trio of Jeff Driskel, Carson Wentz, and Alex Smith tried five corners against Pittsburgh, picking up a first down on every throw (including one touchdown) for a combined total of 135 yards.

The best way to improve performance against corners is to improve the caliber of your corners—cornerbacks, that is. The Steelers, however, did the opposite, losing starter Steven Nelson and nickelback Mike Hilton (a playmaker extraordinaire) in free agency. Your top cornerbacks in Heinz Field this season: Joe Haden (who turned 32 in April), Cameron Sutton (a 2017 third-round pick with eight starts in his first four seasons), and Arthur Maulet (now playing for his fourth team in the past four years). After Pittsburgh's offense cratered at the end of last season, Steelers fans are rightfully concerned about Ben Roethlisberger and his rapidly fading arm, but their team has concerns on defense this season too.

The numbers in the following tables represent DVOA and other passing splits against the targeted routes. As such, pass-related outcomes such as sacks, scrambles, and throwaways are excluded since they don't result in targets. We're only looking at the 13 routes that saw at least 400 targets league-wide. All DVOA figures are calculated against a baseline of only pass attempts, and since this evaluation comes from the defensive perspective, negative DVOA totals are better than positive ones. These DVOA ratings are based on the same DVOA baselines as our "Defense vs. Receivers" tables on FO+, which means that A) ratings are adjusted for the quality of the offenses faced, based on receiver position, and B) interceptions and fumbles both count as negative plays in these DVOA ratings.

Curl

This is the most basic route in football, so defenses had better be prepared for it. And no team played better against the curl than the Pittsburgh Steelers.

Defense vs. Curl Routes, 2020
Team DVOA Tgt Yds Catch% aDOT YAC
PIT -22.1% 83 536 63% 7.5 3.0
IND -19.3% 88 511 72% 5.8 2.6
NE -15.2% 57 447 77% 7.2 3.2
TEN -13.9% 80 529 75% 6.6 2.5
NO -12.7% 42 314 73% 8.4 2.3
SF -10.1% 68 437 76% 6.5 2.3
LV -9.9% 66 507 77% 6.9 2.8
DEN -7.1% 80 515 72% 7.0 2.0
MIA -6.9% 74 550 75% 6.1 4.0
GB -6.1% 90 525 72% 5.7 3.3
LAR -1.9% 101 608 68% 6.0 2.7
CIN -1.5% 68 482 75% 7.1 2.4
NYG 0.1% 91 734 78% 7.9 3.1
BUF 0.1% 92 581 69% 6.5 2.9
CAR 2.7% 103 698 73% 7.3 2.2
CLE 2.7% 79 588 80% 7.1 2.4
BAL 2.9% 61 439 80% 6.6 3.1
PHI 5.7% 65 502 77% 7.0 3.4
LAC 6.9% 90 589 78% 6.6 2.2
ATL 7.0% 83 586 71% 6.5 3.6
WAS 8.0% 65 443 69% 7.0 3.2
TB 8.5% 82 596 79% 6.6 2.7
MIN 9.4% 41 303 78% 7.0 2.7
JAX 11.0% 54 380 85% 6.6 1.9
SEA 13.1% 120 1,002 73% 7.7 4.1
DET 15.4% 66 578 78% 7.7 3.5
HOU 21.2% 73 530 80% 6.6 2.8
ARI 21.3% 64 492 80% 6.6 3.3
KC 24.4% 51 437 82% 5.9 5.0
NYJ 26.5% 119 1,004 83% 7.0 3.5
CHI 28.6% 66 573 84% 7.9 2.4
DAL 34.8% 41 314 85% 6.7 2.6
NFL 3.4% 2,403 17,330 76% 6.8 2.9

The Dallas Cowboys, on the other hand, were the league's worst defense against curls … or were they? They gave up a 34.8% DVOA on curl targets, highest in the league, but then they only faced 41 curls all year, tied for fewest, and a little more than half as many as the average team. Is it more important to perform well against a certain route, or to make sure that route is never thrown in the first place? That's a complicated question. For now we'll just note that for the second year in a row the Seattle Seahawks saw the most curls at 120. That's partly because the Seahawks set a record for most pass attempts allowed in 2020, and also because they play more zone coverage out of base personnel than almost any other team. It's a vanilla scheme that allows a bunch of short completions, and opponents were all too happy to take them.

Out

If you're going to defend the out route, you're going to need to need a quality cornerback. Miami has Xavien Howard, New England has Stephon Gilmore, and New Orleans has Marshon Lattimore, which partly explains why the Dolphins, Patriots, and Saints took gold, silver, and bronze in DVOA against the out.

Defense vs. Out Routes, 2020
Team DVOA Tgt Yds Catch% aDOT YAC
MIA -41.7% 45 245 58% 7.4 2.8
NE -40.2% 40 285 77% 6.9 2.4
NO -34.6% 71 342 54% 7.5 2.3
PIT -34.0% 76 444 53% 7.9 2.6
LAR -30.6% 51 250 71% 5.8 1.6
LAC -29.4% 47 337 62% 8.6 3.1
IND -28.7% 47 231 57% 7.9 1.0
JAX -25.3% 56 328 69% 7.1 2.5
MIN -24.8% 82 493 67% 7.6 2.1
SF -21.5% 60 368 68% 7.0 2.1
BUF -19.9% 47 244 67% 6.2 2.2
KC -19.7% 59 332 64% 6.1 3.2
GB -18.3% 64 429 62% 8.1 2.5
CLE -14.7% 70 370 57% 7.6 2.1
PHI -14.7% 41 240 66% 7.9 1.8
CAR -10.8% 52 375 71% 8.6 2.3
WAS -5.3% 52 364 73% 7.3 2.9
TB -2.0% 64 417 75% 6.7 2.8
BAL 1.2% 72 519 74% 7.0 2.5
HOU 8.6% 48 366 66% 9.2 2.1
CHI 9.7% 53 390 71% 7.3 2.8
TEN 9.8% 69 504 78% 6.8 2.7
ATL 10.7% 59 503 77% 7.5 3.8
DAL 10.9% 74 504 74% 7.0 2.4
NYJ 13.6% 60 460 78% 8.1 2.6
DET 16.7% 50 389 70% 8.4 2.8
NYG 17.1% 56 431 70% 9.1 2.3
SEA 18.6% 71 524 80% 7.5 1.9
DEN 19.1% 71 540 72% 7.8 3.6
LV 30.5% 46 395 85% 7.2 2.9
ARI 32.0% 58 465 75% 6.8 3.9
CIN 42.6% 66 613 81% 8.3 3.3
NFL -4.6% 1,877 12,697 69% 7.5 2.6

By contrast, the Bengals were starting LeShaun Sims, who rarely came off the bench for the Titans; the Cardinals were starting Patrick Peterson, whose shelf life may have run out; and the Raiders were starting Nevin Lawson, who is both aging and should be coming off the bench. That's largely why those three teams were so much worse against the out than anyone else.

The Vikings saw the most outs in the league, 82, though their DVOA against the play was better than most. The Eagles saw the fewest, 41; their DVOA was perfectly mediocre.

Slant

Remember when we mentioned New Orleans had a stud corner in Marshon Lattimore? Turns out that stud corners are good at defending slants too.

Defense vs. Slant Routes, 2020
Team DVOA Tgt Yds Catch% aDOT YAC
NO -48.9% 49 307 57% 5.9 4.1
BAL -38.5% 58 319 57% 6.1 3.7
MIN -38.3% 42 309 68% 7.2 3.4
CLE -33.2% 35 266 66% 6.2 5.3
NYJ -20.1% 35 249 57% 7.0 5.2
KC -13.5% 51 359 66% 6.5 4.8
CAR -11.6% 24 140 54% 7.0 4.2
LV -8.7% 39 317 61% 6.3 7.3
DEN -6.7% 35 249 66% 6.3 4.4
ARI -6.6% 43 337 64% 7.3 4.9
TEN -5.9% 42 265 58% 6.6 3.3
WAS -4.1% 43 289 63% 7.2 4.1
CIN -2.8% 47 284 64% 6.2 3.3
ATL -1.2% 48 370 74% 6.9 3.3
IND 0.9% 46 379 61% 7.6 6.3
LAC 1.4% 31 242 65% 7.1 5.0
SF 2.3% 34 193 62% 5.7 3.4
TB 2.5% 45 318 76% 6.6 2.9
BUF 4.5% 53 374 67% 7.3 3.0
DAL 7.1% 34 208 56% 6.8 3.7
PHI 9.8% 55 506 71% 7.2 5.5
HOU 9.8% 45 316 64% 6.9 4.1
CHI 13.8% 44 334 74% 6.6 3.5
MIA 19.9% 32 302 77% 7.9 4.3
NE 23.7% 18 219 72% 6.9 9.2
JAX 25.0% 35 342 76% 6.9 6.0
LAR 25.7% 26 212 65% 6.5 5.8
PIT 30.6% 44 453 74% 7.0 6.6
NYG 34.5% 26 214 58% 6.9 6.3
GB 34.7% 42 370 71% 6.2 5.6
DET 35.7% 27 276 83% 7.2 5.3
SEA 38.5% 55 467 70% 7.7 4.4
NFL 0.8% 1,283 9,785 66% 6.8 4.6

There's a pretty big gap here between the Saints and the next three teams in DVOA, the Ravens, Vikings, and Browns. The Patriots also deserve notice for allowing only 18 slant targets all season, another effect of having a stud corner such as Stephon Gilmore.

You know who didn't have any stud corners last year? The Seahawks. They allowed 55 slant targets, second only to Baltimore, and they gave up a league-worst DVOA of 38.5% on those throws. The Jaguars saw this performance and decided that guaranteeing Shaquill Griffin nearly $30 million was a wise use of their resources.

Flat

We have talked about cornerbacks a lot today, but coverage against flat routes is more typically the responsibility of a team's linebackers. It's surprising, then, that the Green Bay Packers had the best DVOA against flat routes in 2020.

Defense vs. Flat Routes, 2020
Team DVOA Tgt Yds Catch% aDOT YAC
GB -68.2% 26 138 85% 1.5 4.9
BAL -57.5% 22 48 59% 0.0 3.9
PIT -45.6% 16 61 69% 1.2 4.5
WAS -42.8% 33 122 61% 1.7 4.3
BUF -33.8% 31 135 68% 1.2 5.1
NYJ -24.2% 33 155 85% 1.3 4.3
ARI -19.0% 27 98 70% 1.6 3.4
MIA -11.1% 29 162 72% 1.7 6.1
LAC -9.9% 40 168 75% 1.2 4.3
CHI -8.9% 29 138 93% 1.4 3.6
TB -6.9% 39 205 72% 1.5 5.6
HOU -6.9% 34 208 82% 0.8 6.5
SEA -6.4% 40 167 73% 0.9 4.8
NYG -2.7% 28 140 71% 0.5 6.4
ATL -1.1% 50 252 86% 0.9 4.9
SF -1.0% 35 149 86% 1.0 4.0
TEN -0.1% 32 169 88% 1.3 4.6
CAR 0.9% 30 122 73% 1.6 3.7
NO 2.3% 43 166 86% 1.3 3.3
KC 3.1% 42 259 86% 1.3 5.9
MIN 4.9% 37 173 86% 1.7 3.8
DAL 5.1% 46 213 76% 1.0 5.2
PHI 7.1% 33 166 85% 2.0 4.1
NE 7.2% 29 193 97% 1.1 5.7
LV 11.8% 45 283 84% 1.7 5.6
DEN 13.1% 40 244 93% 1.3 5.1
CIN 15.1% 34 200 91% 1.3 5.3
JAX 19.1% 25 171 80% 1.2 7.4
LAR 19.9% 34 194 82% 1.0 6.1
IND 20.2% 23 129 91% 1.9 4.3
CLE 21.8% 34 230 91% 1.0 6.5
DET 36.3% 37 246 86% 1.9 5.7
NFL -2.6% 1,076 5,504 81% 1.3 5.0

Christian Kirksey failed to make the top 30 linebackers in either success rate in coverage or yards allowed per target. Krys Barnes failed to make the top 50. So how was Green Bay so effective against flats? With defensive backs. Nine of the flats thrown against the Packers had no defender listed in coverage, because if you're going to leave a guy uncovered, he'd better be near the line of scrimmage. Of the 17 others, the primary defender was listed as a linebacker only three times (and that includes one target for edge rusher Preston Smith). The others all went to defensive backs, including four by starting corner Chandon Sullivan and three by Pro Bowler Jaire Alexander. A lot of offenses try to get their running backs matched up against linebackers in coverage, but the Packers rarely let that happen.

The Lions had the league's worst DVOA against flats, because they were the Lions and they were terrible at most everything. The Falcons saw the most flat targets with 50; the Steelers saw the fewest at just 16, six fewer than anyone else.

Dig

There are a lot of reasons why Tampa Bay won the Super Bowl last year, but a big one is that the Buccaneers' DVOA against digs was so much better than anyone else's.

Defense vs. Dig Routes, 2020
Team DVOA Tgt Yds Catch% aDOT YAC
TB -83.7% 36 250 47% 12.1 3.6
DEN -61.7% 39 284 51% 9.8 4.1
MIA -58.1% 25 198 52% 11.2 5.3
CHI -49.1% 44 348 56% 9.1 4.8
TEN -41.6% 38 257 58% 9.3 2.1
MIN -39.8% 34 293 59% 10.5 5.4
IND -35.2% 41 306 63% 10.1 2.0
PIT -35.0% 38 221 46% 9.4 3.4
CIN -34.5% 22 172 59% 8.4 4.8
BUF -34.3% 38 262 49% 10.6 4.3
DAL -28.6% 27 214 63% 9.1 3.6
NYG -28.4% 32 249 59% 11.4 2.7
NO -12.6% 37 343 70% 9.6 2.7
BAL -2.4% 38 282 65% 7.8 3.7
CLE -2.1% 27 182 59% 10.1 2.8
WAS 3.5% 41 265 58% 8.8 2.3
LAC 7.7% 30 240 57% 10.3 4.0
CAR 9.9% 38 323 61% 11.2 2.3
GB 10.5% 43 370 63% 9.4 4.2
LAR 10.7% 36 319 69% 9.4 3.2
KC 12.2% 31 241 71% 8.6 2.6
SEA 17.2% 42 357 60% 11.0 4.3
HOU 19.5% 28 280 61% 12.0 4.1
JAX 24.7% 44 406 77% 10.0 3.1
LV 28.3% 35 327 61% 11.3 4.0
ARI 30.2% 23 193 64% 10.0 2.8
ATL 35.8% 29 305 69% 12.9 2.1
SF 37.1% 18 190 67% 10.3 4.9
NE 40.9% 30 348 67% 12.4 5.8
PHI 43.6% 19 225 74% 11.7 3.5
DET 58.7% 36 463 81% 11.6 4.8
NYJ 82.1% 32 386 75% 11.2 4.4
NFL -4.5% 1,071 9,099 62% 10.3 3.6

The Steelers' numbers in this table are very similar to the Buccaneers, but turnovers made the difference. Pittsburgh forced just one turnover on digs, an interception of Jeff Driskel. The Bucs forced six, five interceptions and a fumble. The Jets, meanwhile, finished last by a mile in DVOA against digs, in part because they forced only one turnover on the play (a fumble by Henry Ruggs of the Raiders).

The Bears and Jaguars both saw 44 dig targets, tied for most in the league. The 49ers saw the fewest, only 18.

WR/TE Screen

Screens to wide receivers and tight ends are terrible plays in the NFL, and so most defenses had great numbers on those routes. None of them, however, had any better numbers than the Saints.

Defense vs. WR/TE Screen Routes, 2020
Team DVOA Tgt Yds Catch% aDOT YAC
NO -59.0% 22 84 91% -2.1 6.5
CHI -53.7% 23 172 83% -2.7 11.8
PIT -53.6% 24 89 92% -2.7 6.7
LAR -48.5% 36 159 92% -2.3 7.2
TEN -44.8% 28 152 100% -1.5 6.9
CIN -39.0% 26 153 88% -1.8 8.8
BUF -37.7% 30 191 83% -2.3 10.0
NYJ -36.5% 29 109 86% -2.6 7.0
IND -35.0% 33 150 91% -1.4 6.5
WAS -33.3% 50 217 94% -2.0 6.5
ATL -31.4% 40 240 98% -1.7 7.9
DET -30.7% 27 156 81% -2.3 9.2
LV -29.6% 30 195 83% -2.4 10.1
CAR -29.3% 41 232 93% -2.1 8.2
SEA -28.0% 43 187 95% -2.3 6.9
MIN -27.9% 27 156 100% -2.7 8.4
LAC -25.9% 26 135 85% -1.8 8.0
DEN -19.7% 36 199 94% -1.6 7.4
CLE -18.0% 33 189 91% -1.9 8.4
KC -16.6% 28 151 96% -2.3 7.9
JAX -16.1% 28 225 93% -2.4 11.0
SF -15.7% 46 219 96% -1.7 6.7
GB -13.3% 31 190 90% -2.4 9.1
NYG -10.2% 36 183 97% -2.4 7.6
TB -7.3% 38 256 92% -2.0 9.5
MIA -6.2% 34 330 85% -1.3 13.1
HOU -3.6% 28 182 86% -1.6 9.2
DAL 0.6% 21 138 90% -1.6 8.8
NE 8.4% 29 213 97% -1.7 9.4
ARI 10.3% 38 307 92% -2.3 11.2
PHI 12.9% 42 317 95% -1.1 9.2
BAL 28.8% 39 287 95% -1.7 9.5
NFL -21.0% 1,042 6,163 92% -2.0 8.5

So, yeah, congratulations to New Orleans for doing what pro defenses are supposed to do and shutting screens down, but I'm more intrigued by the team at the other end of the spectrum. How on earth did the Ravens give up a 28.8% DVOA to WR/TE screens? They can blame a lot of that on the Colts and Football Team. In two games against Indianapolis, and Washington, the Ravens gave up seven WR/TE screens that gained at least 10 yards. That includes a 40-yard gain by Antonio Gibson on third-and-11 and a 15-yard gain by Zach Pascal on fourth-and-6. There were 19 teams that didn't give up that many 10-yard gains on WR/TE screens all season. (The Ravens gave up 10 such plays in all; the Cardinals gave up the most with 14.)

Washington faced 50 WR/TE screens, most in the league, mainly against Philadelphia, San Francisco, Pittsburgh, and Cincinnati. New Orleans saw the fewest, only 22.

Broken Play

There were a lot more of these across the league than in the season before, and the 49ers saw plenty of them, sharing a division with both Russell Wilson and Kyler Murray. So it's a good thing they had the league's best defense on broken plays.

Defense vs. Broken Play Routes, 2020
Team DVOA Tgt Yds Catch% aDOT YAC
SF -53.6% 32 184 45% 12.6 2.6
GB -47.6% 17 93 47% 8.4 1.0
NYG -42.3% 42 276 52% 8.7 4.0
WAS -40.8% 33 168 39% 11.2 2.1
JAX -37.6% 17 126 65% 12.5 2.4
HOU -37.3% 20 82 37% 12.0 4.9
PIT -30.3% 28 234 59% 14.8 4.8
CIN -28.8% 26 206 52% 10.0 5.5
NE -28.2% 43 192 49% 10.0 2.4
DET -26.4% 33 241 58% 10.8 3.4
NO -24.2% 25 200 40% 15.9 4.0
LAC -23.9% 22 167 45% 12.7 2.2
NYJ -22.3% 37 225 51% 11.1 2.9
ARI -22.0% 27 163 32% 15.3 2.4
SEA -20.8% 33 240 52% 11.8 1.2
DAL -19.9% 37 299 56% 13.2 2.9
TB -19.5% 26 206 62% 8.2 4.2
LAR -19.5% 27 146 42% 9.6 5.9
BUF -15.5% 31 356 50% 15.4 3.2
PHI -13.9% 32 178 50% 11.6 3.6
BAL -12.0% 36 263 60% 9.4 2.8
LV -10.3% 37 220 49% 12.9 1.1
CHI -10.3% 22 118 45% 12.7 1.9
DEN -7.1% 26 194 54% 11.5 3.3
MIN -5.1% 23 182 61% 12.4 4.4
CAR -0.8% 23 210 57% 11.4 6.6
MIA -0.5% 35 304 49% 11.6 7.4
KC -0.1% 25 176 52% 9.4 7.0
IND 5.2% 33 292 55% 11.4 3.7
CLE 18.4% 27 238 56% 12.4 4.3
TEN 30.8% 38 287 66% 10.6 3.0
ATL 45.0% 30 287 57% 13.2 3.8
NFL -15.8% 943 6,753 52% 11.7 3.6

The Falcons gave up the worst DVOA on broken plays. In Week 17 alone, Tom Brady (who is very old) threw 25- and 30-yard touchdowns to Antonio Brown on broken plays against the Falcons.

Totals on broken plays are funny, because these are plays where the defense chased the quarterback out of the pocket. That's a good thing! You want that to happen as often as possible! And the average DVOA on these throws backs it up. So congratulations to the New England Patriots for facing a league-high 43 broken-play passes, and better luck next year to the Green Bay Packers for facing a league-low 17.

Drag

There were four defenses that allowed a DVOA of -40.0% or lower. Three of them won a combined total of nine games (plus a tie). The fourth went 13-3 and reached the AFC Championship Game.

Defense vs. Drag Routes, 2020
Team DVOA Tgt Yds Catch% aDOT YAC
JAX -54.4% 21 110 67% 4.1 4.2
BUF -52.7% 25 114 56% 3.0 4.6
ATL -52.6% 27 120 63% 3.3 3.8
CIN -43.4% 25 104 54% 3.2 4.7
MIA -39.1% 27 153 78% 4.0 3.7
SF -39.0% 25 149 72% 3.4 4.4
NE -36.7% 30 228 77% 3.7 6.3
DEN -36.3% 20 88 60% 3.3 3.6
PIT -35.0% 22 127 59% 4.2 5.6
DAL -33.3% 21 192 71% 2.7 9.7
TB -26.9% 27 163 85% 3.2 4.0
KC -25.9% 24 106 58% 2.8 4.8
CLE -20.1% 20 180 65% 3.9 9.9
LV -18.2% 31 159 65% 2.8 5.6
BAL -14.1% 31 204 74% 2.3 6.6
NYJ -3.9% 17 122 71% 3.2 6.5
LAR -3.7% 23 120 74% 3.6 3.2
PHI 2.6% 19 135 79% 4.0 5.3
IND 7.8% 24 152 67% 4.0 5.8
WAS 9.3% 13 77 85% 3.7 3.5
CAR 9.4% 23 149 74% 3.3 5.4
LAC 11.0% 25 183 75% 4.3 6.0
DET 11.7% 49 337 71% 4.1 6.2
SEA 14.5% 30 206 83% 2.9 5.2
NO 15.3% 21 161 75% 4.1 6.5
NYG 16.9% 30 233 87% 3.6 5.4
ARI 18.7% 40 319 83% 3.1 6.5
CHI 26.6% 23 182 78% 3.7 5.8
TEN 27.7% 30 235 83% 4.0 6.0
GB 33.6% 17 154 82% 4.1 6.7
HOU 38.0% 38 276 84% 2.8 5.8
MIN 67.7% 19 214 84% 3.6 9.6
NFL -6.2% 817 5,452 73% 3.5 5.7

The Jaguars did not allow a single gain of more than 20 yards on a drag route. They gave up only two touchdowns on the play (a pair of 4-yard scores to Tennessee and Baltimore) while intercepting Aaron Rodgers in their near-upset of Green Bay, one of five interceptions Rodgers threw all season.

And then you have the Vikings. They gave up three touchdowns on drags, including a 41-yarder to Carolina that was the best drag play of the season and a 35-yarder to Hayden Hurst of the Falcons.

The Lions saw a league-high 49 drag routes in 2020. Their division rivals in Green Bay saw only 17. That was tied for the lowest total in the NFL along with the New York Jets. And speaking of the Jets…

Deep Cross

There are not a lot of things you can point to and say "the Jets did that better than anyone else in 2020." But defense against the deep cross is one of those things.

Defense vs. Deep Cross Routes, 2020
Team DVOA Tgt Yds Catch% aDOT YAC
NYJ -211.1% 7 40 29% 12.0 3.5
WAS -91.3% 6 45 40% 16.0 0.5
ARI -55.6% 27 152 37% 12.4 3.4
GB -52.1% 18 156 44% 18.1 3.0
PIT -22.2% 6 36 33% 14.5 2.5
CHI -21.8% 17 91 41% 13.4 1.9
BAL -20.7% 22 244 55% 15.5 3.9
CAR -14.7% 16 193 50% 16.8 6.0
DET -8.2% 18 206 67% 14.2 3.3
JAX -2.7% 33 427 58% 15.7 5.3
ATL -0.4% 29 322 59% 17.8 3.3
CIN 5.8% 13 110 54% 17.2 3.3
KC 8.0% 21 289 53% 19.1 6.7
NO 12.2% 21 238 52% 16.0 5.8
SEA 13.8% 22 294 64% 17.1 3.1
MIA 18.3% 47 540 60% 16.4 4.6
PHI 24.0% 23 236 65% 11.4 3.8
SF 27.2% 11 146 45% 20.8 6.0
LAC 40.2% 24 277 54% 19.0 3.4
DEN 42.1% 23 348 61% 17.2 6.4
LV 42.4% 32 402 63% 15.6 3.0
CLE 43.1% 22 287 59% 18.2 5.8
NE 50.2% 29 345 62% 14.3 4.8
MIN 62.6% 14 224 64% 16.4 8.1
DAL 63.0% 18 252 59% 15.9 6.6
TEN 65.6% 29 395 61% 17.3 4.8
BUF 72.2% 21 270 65% 14.5 5.4
IND 87.3% 30 500 83% 16.5 3.1
HOU 88.3% 23 359 81% 13.7 5.2
LAR 111.3% 12 221 75% 18.8 7.2
TB 131.0% 14 279 79% 17.4 6.0
NYG 190.2% 12 233 92% 13.9 7.6
NFL 26.9% 660 8,157 60% 16.0 4.7

First of all, New York only allowed seven deep cross passes all year, fewer than all but two other teams (Washington and Pittsburgh allowed a half-dozen each). Second, they intercepted two of those passes—or, one for every completion they allowed. That's a DVOA of -211.1% that lapped the field. Obviously, the tiny sample size is part of that, but that's almost unavoidable at this stage of the game.

While the Jets were almost impossibly good against deep crosses, their roommates at MetLife Stadium were almost impossibly bad. Opposing quarterbacks threw 12 deep crosses against the Giants, completing 11 of them for 233 yards and four, count 'em, four touchdowns. That's how you get a DVOA of 190.2%.

The Dolphins' DVOA against deep passes wasn't nearly that bad—it was right in the middle of the pack, in fact—but they were still oddly vulnerable to the play. They faced 47 deep crosses in all, 13 more than anyone else and more than twice as many as the average team.

Go/Fly

Unlike the Jets, the Rams defense did quite a few things better than anyone else in 2020. But coverage against go/fly routes may have been the thing they did best of all.

Defense vs. Go/Fly Routes, 2020
Team DVOA Tgt Yds Catch% aDOT YAC
LAR -105.5% 18 67 6% 39.9 0.0
TEN -47.2% 15 134 29% 28.5 1.5
NE -37.1% 12 127 18% 36.8 15.0
CIN -35.3% 19 158 22% 32.3 4.0
KC -28.4% 21 232 25% 29.4 12.4
BAL -26.8% 11 83 11% 32.2 18.0
NYG -26.7% 9 125 33% 33.4 5.0
ATL -2.4% 24 234 25% 31.0 4.8
MIA 1.1% 19 266 29% 31.6 9.8
BUF 16.2% 12 140 27% 29.2 2.7
ARI 22.6% 19 210 29% 27.2 3.8
GB 24.2% 11 134 30% 35.0 3.0
MIN 25.6% 12 201 36% 36.2 5.8
NYJ 34.9% 22 320 26% 33.5 4.4
CAR 36.0% 11 140 30% 30.4 9.7
DET 37.1% 16 247 36% 33.9 7.4
PIT 40.6% 21 246 33% 30.1 13.6
PHI 41.2% 7 98 40% 26.7 8.0
SEA 42.8% 12 173 33% 35.0 5.5
HOU 47.8% 12 219 33% 34.0 15.5
CHI 48.9% 15 263 36% 36.7 13.0
WAS 51.3% 13 217 33% 34.9 8.3
LV 56.1% 17 259 40% 30.3 1.7
NO 71.0% 15 271 33% 38.5 7.8
CLE 71.5% 16 265 40% 30.1 7.7
IND 72.3% 10 155 29% 28.8 0.0
LAC 76.9% 19 330 39% 33.9 2.1
DAL 105.8% 14 252 42% 35.1 5.2
DEN 108.4% 14 316 46% 37.6 10.8
TB 134.1% 9 222 56% 36.3 12.2
JAX 136.2% 19 530 53% 31.9 18.4
SF 143.7% 16 277 50% 33.9 2.3
NFL 33.4% 480 6,911 32% 32.9 7.7

L.A. allowed just one completion on a go/fly route all season, a 42-yard gain by Allen Robinson of the Bears in Week 7. DeAndre Hopkins also drew a 25-yard DPI on a go/fly in Week 13. The other 16 go/fly routes the Rams saw in 2020 were all incomplete—or intercepted, like the one Tom Brady threw in Week 11.

L.A.'s NFC West rivals in San Francisco were the worst defense against go/flies. The first two go/flies attempted against the 49ers both resulted in DPIs. That never happened again, but the next 14 go/flies thrown against them resulted in seven completions for 250 yards and five touchdowns.

The Falcons defense saw 24 go/fly attempts, most in the league; the Eagles only saw seven, fewest.

Corner

Yup, it's the Rams at the top of the table again.

Defense vs. Corner Routes, 2020
Team DVOA Tgt Yds Catch% aDOT YAC
LAR -206.4% 7 79 43% 20.4 1.7
IND -103.4% 14 91 31% 15.5 2.8
WAS -87.7% 8 77 50% 17.4 1.5
NE -68.6% 15 130 29% 21.0 8.5
CLE -59.9% 14 52 14% 20.4 1.5
CHI -53.0% 17 98 35% 16.8 1.2
KC -45.3% 17 127 31% 18.0 3.4
SF -41.9% 14 178 36% 22.5 12.2
MIA -41.7% 16 122 27% 22.5 0.5
TEN -31.9% 14 107 23% 21.4 1.3
JAX -28.9% 22 193 45% 17.6 2.1
ARI -22.5% 14 102 36% 17.8 4.2
BUF -17.5% 15 131 53% 14.6 1.1
LAC -2.1% 9 50 25% 17.2 1.5
DAL 0.7% 9 74 14% 21.9 0.0
NYG 3.3% 8 58 29% 23.1 1.5
CIN 14.1% 21 230 43% 18.7 3.1
TB 15.1% 14 119 50% 21.5 0.4
HOU 32.3% 12 117 50% 21.3 1.2
NO 42.5% 20 234 58% 19.8 2.0
LV 43.5% 24 221 50% 17.9 1.7
DEN 48.5% 11 134 64% 19.5 3.4
DET 49.4% 12 190 50% 18.8 9.8
BAL 55.1% 21 222 53% 19.2 3.6
CAR 63.9% 13 152 58% 20.3 1.4
MIN 81.4% 26 392 60% 18.5 6.2
NYJ 88.9% 13 180 62% 17.8 3.5
GB 88.9% 13 221 75% 17.6 3.9
SEA 90.8% 9 130 56% 18.6 6.4
PIT 91.3% 13 188 64% 17.5 2.3
ATL 98.6% 11 197 60% 24.6 4.7
PHI 125.0% 13 110 73% 13.3 0.6
NFL 9.8% 459 4,706 46% 19.0 3.2

L.A. faced the fewest corner targets in the league. They did give up a 43% catch rate and 11.3 yards per throw on corners, both right around league average. Their DVOA against the play was comically low because they intercepted three out of seven passes without surrendering a touchdown. You'll put up with a lot when you can get that kind of turnover rate.

The Eagles, meanwhile, allowed the worst DVOA on corners. The 13 corner targets they faced resulted in only three incompletions, with two DPIs for 24 yards and eight completions for 86 yards and five touchdowns.

The Vikings defense face the most corner routes, 26 in all.

Post

The Patriots had by far the NFL's top defense against post route, and their total DVOA honestly doesn't do them justice.

Defense vs. Post Routes, 2020
Team DVOA Tgt Yds Catch% aDOT YAC
NE -223.7% 9 21 11% 28.7 0.0
NYJ -133.3% 10 28 11% 21.2 3.0
DEN -124.0% 8 46 25% 21.1 7.5
ARI -63.2% 12 64 33% 21.9 2.3
TB -49.4% 11 109 45% 24.4 6.0
LAR -46.3% 15 102 29% 21.9 4.3
BAL -39.9% 19 101 26% 19.3 4.2
KC -30.1% 18 198 41% 24.1 6.4
TEN -18.9% 11 125 36% 29.8 3.5
CAR -2.8% 13 104 31% 29.7 3.8
ATL 0.2% 24 384 58% 21.8 5.0
DET 4.0% 9 98 44% 20.6 3.3
PIT 5.2% 22 202 38% 24.7 2.3
GB 9.6% 12 95 33% 23.1 1.5
WAS 22.3% 16 168 44% 28.4 2.6
CIN 22.3% 19 211 47% 27.1 2.2
MIN 27.1% 15 207 53% 24.1 3.8
IND 29.1% 18 190 50% 21.9 3.6
NO 34.9% 20 351 42% 32.9 10.8
CHI 39.0% 19 293 58% 17.8 8.0
LV 40.1% 14 232 50% 21.0 11.7
SF 46.4% 6 42 33% 31.2 2.0
DAL 47.5% 12 152 50% 22.8 4.3
JAX 49.5% 7 87 40% 24.7 1.5
BUF 50.8% 20 274 63% 16.8 4.0
CLE 64.0% 21 261 57% 21.5 2.0
SEA 66.4% 11 145 55% 22.5 3.7
HOU 68.9% 11 176 40% 28.3 6.0
LAC 70.1% 10 136 60% 19.9 3.8
NYG 90.6% 15 187 67% 15.7 2.7
MIA 117.2% 14 300 54% 32.2 6.1
PHI 140.9% 8 191 71% 28.4 2.2
NFL 14.2% 449 5,280 45% 23.7 4.5

The first five post attempts against New England had these results: interception, incomplete, interception, interception, interception. They did not intercept another post target after that, but they did force incompletions on every throw that both Deshaun Watson and Justin Herbert tried. They did not allow a single completion on a post route until a meaningless Week 17 game when they gave up a touchdown to—of all people—Sam Darnold and Chris Herndon of the New York Jets.

As for the worst defense on a given route, hey, it's the Eagles again!

(I don't actually hate the Eagles—the band or the NFL franchise—but I can't pass up a chance to make a Big Lebowski reference.)

The Eagles defense faced eight post routes in 2020. Two fell incomplete. One resulted in a DPI for 34 yards. The other five were caught for a total of 157 yards. Every completion resulted in a first down, of course. Hey at least none of them scored.

The Falcons are coming up a lot as the team that saw the most targets on a given route especially here at the bottom amidst the lower levels. I'm not sure what that says about them (other than they saw the third-most total passes in the league), but they were targetd on post routes an NFL-high 24 times last season. The 49ers saw the fewest posts with just six, with no completions allowed after Week 6.

Fade

We close with the fade, and a table where virtually everything is a result of small sample sizes.

Defense vs. Fade Routes, 2020
Team DVOA Tgt Yds Catch% aDOT YAC
DEN -90.1% 11 25 18% 17.9 0.0
ARI -77.4% 19 94 22% 16.9 0.0
SF -57.1% 16 108 15% 18.8 7.0
BUF -53.3% 9 41 13% 12.4 1.0
MIN -53.3% 21 156 29% 21.1 1.3
WAS -49.0% 9 64 25% 22.0 0.0
JAX -43.4% 17 117 24% 21.8 6.3
LV -41.3% 7 41 29% 13.3 2.5
NYJ -33.0% 8 30 29% 13.0 1.5
IND -29.5% 11 123 30% 19.7 15.0
LAR -22.4% 10 102 38% 24.8 2.7
MIA -8.5% 17 129 31% 21.4 2.4
NYG -7.5% 12 56 36% 14.0 0.0
NO -6.7% 15 148 33% 18.5 2.0
CIN -4.0% 24 181 23% 15.2 9.4
PIT 0.6% 9 74 33% 24.1 3.0
CLE 9.6% 16 125 25% 21.1 7.3
TB 11.3% 8 90 50% 15.8 1.0
KC 17.2% 23 214 45% 17.8 1.8
ATL 17.5% 14 183 45% 22.3 4.6
NE 19.1% 12 137 50% 21.4 3.0
TEN 21.8% 21 201 40% 17.6 4.1
LAC 22.5% 12 148 36% 26.3 4.3
HOU 28.0% 13 120 42% 19.5 2.4
BAL 36.6% 15 171 27% 23.4 1.7
SEA 47.7% 12 101 45% 19.2 1.2
DAL 48.3% 9 90 50% 24.1 3.5
DET 74.1% 18 252 44% 24.2 4.7
PHI 76.3% 24 290 43% 21.1 3.2
CAR 79.7% 10 128 56% 19.2 7.0
CHI 106.8% 12 184 50% 26.8 2.0
GB 357.9% 2 60 100% 20.5 9.5
NFL 3.5% 436 3,983 35% 19.9 3.6

The Broncos gave up one touchdown on a fade to Tampa Bay's Mike Evans in Week 3, and another to Mike Williams of the Chargers in Week 8. They intercepted another fade intended for Williams later in that same game; the other eight fade targets they saw all season were incomplete.

And then there are the Packers. They only saw two fade targets all year, a 19-yard touchdown by Adam Thielen in Week 1 and Richie James' 41-yard touchdown against busted coverage in Week 9. The Packers weren't perfect on defense last season, but despite their DVOA here, they did a fine job of taking the fade away.

The Eagles and Bengals tied for the top spot in targets with 24 fades faced each. Daniel Jones by himself threw seven of them against Philadelphia, with four completions for 121 yards and a touchdown, plus a 22-yard DPI. The Bengals were most frequently victimized by Tua Tagovailoa, who threw five fades in one game (including four in the second half!) in Week 13. Those five throws picked up a total of only 13 yards, though one was completed for a touchdown.

Detroit's Dilemma

We opened with a look at the Pittsburgh Steelers, last year's top pass defense. We close with the worst pass defense in the league, the Detroit Lions.

DET Defense vs. Most Common Routes, 2020
Route DVOA Rk Tgt Rk Avg. Rk
WR/TE screen -30.7% 12 27 7 5.8 16
Broken Play -26.4% 10 33 22 7.3 15
Deep Cross -8.2% 9 18 13 11.4 12
Post 4.0% 12 9 6 10.9 14
Drag 11.7% 23 49 32 6.9 18
Curl 15.4% 26 66 12 8.8 32
Out 16.7% 26 50 9 7.8 28
Slant 35.7% 31 27 5 10.2 30
Flat 36.3% 32 37 23 6.6 29
Go/Fly 37.1% 16 16 19 15.4 19
Corner 49.4% 23 12 10 15.8 30
Dig 58.7% 31 36 19 12.9 32
Fade 74.1% 28 18 26 14.0 30

We can summarize Detroit's strengths in exactly three bullet points:

  • The Lions were good when the offense called a WR/TE screen, which is a terrible play for most any offense.
  • They were good when they chased the quarterback out of the pocket, like most any other defense.
  • They were good against deep crosses.

That's it. They were bad against literally everything else. So long as you avoided screens against Detroit and kept your quarterback upright, he was bound to have a good day.

Comments

12 comments, Last at 02 Sep 2021, 2:15pm

1 Confusing

NE rates really well in several of these, but overall their pass defense was 18th by DVOA.  Was it really sucking on other routes or what?

4 Well, just in the routes…

In reply to by Raiderfan

Well, just in the routes listed here, they're in the top 10 in DVOA in seven of them, but 21st or worst in the other six. Not a lot of middle ground there.

This article also only includes the most common routes. New England had 130-ish targets on other routes not listed here (depending on precisely how you define screens), and their DVOA on those throws was 5.0%, 20th. (Pittsburgh was first, but the second-place team was Seattle, which is mind-blowing.)

Finally, and probably most importantly, none of these numbers include sacks. New England was 24th in adjusted sack rate, which also drags their overall pass defense DVOA down.

7 No Middle Ground

NE had an excellent back end and bad, no, terrible, no, awful, linebackers and defensive linemen. Given how weak the pass rush was and how poor they were at defending TEs, it's surprising that the DBs had the success they did. They had to cover for a long, long time.

2 Passer rating vs DVOA

I remember reading years ago that passer rating declines the further the ball is thrown down the field.

Passer rating is best when throwing the ball behind the line of scrimmage and is worst on deep passes.  DVOA I believe is the exact opposite.

This is the best that I could find regarding my point on passer rating:

https://www.sharpfootballstats.com/directional-passer-rating--off-.html

Does DVOA completely contradict the QB passer rating statistic?  The WR/TE screen is an awful play but I believe that it would result in a great passer rating due to its massive completion percentage and low interception rate.  

Conversely, the fly/go route results in great DVOA, but I believe a poor passer rating due to a low completion rate and high interception rate.

Here is a segment on passer rating taken from the Wikipedia page:

In 2011, Sports Illustrated published an article by Kerry Byrne of Cold Hard Football Facts highlighting the importance of passer rating in determining a team's success.[7] "Put most simply," the article states, "you cannot be a smart football analyst and dismiss passer rating. In fact, it's impossible to look at the incredible correlation of victory to passer rating and then dismiss it. You might as well dismiss the score of a game when determining a winner. [...] Few, if any, are more indicative of wins and losses than passer rating. Teams that posted a higher passer rating went 203–53 (.793) in 2010 and an incredible 151–29 (.839) after Week 5." Byrne made an expanded defense of the passer rating and its importance for the Pro Football Researchers Association in 2012.[8] The study showed that all of the eight teams since 1940 that led the league in both offensive passer rating and defensive passer rating won championships.[9]

A high QB passer rating correlates highly with winning, but so does passing DVOA.  What am I missing here?

5 Well, for starters, your…

Well, for starters, your basic premise is wrong. Passer rating is pretty stable regardless of pass distance.

 

Passer rating by depth of target, 2020

At or behind line of scrimmage: 90.7

1-10 yards downfield: 98.6

11-20 yards downfield: 92.1

21+ yards downfield: 94.9

 

Yes, short throws get lots of completions and few interceptions, but they don't produce yards or touchdowns. Deeper throws produce lots of yards and touchdowns, but more incompletions and turnovers. In raw numbers, most of those balance out. 

DVOA, however, knows that not all interceptions are created equal, and that an interception on a deep post is not nearly as bad as an interception on a screen.

 

Passing DVOA on INTs by depth of target, 2020

At or behind line of scrimmage: -960.2%

1-10 yards downfield: -912.4%

11-20 yards downfield: -823.3%

21+ yards downfield: -505.8%

 

That somewhat nullifies the extra interceptions that come with deeper throws, because those interceptions are not as bad as short-throw picks. Arm punts are a real thing. Meanwhile, deep throws produce dramatically higher yardage totals on average than shorter throws. That's why, on average, deep throws are better for DVOA than short passes.

I'm not at all surprised that the team with the higher passer rating usually wins. Teams with high passer ratings usually have lots of touchdowns and few interceptions, and teams with lots of touchdowns and few interceptions usually win a lot of games. That doesn't mean we can't do better, though, and DVOA attempts to do that by adding context to the raw numbers.

9 As a follow up, how does DVOA account for sacks and penalties?

Thanks, your explanation was excellent with regards to explaining DVOA vs QB rating. 

Now as a follow up, QB rating clearly does not take into account sacks, penalties or scrambling.

How does DVOA take into account sacks, penalties and scrambles? It is seemingly extremely difficult or impossible to do this as no pass was thrown at all in the case of sacks and scrambles.  However, deep pass play calls will clearly lead to more sacks, scrambles, and offensive holding penalties (they will also lead to more broken plays, but you have that covered in your data above).  The sacks and offensive holding penalty risks are clearly a drawback of throwing deep, however throwing deep will lead to more PI, illegal contact and defensive holding penalties, the great drive continuators of 3rd down passes.

Of course you would never know that long passes lead to more sacks or scrambles with Tom Brady running "bombs away" Bruce Arians offense.  Maybe my premise is incorrect to some extent?

10 The numbers you're looking…

The numbers you're looking at here are the numbers we use for "defense vs. receivers," broken down by route. So the average baseline is based on the average performance of attempted passes.

Our team defense numbers are based on the average baseline for all plays. Sacks, obviously, are much better than the average play for a defense, so DVOA rewards them highly. 

It's a general rule that the deeper you throw, the more sacks you will take, but there are plenty of exceptions. Tampa Bay had the highest aDOT in the league last year, but one of the lowest sack rates. Buffalo and Denver stood out in that department too. On the flip side, Washington never threw deep but still gave up lots of sacks. To a degree, so did New England, San Francisco, and New Orleans.

11 However, deep pass play…

However, deep pass play calls will clearly lead to more sacks, scrambles, and offensive holding penalties (they will also lead to more broken plays, but you have that covered in your data above). 

Note that you're shifting here from deep pass plays to deep pass calls. There's a huge difference between the two (inasmuch as there is such a thing as a 'deep pass call'), and obviously without knowing what the play call was you can't group things together like that.

That's the difficulty with looking at breakdowns like this - you're only seeing the result, not the process. It's a selection effect - you're not seeing all the plays that weren't called (or weren't thrown). It's also tough because even if you situationally correct the results of the play, you can't situationally correct the route fraction. As in, if all of the deep routes a team faces occurred on, say, 3rd and long (where they're practically forced to do it), that's different in comparing to another team where the deep routes happen all on 1st and 10 because the defense is just that bad.

For shorter routes I don't think that's much of an issue, because I have to imagine most of those are intentional. The deeper routes have gotta be more opportunistic/situational.

12 These are all good points…

These are all good points. One of the reasons the DVOA numbers for deep routes are so good is that QBs only throw them when their receiver is open. Nobody checks down to a deep cross or corner. That's one of the reasons volume is meaningful along with efficiency.

6 Given how often they're run…

Given how often they're run in combination, it's interesting to me that Detroit was so bad on slants and flats. You'd think that teams incapable of defending both well might make schematic choices that effectively shut down one of the two, but apparently not!

8 The Jets were pretty good to…

The Jets were pretty good to very good on routes in the middle of the field, indicating good safety play, but very bad on any downfield passes to the outside, indicating bad CB play.  The exception is the fade route, where size and physicality of a CB matter more than speed/quickness/loose hips, etc.  That matches the eye test.  Unfortunately for the Jets, the only additions to the CB room this year are mid to late round rookies.  It's going to be a long year of 3rd down conversions on outs, digs, and curls.