<
        >

          Nine NFL stats, trends and numbers you should know: Replacing Robert Woods, A.J. Terrell's breakout and a shocking Seahawks issue

          • Seth WalderNov 21, 2021, 06:55 AM ET
            Close
              Seth Walder is an analytics writer at ESPN, specializing in quantitative analysis. He is also a regular on "ESPN Bet Live" and helps cover sports betting. Seth has been at ESPN since 2017. He previously worked at the New York Daily News covering the Jets and Giants. You can follow Seth on X via @SethWalder.

          Dallas' Trevon Diggs is getting all the glory -- eight interceptions! -- but he isn't the only second-year cornerback breaking out in 2021. A close look at the metrics show A.J. Terrell is having a huge season too, just in a quieter way. And future Hall of Famer Aaron Donald? His pass-rush numbers show that he's still at his peak.

          As I comb through our metrics each week researching angles and stories, it's impossible to not come across numbers and trends that catch my eye. Like Terrell and his tiny yards per coverage snap allowed, Donald's greatness, or a shift in the coverages teams are employing to stop modern offenses.

          Consider this an emptying of the proverbial metrics notebook; here are nine stats that caught my eye and tell us something about what we've seen in 2021 and where we're going:

          Jump to a topic:
          Eagles' strength vs. Saints' strength
          Can the Rams replace Woods?
          Rookie tackles are struggling
          Inside A.J. Terrell's breakout
          Seattle's late-down struggles
          A strange stat for a Lions RB

          Aaron Donald is breaking charts again

          It took a few weeks for the star Rams defender to take over the No. 1 defensive tackle spot in PRWR from Philadelphia's Javon Hargrave, but now Donald has blown past the competition.

          In addition to his pass-rushing prowess, the double teams he draws are even more incredible than this chart lets on. When Donald plays defensive tackle, he's almost always playing the 3- or 4-technique. Many other defensive tackles -- including the two ahead of Donald in double-team rate -- the Patriots' Christian Barmore and the Giants' Dexter Lawrence -- spend some time at nose tackle. The farther inside a defensive tackle is, the more likely he is to be doubled.

          Both ways Donald excels in this chart -- his win rate and his double-team rate -- should help out new teammate Von Miller rack up sacks for the Rams.

          Strength on strength: Eagles' ground game vs. Saints' run stop

          The Saints have allowed 2.9 yards per designed carry this season, which is on pace to be the lowest rate allowed in the entirety of our dataset (since 2006). To put in perspective how much of an outlier they are in 2021, the Bucs rank second at 3.6 yards per designed carry.

          No surprise given those numbers: New Orleans ranks first in run stop win rate (RSWR) this season, too. And Cameron Jordan is first in RSWR at edge.

          That makes for a pretty great strength for the Saints this weekend because the Eagles are one of the most efficient running teams in the league. They rank third in EPA per designed run play, and it's not just quarterback Jalen Hurts, either; they rank fourth in EPA per play when a running back carries the ball, too. Something's going to have to give on Sunday.

          Rookie tackles are struggling in pass protection

          The exceptions to the headline above are the Bills' Spencer Brown and the Chargers' Rashawn Slater, who have recorded above-average pass block win rates (PBWR) this season. Here are the rankings for the five other rookies who qualify for our tackle leaderboard, which goes 71 players deep:

          That list includes three of the four first-round tackles in Sewell, Darrisaw and Leatherwood, the last of whom has since kicked inside to right guard. (By the way, Brown ranks 16th; Slater is 29th.)

          It's a reminder that first-round tackles should not be thought of as a quick fix. In fact, they're usually below average in PBWR as rookies. One piece of optimism for Sewell going forward: In Week 10, he moved to right tackle with Taylor Decker returning from injury, and the No. 7 overall pick had by far his best game by PBWR.

          A.J. Terrell's second-year breakout

          Quantifying the play of cornerbacks remains an unsolved problem, but one simple metric I like to lean on is yards per coverage snap allowed, which is based on nearest defender data from NFL Next Gen Stats. The key here is that part of a corner's objective is to discourage targets. Yards per coverage snap rewards them from preventing a completion, either by preventing a target or forcing an incompletion.

          And guess who is No. 1 among outside corners this year? Terrell, whom the Falcons picked No. 16 overall in last year's draft. He's excelling in both areas; opponents are targeting him 13% of the time -- below the 17% average for outside corners -- and when they do, their completion percentage over expectation is minus-14%.

          The only area in which Terrell lacks is interceptions -- he has just one, on Thursday night off Mac Jones. Another way to look at corners is expected points added (EPA) per coverage snap, which rewards them for an incompletion when they are the targeted defender and really helps if they record an interception. With only one pick, Terrell almost by definition can't be at the top of the chart, but he's still well above-average because of how hard it has been to throw in his direction.

          Atlanta's defense remains a problem, as it has been for years. It ranks 25th in EPA per play overall and 28th against the pass. If the Falcons are going to turn around that unit, they'll need some stars beyond Grady Jarrett. It looks like they found one in Terrell.

          Replacing Robert Woods

          Before suffering a torn ACL last week, the Rams wide receiver led the league in separation over expectation this season. In other words: When targeted, Woods is averaging more separation than any other receiver after controlling four factors such as down and distance, air yards, distance from sideline and route run. He had generated 0.7 yards of separation over expectation on targets according to the model, which uses data from NFL Next Gen Stats.

          This is an area in which some X's and O's expertise comes in handy, so I ran the information and a few plays past my colleague Matt Bowen. He thought that Woods' separation was a result of scheme vs. zone -- frequently in-cutting intermediate routes -- and his ability to get open vs. man.

          In some ways, that's good news for the Rams because if Sean McVay can scheme open Woods, he should be able to scheme open Odell Beckham Jr., too. At the same time, No. 1 is No. 1. While there are several factors that determine separation, being on top of the leaderboard suggests specific strengths that Woods had to get space between him and defenders. The Rams are definitely going to miss him.

          A D'Andre Swift stat oddity

          Here's a weird stat combination: The Lions running back has recorded negative-127 rush yards over expectation (RYOE), the worst total among all players this season, but also plus-113 yards after catch (YAC) over expectation, the best among all running backs. Both metrics from NFL Next Gen Stats are based on similar models, which set yardage expectations based on factors including the location, speed and orientation of all 22 players on the field at the time of handoff or reception. Strange!

          It is worth noting that Swift's YAC numbers are partially driven by volume. His YAC over expectation per reception is only above average, but he has caught 50 passes. Receiving volume is a skill though, so it's still notable. The disparity between the two metrics is odd, and I think it accentuates the point that gaining yards via handoff and via YAC should be thought of as separate skills.

          It also makes me wonder if Swift's long-term role will be similar to Alvin Kamara's. Kamara has long been a dangerous receiving weapon, but this season the Saints have tried running him far more ... and it has not worked, at least in terms of RYOE. He has accumulated the second-worst RYOE at negative-104, only ahead of Swift.

          A coverage shift across the league

          Here's the rate of coverages using two-high safeties (2-Man, Cover-2, Quarters and Tampa-2) across the league as a portion of all coverages, with goal line and prevent excluded, over the past several years, according to data from ESPN coverage metrics using NFL Next Gen Stats:

          • 2017: 37%

          • 2018: 37%

          • 2019: 38%

          • 2020: 38%

          • 2021: 41%

          That might not seem like much of a shift, but when we're talking about the entire league, even three percentage points in a single season (3.5% if we remove rounding) is significant.

          What's the takeaway? Defenses are prioritizing the pass game. It's a shift that makes sense. Across the board, pass plays are worth about a 10th of an expected point added more than run plays. Think about it this way: For every 10 run plays a team replaces with pass plays, it improves its point differential by about a point.

          We've known for a while that passing is the most important thing to get right on offense. The same is true for defending it on other side of the ball.

          Will the Seahawks' late-down struggles continue?

          Seattle's offensive splits between early downs and late downs are truly bizarre. It ranks fifth in EPA per play on first and second down ... and last on third and fourth down. Last! With Russell Wilson (mostly) as the quarterback!

          A little bit of this can be explained by the fact that the Seahawks run at above average rates on third-and-long, but that's just a piece of it. Because even if we split to look at their early-down pass plays vs. their late-down pass plays, the gap is almost as wide (sixth vs. 30th). A huge culprit here is sacks, a longtime weakness of Wilson's, as the Seahawks have taken sacks on 19% of their late-down dropbacks.

          It's not impossible that part of the explanation is an imperfection within EPA itself, because the Seahawks do average one of the longest distances to go on late downs. The best guess here is just variance, though. Late downs are a relatively small set of plays, but a highly impactful set. That's probably good news for the Seahawks.

          I looked at teams from 2010 to 2020 and compared their EPA splits from the first half of the season to the second half. First-half early-down EPA per play correlated better with second-half late-down EPA per play than first-half late-down EPA per play. That actually makes sense if you think about it: There are more early-down plays, so it's just more information. Of course, the highest correlation is with first-half-overall EPA per play (Seattle ranks 21st).

          Either way, it's reasonable to expect significant positive regression from the Seahawks' late-down efficiency, though we should probably expect slightly worse performance on early downs, too.

          Motion at the snap is still an edge ... in the pass game

          For the fourth consecutive season, motion at the snap rates have increased leaguewide. In 2017, teams employed a man in motion at the snap just 4% of the time, and that number has more than tripled to 14% this season.

          We wrote back in 2019 that motion at the snap was an edge for the offense in terms of EPA per play relative to non-motion-at-the-snap plays in both the run and pass game. That remains only partially true. You can see in the charts below that as motion at the snap usage on run plays increased from 6% to 20% from 2017 to 2021, the advantage over non-motion at the snap plays dissipated (though with a small bounce back this year). Zero on this chart indicates when the effectiveness of motion at the snap plays was the same as non-motion at the snap plays.

          Motion at the snap has also increased on designed pass plays -- though at a lower level -- but the advantage still exists, to the tune of .08 EPA per play in 2021.

          What does this mean going forward? I'd expect that the growth of motion at the snap to slow in the run game, but still continue to increase steadily in the pass game. If there's an advantage to be found, teams likely will grab it.