Why Alex Leatherwood Holds Key to Raiders Season

Las Vegas Raiders OL Alex Leatherwood
Las Vegas Raiders OL Alex Leatherwood
Photo: USA Today Sports Images

NFL Offseason - After the Raiders made two of the biggest personnel moves of the offseason, Las Vegas fans probably can't wait to see what Derek Carr can do with Davante Adams on offense, or how Chandler Jones will work with Maxx Crosby on defense. But the player who ultimately decides their season could be Alex Leatherwood, because if the Raiders are going to keep up with the arms race in the AFC West, they're going to need major improvement from last year's first-round draft pick.

The Raiders blew up their offensive line in the 2021 offseason, when former general manager Mike Mayock traded away starters at center (Rodney Hudson), right guard (Gabe Jackson), and right tackle (Trent Brown). Leatherwood, the 17th pick in the draft and the fourth offensive lineman to go off the board, was supposed to fill Brown's role on the right edge and lock down that side of the line for years. His collegiate resume suggested he would be up to the task—he was a three-year starter for Alabama at guard and at both tackle spots, with two all-SEC awards, two national championship rings, and an Outland Trophy as the nation's best lineman to his name.

Unfortunately, Leatherwood failed to perform at that level in Vegas. He started the first four games of the year at right tackle, and in that month he was credited with (blamed for?) a dozen blown blocks by Sports Info Solutions, tied for third most in the league in that timeframe. At that point the Raiders gave up and moved Leatherwood inside to right guard, where Denzelle Good had suffered a torn ACL in Week 1. Between the two positions, Leatherwood ended up starting all 17 games, as did left tackle Kolton Miller, left guard John Simpson, and center Andre James. With Brandon Parker taking over for Leatherwood at right tackle, the Raiders were the only team in the NFL in 2021 to start the same five linemen for 13 straight games (a streak that stretched to 14 games if you include the wild-card loss to Cincinnati). That's the biggest reason the Raiders finished first in Football Outsiders' offensive line continuity score, which measures an offense's ability to field a consistent starting quintet week-in and week-out. (The details of continuity score are explained below.) It was a big change from 2020, when the Raiders finished in the bottom five in continuity—which was one of the reasons they got rid of Hudson, Jackson, and Brown in the first place.

Biggest Year-to-Year Improvements
in Offensive Line Continuity, 1999-2022
Team Years Cont
Yr1
Cont
Yr2
Dif
IND 2018-19 24 48 24
NE 2015-16 15 38 23
LV 2020-21 21 43 22
GB 2002-03 26 48 22
JAX 2018-19 22 44 22
PIT 2003-04 27 48 21
MIN 2014-15 28 48 20
WAS 2011-12 23 43 20
CHI 2007-08 28 48 20
4 Teams Tied 19

For all that stability, though, Las Vegas' line play wasn't very impressive. They finished 17th in adjusted line yards and 18th in adjusted sack rate, and SIS had them just 20th in pressure rate allowed. That mediocrity was good enough to win a fluky playoff berth in a down year in the AFC, but with the conference as a whole and the AFC West in particular adding extra firepower for 2022, the Raiders will need to improve to just to get back to the postseason, let alone win their first playoff game in two decades. And improvement for the offensive line will need to come from the players already on the roster, because the same five starters are expected to return. The Raiders are fresh out of draft capital (they gave up their first- and second-round picks in the Adams trade) and free-agent cash (they guaranteed Jones over $30 million).

With the Raiders unable to make an upgrade in offensive line personnel, they must get an upgrade from the personnel they already have on hand, and Leatherwood stands out as the blocker with the most room for improvement. He led all guards with 40 blown blocks and 26 blown blocks on passing plays, the only guard to make the top 10 in either category. No other Raiders lineman finished within 10 blown blocks of Leatherwood, and remember that three of them also started every game.

The good news is that Leatherwood was a rookie last year, and most NFL players make their biggest improvements in their second seasons. Right now, however, Leatherwood is pretty clearly the biggest weakness on the Las Vegas offense. And as stacked as the AFC West looks, one weakness could be the difference between a playoff berth and a last-place finish.

Other Good Teams

Tampa Bay's offensive line has a legitimate gripe about their second-place ranking in offensive line continuity. The starting quintet of Donovan Smith, Ali Marpet, Ryan Jensen, Alex Cappa, and Tristan Wirfs missed only one start between them, when Marpet sat out the Week 12 win over Indianapolis with an abdominal injury. Because that absence came near the middle of the year, it limited Tampa Bay's longest streak of starting the same five linemen to 10 weeks. If that same absence had taken place before Week 5 or after Week 14, the Buccaneers would have been alone in first place. Subjectively, it's weird to see the Bucs behind the Raiders when they were one start away from a perfect season. It's a similar story in Atlanta, where Jake Matthews, Jalen Mayfield, Matt Hennessy, Chris Lindstrom, and Kaleb McGary started 83 of a possible 85 games.

Four teams were tied for fourth place behind the Raiders, Buccaneers, and Falcons. Let's just bullet-point 'em:

  • Los Angeles Rams: Sean McVay's teams almost always fare well in offensive line continuity—in his five seasons as head coach, the Rams have now finished first, first, 12th, tied for first, and tied for fourth. Right guard Austin Corbett started all 17 games for L.A., and Andrew Whitworth, David Edwards, Brian Allen, and Rob Havenstein started at least 15 games each. Unfortunately those absences were scattered throughout the year, and so L.A. never had the same five starting linemen for more than seven games in a row.
     
  • Minnesota Vikings: The left tackle spot was a bit of a mess for Minnesota, with three different players swapping in and out of the lineup. But left guard Ezra Cleveland and right tackle Brian O'Neill each started every game, and when center Garrett Bradbury or right guard Oli Udoh were unavailable (either due to injury or because Udoh was filling in at left tackle), Mason Cole was able to fill in for either of them.
     
  • New England Patriots: The Patriots rank remarkably high considering they used 10 different starters on the line, including five different right tackles in their first seven games. But once Trent Brown finally got healthy at that spot, New England had the same starting quintet in each of their last eight contests. Only Las Vegas had a longer active streak at the end of the year.
     
  • New York Jets: A reminder that "continuity" is not necessarily synonymous with "health." Left tackle Mekhi Becton suffered a knee injury in Week 1 and didn't play again all year. The Jets moved right tackle George Fant to the left side and started Morgan Moses in his place; they started 15 and 16 games, respectively, which boosted New York's continuity even though they were basically playing backups at two positions. Left guard Ali Vera-Tucker and center Connor McGovern also started 31 games between them, though the right guard spot was something of a turnstile.
     
  • San Francisco 49ers: The 49ers were perfect on the interior, with Laken Tomlinson, Alex Mack, and Daniel Brunskill each starting all 17 games. Left tackle Trent Williams missed only two starts. At right tackle, Mike McGlinchey started eight games before suffering a torn quad; his spot was filled first by Jaylon Moore and then by Tom Compton.

Saints, Panthers, Eagles Threaten Futility Records

At the bottom of the table we find two teams tied for last place … and it's two teams from the NFC South, which is remarkable considering the other two teams in the division finished second and third. The Carolina Panthers and the New Orleans Saints each had an offensive line continuity score of just 17, threatening the all-time low mark of 15 set by the 2015 Patriots and 2020 Eagles (covered in great detail last year). Neither team had a stretch longer than two games of starting the same five linemen in the same five positions, tying the all-time record low in that category with a half-dozen other teams. And the Panthers also tied the record of those 2015 Patriots by taking the field 14 times with a different starting lineup than they had used the week before.

The Panthers had exactly one reliable name on the offensive line: Taylor Moton, who started every game (16 at right tackle, one at left). But Dennis Daley was the only other Carolina lineman who hit double digits in starts. The Panthers had four different starters making seven lineup changes at both left tackle and at right guard. Moton's not the only lineman who found himself starting games at multiple positions: Daley made starts at right and left guard and left tackle; Brady Christensen started at left and right tackle and right guard; and Pat Elflein started at both left guard and center.

While the Panthers had Moton to lean on, the Saints had Cesar Ruiz. The 2020 first-rounder started all 17 games (13 at right guard, four at center). New Orleans also got 13 starts out of James Hurst (eight at left tackle, three at right tackle, two at left guard,) 12 out of Erik McCoy (all at center), and 10 out of Ryan Ramczyk (all at right tackle). No other Saints lineman made it to double-digit starts. Things were especially chaotic on the perimeter. New Orleans had a change at right tackle in each of their final seven games; they also had a different (but overlapping) streak of seven consecutive games with a change at left tackle.

Before we move, we need to mention those Eagles again. They improved from "as bad as any team we have ever measured" to "tied for third worst in this specific season," so hey, good for them! Along the way, however, they used a dozen different starters on the offensive line, which is somehow one more than they used in their apocalyptic 2020 campaign (though that total of 11 included neither Brandon Brooks nor Andre Dillard, expected starters who were injured in the preseason and missed the entire year). Only two teams have ever used more starters: the 2007 Rams and 2018 Cardinals, who both used 13.

What's most amazing about this back-to-back run of Philly futility is that Jason Kelce started every game at center in both seasons. (In fact, he has not missed a start since 2014.) The Eagles only had three linemen make double-digit starts: Kelce, plus Landon Dickerson and Lane Johnson, who had 13 starts apiece. The Eagles used three different starters at left guard, four at both left and right tackle, and five at right guard (including four in their first five games). Philadelphia still has two picks in the first round of the draft this year, and as much as they need a receiver, it sure would be nice to find a lineman who could survive a full season.

The Nuts, the Bolts, and the Elephant in the Room

Offensive line continuity scores were originally developed by Jason McKinley in the early days of FO (when FO Almanac was still Pro Football Prospectus) and we have subsequently gone back and calculated it for every team since 1999. The continuity scores are based on three variables:

  • number of starters used;
  • number of week-to-week changes in starting lineups;
  • and the longest starting streak of any one five-man unit.

A team can earn a maximum of 17 points in any one category (one point per game), meaning a team that started the same five linemen in all 17 games would get a perfect score of 51; 39 teams have achieved perfect scores, most recently the Indianapolis Colts in 2019. Hypothetically, if a team started five brand new linemen every week of the year, they would get a "perfect" score of -61, though of course nobody has ever come close to that. As mentioned earlier the worst continuity score on record belongs to the 2015 New England Patriots and 2020 Philadelphia Eagles at 15.

That's a boilerplate paragraph that we copy-and-paste every year, usually with only a few minor updates. This year, however, we had to make some significant changes because the NFL schedule jumped from 16 games to 17. This means the maximum score got a little higher, and the minimum score got a little lower. Those 39 teams with perfect continuity scores? They all pulled it off in 16-game schedules, meaning they had perfect scores of 48, not 51. We considered trying some mathematical magic to fudge the numbers for 2021 and squeeze them to fit the old scale, but ultimately decided to go ahead and leave them be. After all, the extra week meant not only an opportunity for scores to go up, but also for scores to go down based on injury/illness/suspension/benching/what have you.

The average team in 2021 had a continuity score of 26.9, up from the 25.5 average score we saw in the 2020 season. Remember, though, that 2020 was the NFL season most affected by the COVID pandemic, with multiple postponements and entire position groups removed from some games. In that chaos, it's only natural that continuity on the offensive line (and all other positions) would go right to hell. The continuity scores we measured in 2020 were the lowest on record, but the continuity scores of 2021 were second lowest.

This has been part of a long-term trend. From 1999 to 2008, league-average continuity scores were usually in the mid-30s. They started to drop after that, falling into the 20s every year since 2015. As the following chart shows, this is partly because teams are using more starters, but mostly because they are making more changes in the lineup, and thus the average longest stretch without making changes is also going down. That was especially true last season. We have counted 43 teams that made at least 10 lineup changes in a year; 13 of those teams (30%) played in 2021.

Changes in OL Continuity

The following table shows full offensive line continuity results for all teams in 2021. We had two teams in the 40s, nine in the 30s, 19 in the 20s, and two in the teens. That distribution falls right in line with the averages in the prior five seasons.

Offensive Line Continuity Scores, 2021
Tm Starters
Used
Lineup
Changes
Longest
Streak
Continuity
Score
LV 7 2 13 43
TB 6 2 10 41
ATL 7 3 10 39
NE 10 5 8 32
MIN 7 5 5 32
SF 8 5 6 32
LAR 8 6 7 32
NYJ 9 6 8 32
CHI 9 5 6 31
DAL 7 8 6 30
KC 7 7 5 30
BAL 9 8 6 28
PIT 9 7 5 28
GB 9 8 5 27
LAC 10 8 5 26
DET 9 9 4 25
MIA 9 9 4 25
BUF 8 10 3 24
CLE 9 10 4 24
NYG 10 10 5 24
SEA 10 9 3 23
DEN 10 10 4 23
IND 9 10 3 23
JAX 10 10 4 23
ARI 8 11 3 23
CIN 11 9 3 22
TEN 10 11 4 22
WAS 10 11 4 22
PHI 12 10 3 20
HOU 11 12 4 20
CAR 10 14 2 17
NO 11 13 2 17

Comments

14 comments, Last at 07 Apr 2022, 12:57pm

1 The aspect ratio on your…

The aspect ratio on your plot image looks off.

Is the decrease in continuity score driven by injury or by players changed for performance reasons? (Does it correlate to the injury list?)

2 Is there any evidence that…

Is there any evidence that continuity could change for anything other than injury or performance?  I.e. teams having different starters for matchup reasons rather than the other two? Seems like a possible reason for it to go down over time. 

 

 

5 This rarely happens in the…

This rarely happens in the NFL. It's difficult enough swapping from the left side to the right that you wouldn't want guys to do it midseason if you didn't have to. Though there might be cases of tackles being kicked inside to guard -- Leatherwood probably would have wound up there whether Denzelle Good had been injured or not. 

The exception to this rule is Sean Payton. He has always liked to shuffle guys around, usually ranking high in the number of lineup changes but low in the number of starters used. In the past 10 years, the Saints have ranked in the bottom 10 in continuity score six times, but in the top 10 just once.

8 Not that I can recall. Maybe…

Not that I can recall. Maybe in his rookie year, when they would start Chris Chandler but bring in Vick for special packages. But that wouldn't affect the starting lineup, and the starting lineup is what continuity score is based on.

6 Usually when this happens,…

Usually when this happens, the sixth lineman is just listed as "T" or "OL" with no LT/LG/C/RG/RT designation, so it has no effect. There were one or two cases where we had to go back and manually check which linemen played the whole game and which were specialists.

9 What's the reason for using …

What's the reason for using "longest streak of any five-man unit" rather than "number of games of most commonly used five-man unit"?  As you point out with TB, the current formula penalizes a short change in the middle of the year versus the same length of change at either the one-quarter or three-quarter mark of the season.  Even in an hypothetical situation where a team went back to the same unit every second game, with a series of one game injuries to one player or another interrupting the appearance of the regular unit, the team would still be fielding a consistent line up 50% of their games and the "number of week-to-week changes" would seem to adequately capture the discontinuity; no need for longest streak to double count this, I wouldn't think.

12 Hi Lost Ti-Cats Fan, I'm the…

Hi Lost Ti-Cats Fan, I'm the original researcher on this from way back when. At the time, I had a few decent ideas to derive a continuity score which I could then use to run simple correlations against various metrics (offensive DVOA, adjusted sack rate, etc.). After gathering the data I couldn't really decide which one would best capture continuity so I threw my hands up and combined the three. I think in the original article I probably mentioned other ideas to improve it. Something similar to your idea may have even been one of them because I remember being very annoyed that the Patriots kept swapping out starting guards for the first half of a (typically great offensive) season but they had a very low continuity score because of that deliberate strategy and switching to a "most common five" would have muted that a bit. It sounds like that has happened with the Saints as well. But, I wanted to get it into the book so I went with what I had and made it in time for publication.

All this stuff is always a work in progress, so ideas like yours that might improve results should be considered. Limits to implementation might exist around how the data is collected and stored - it may not be easy to run it the way you suggested. (Maybe it is, I don't know; I'm not the one running it!) My goal at the time was to just explore the "Continuity on the O-line is critical to offensive success!" mantra that I always heard. Surely there are better methods to capture continuity, but I haven't really seen anyone else try so I'm glad FO still keeps running this and publishing it. 

Unrelated to your question but related to the impact of continuity: The correlation I would probably run now would be against the team pass block win rates and run block win rates. Those metrics have only come along in the last three(?) seasons but I think they tie in really well since they really only focus on the lines' ability to block as a unit and are a lot less reliant on QB or other "skill" positions.

10 GB w/Bak

*cough* shouldn't have played week 18 *cough*

But he should (hopefully) be ready by week this year, along with Josh "picked before Creed Humphrey" Myers. But Turner and Patrick who also started most games are gone. And Elgton likely will miss the first half of the season or so.

Hopefully the OL goes something like

  • LT: Bak
  • LG: Runyan
  • C: Myers
  • RG: Rookie (any G on the block?)
  • RT: Newman (866 snaps in college, wasn't super great at G anyway, yeah it may be counter intuitive idc)

Gonna be different again. Elgton gonna slide into RT when he's healthy I presume. As we saw in the playoff game they didn't trust UDFA Yosh and truthfully they're probably right not to in contrast to the clamoring for him.