How many good offensive linemen is enough?

Let me beat all you readers to the punch: I know, you can never have enough good offensive linemen, so “all of them.” OK, now that you’ve gotten that out of your system, let’s return to reality.

Good offensive linemen are a scarce resource at the NFL level. Every fan wishes that their team was rock-solid all along the line, but very few teams are. There are more misses along the offensive line in the draft than most of us would like to admit. Furthermore, the NFL is a salary-capped league. Even if you could put together a set of five great offensive linemen on one team, they would soon price themselves out and several of them would have to go elsewhere to get a big second payday.

So how good is good enough to succeed?

The ‘weak link’ theory of offensive lines

There have been several studies that have popularized the notion that football is a “weak link” game, i.e. the performance of any unit on the field, whether it be the offensive line, the defensive backfield, etc., is only as good as the performance of its weakest member. Intuitively this makes sense. How often do we hear of offensive play callers staying away from a team’s shutdown cornerback and attacking one of the other defensive backs that they feel can be beaten more easily?

A study in Pro Football Focus several years ago demonstrated how much difference it makes to both rushes and passes to have perfect blocking on any given play. Here are the stats for rushes (perfectly blocked is defined as a play on which each lineman received a positive or neutral PFF blocking grade):

Courtesy of Pro Football Focus

The difference is stark, and almost as surprising is how seldom all five offensive linemen do their job on running plays. To put more of a face on this, Arjun Menon (now working for the Jets) plotted the yards per carry on perfectly blocked vs. non-perfectly blocked plays for individual running backs after week 5 of last season:

No one averaged even five yards per carry when the blocking wasn’t perfect. Some were hardly any better when the blocking was perfect, but a couple averaged more than 12 yards per carry when the blocking was perfect (including Saquon Barkley).

Surprisingly, the difference wasn’t as stark on passing plays, but there is still a clear difference between perfectly and non-perfectly blocked plays:

Courtesy of Pro Football Focus

Note also the surprising result that pass plays are much more often perfectly blocked than not, the opposite of rushing plays. A blocking failure on a pass play leading to a sack or quarterback hit is much more obvious than a miss on a run block that leads to a back only gaining two yards, but the latter is apparently much more common.

So it’s better to have five adequate offensive linemen of similar skill than five of varying skill from excellent to poor. As stated by Eric Eager of Sumer Sports:

Consider a group of n players that each have a probability of succeeding on each individual play. Assuming independence, an assumption which has different levels of validity depending on the situation, the likelihood that the entire unit succeeds – e.g. the system doesn’t fail – is the product of all of the probabilities. Assuming that there is a fixed sum of all of the probabilities in the group – an assumption that is reasonable given the constraints of the salary cap – the probability that the system doesn’t fail is maximized when all of the individual probabilities are the exact same.

What can we learn from last year’s playoff teams?

Here is a plot of 2022 regular season PFF team run blocking vs. pass blocking grades for all 32 NFL teams:

Data from Pro Football Focus

Teams above and to the left of the 1:1 line were better at run blocking than pass blocking, and teams below and to the right better at pass blocking than run blocking. There are many more teams in the latter category than the former, supporting the stats presented earlier that perfectly blocked running plays are rarer than perfectly blocked passing plays. (There could also be a league selection effect in that the dominance of passing in today’s NFL may cause teams to value and retain good pass blockers preferentially.). I’ve labeled some of the points corresponding to the best overall blocking teams, the NFC East teams, all other playoff teams, and a few outliers of interest.

There were only four teams who were strikingly better at run blocking than pass blocking last season. It’s no surprise to see Atlanta and Tennessee in this group - they are two of the most run-oriented teams in the NFL. It’s a shock, though, to see pass-happy Miami and Minnesota with them in that group. That might suggest an equation of the form:

Pass-heavy offense + better run- than pass-blocking OL = first-round playoff exit

By contrast, there were a large number of lousy run-blocking but good or very good pass-blocking teams. The biggest discrepancies belonged to the Chargers, Jacksonville, Tampa Bay, and Green Bay. Coincidentally these teams had among the most prolific but not overly mobile quarterbacks in the league.

The best blocking teams in the NFL

In 2022 there were only four teams that graded above average (70+) in both run- and pass-blocking as a team. Two of them (Philadelphia and Kansas City) played in the Super Bowl, so there’s your argument that you need a great OL to win a championship in the NFL. However the third, Baltimore, went out quietly in the first round of the playoffs, and the fourth, Atlanta, didn’t make the playoffs. Apparently a quarterback has something to say about team success also (remember that Lamar Jackson was hurt late in the season and missed the playoff game).

Here are the stats for the five offensive linemen on each team who got the most blocking snaps:

Data from Pro Football Focus

The Eagles were in a class by themselves last year. All five starters graded above average in pass-blocking, and only two were even as bad as average in run-blocking. Four of the five starters totaled two sacks among them. (Lane Johnson has not given up a sack since 2020.) Still, they had something of a weak link in tackle Jordan Mailata, who surrendered 6 sacks and 38 total pressures.

The Ravens were almost as good. They had three elite pass blockers and two elite run blockers. They weren’t perfect, though. Rookie center Tyler Linderbaum was poor in pass blocking and now-departed Ben Powers was an inadequate run blocker.

The Falcons had their own weak link in Drew Dalman, who was subpar on passing downs although he didn’t give up any sacks. As a team, Atlanta gave up surprisingly few pressures, but this was most likely the product of their run-first offense.

The Chiefs had one of the most unusual offensive lines in the NFL last year. They had two arguably elite players in Joe Thuney and Creed Humphrey, but one was better run blocking and the other better pass blocking. Other than that, they had mostly average play on the line, but with lots of pressures surrendered. Overall this team graded as one of the best OLs but had considerable flaws. Would this OL have graded as highly if they didn’t have a future Hall of Fame quarterback behind them, one with great “escapability” among all his other elite traits?

A better comparison

None of the four teams discussed above is an especially good model for the New York Giants. The Giants are never going to have an OL the quality of Philadelphia’s, and probably no other team will either. Baltimore’s may be a better comparison but they didn’t get a chance to show how far they could go in the absence of Lamar Jackson. Atlanta’s was too one-sided and didn’t get them to the playoffs, and Kansas City’s played in front of what appears to be an all-time great.

Let’s try a different comparison:

Data from Pro Football Focus

The Giants and the Jacksonville Jaguars were two pretty evenly matched and similar teams last year in many respects. They played a game that came down to the final play and one yard, they both made a surprise playoff appearance, and each won a playoff game before being eliminated by their conference’s Super Bowl participant. Both have young quarterbacks just now coming into their own. Both last year had a committee of receivers to throw to with no clear elite receiver in either group, though Jacksonville’s was somewhat more established.

The other thing they had in common was mediocre offensive lines. The Giants as a whole were close to average quality (barely) in both run and pass blocking. The Jaguars were much better pass blockers but much worse run blockers. The two teams gave up a similar number of sacks and pressures.

The big difference between the two was that the Jaguars had no elite offensive linemen, but no disasters either, at least in the passing game. Luke Fortner was slightly below average in pass blocking while the other four were above average or average. (They were also surprisingly well matched in run blocking - all of them were poor.) The Giants on the other hand had one elite tackle, Andrew Thomas, one horrendously poor one in Evan Neal, and three average to below average players. Both teams lost in the Divisional Round, but the Jaguars put up a great fight against the Chiefs while the Giants were “smoked” by the Eagles and unable to do anything on offense. Trevor Lawrence was pressured on only 29.5% of his dropbacks in the Kansas City game, while Daniel Jones was pressured on 45.7% of his in the Philly game. That’s a difference the Giants need to narrow if they are to take the next step.

Remembering the weak link theory view, the Giants’ offensive fortunes last season were more likely to be sabotaged by the weak link (most often Neal, but not exclusively) than they were helped by the virtuosity of Andrew Thomas. The Giants’ objective for this season, then, does not have to be that Evan Neal become a Pro Bowl-level tackle. That would be nice but it is not necessary. If he becomes just an average or somewhat better than average NFL tackle, that will be enough. Then if the Giants can get average to above average play from rookie John Michael Schmitz while Mark Glowinski and Ben Bredeson (or whoever gets most of the snaps at left guard) just play at an average level, that may be sufficient to take the Giants’ offense a long way. The Giants shouldn’t ask too much of Schmitz as a rookie (see Tyler Linderbaum’s 53.5 pass blocking grade as a reality check), just that he be average in pass protection and a plus in run blocking. (In 2022 he had an 81.0 PFF pass block grade but a 92.4 run block grade, per the 2023 PFF NFL Draft Guide.)

Effect on quarterback play

To see the effects of the Giants’ leaky pass protection in 2022, let’s compare the performances of Daniel Jones and Trevor Lawrence as a function of their offensive line’s play:

Data from Pro Football Focus

Data from Pro Football Focus

No quarterback wants to be pressured - both Jones and Lawrence did better with a clean pocket than under pressure. There were several major differences between Jones and Lawrence last year, though, and one similarity:

• Jones was pressured a lot more often (42% of dropbacks) than Lawrence (29%).
• Lawrence completely fell apart under pressure: A 91.2 PFF passing grade and 104.7 NFL passer rating when kept clean, vs. 29.5 PFF passing grade and 53.7 NFL passer rating when pressured. 25 TDs/5 INTs when kept clean and 5 TDs/8 INTs when pressured.
• Jones wasn’t affected nearly as much by pressure: 82.6 PFF passing grade and 99.1 NFL passer rating when kept clean, vs. 48.7 PFF passing grade and 78.8 NFL passer rating when pressured. 11 TDs/4 INTs when kept clean and 6 TDs/2 INTs when pressured.
• Neither QB was affected all that much by blitzes.
• Under pressure, Jones took an average of 3.63 seconds to throw (TTT column on the far right), while Lawrence got the ball out on average in 3.21 seconds. Whether Jones was less decisive or just didn’t have anyone open as often as Lawrence is difficult to say.

2022 was no fluke, either. Jones has shown similar splits for clean vs. pressured situations throughout his Giants’ career. He may be bland, but he is a cool customer - so much so that he was able to do almost as much with worse pass blocking in front of him than Lawrence was able to do. Jones has yet to show that he can be really elite when kept clean - that 82.6 grade, while very good, pales next to Lawrence’s 91.2 grade - and where that ceiling is will be one big thing to watch for in 2023 with what is expected to be an upgraded set of receivers.

How does Jones do under pressure compared to the rest of the league? Here is perhaps the first scatter plot ever created that has Jones side-by-side with Patrick Mahomes, courtesy of Jonathan Fortgang of PFF. It shows QB expected points added (EPA) when under pressure for two outcomes: sack + scramble plays vs. passes:

No one makes lemonade out of lemons better than Josh Allen, and next best is Mahomes, but right there with him is Jones.

Still, Jones like other QBs does better when not pressured. If the Giants’ can just get their OL to the point that they have one elite and four adequate to above average starters, and Jones begins to see clean pockets as often as Lawrence and many other QBs did last year (Tom Brady, 80.0%; Aaron Rodgers, 73.3%; Joe Burrow, 72.2%; Jalen Hurts, 70.3%; Dak Prescott, 68.8%; Lamar Jackson, 65.9%; Patrick Mahomes, 65.7%; Josh Allen, 64.7%; Justin Herbert, 64.5%), that’s when Joe Schoen will know that the offensive line is good enough.