I'm Over This Justin Herbert Hype

So far, he's more Carson Wentz than Patrick Mahomes

The Justin Herbert hype train is officially out of control. 

There are a couple good reasons…

… one, he looks the part. At six-six and 250 pounds, with a 4.7 forty-yard dash and an arm that can throw footballs into the atmosphere, the Los Angeles Chargers signal caller has every attribute you look for in a quarterback. He’s bright too, winning the “academic Heisman” at Oregon. Pretty much straight out of central casting - Margot Robbie in Barbie levels.  

… two, he has the stats. In his three years since being drafted sixth overall in 2020, he’s slung a lot of pigskin. He took over the starting job two games in - when team doctors nearly killed the guy ahead of him - and never gave it back. His 49 career starts in three seasons are a league record, just one of the marks he’s set.

… three, he’s got expert approval. An ESPN poll of execs and scouts called him the 5th best quarterback in the league; other prominent football sites agree:

• The Ringer has him 2nd

• Pro Football Focus has him 4th

• NFL.com has him 5th

… and four, he’s got the contract. This week, he signed a quarter-billion contract extension. Billion with a b. He’s set to start making $50M a year in 2025, the most of any NFL player ever. 

I’ve always been a bit skeptical of the hype because I’m a reflexive contrarian - so this week I checked myself, and did some homework. I came away convinced the crowning is more than a bit premature. 

Herbert’s production is hardly unique. His league records in fantasy stats - yards, touchdowns - are volume creations; a product of having thrown 200 more passes than any QB in his first three years. If we look at his efficiency and the production of his offense, it’s above-average but not special. He looks a lot more like Carson Wentz than Patrick Mahomes.

Let’s dive in. 

The obvious Herbert pushback is he hasn’t won much. He’s just 25-25 in his career and 0-2 in playoff games¹. Last year ended with the most embarrassing choke job in NFL history, giving up a 27-0 FIRST HALF lead while playing a pretty sloppy game (45.6 QBR; 6.3 yards per throw). 

Sure, wins are not entirely a quarterback stat. Herbert doesn’t play defense and his teammates that do, suck at it. In his three seasons, they’re 23rd, 29th, and 23rd in points allowed. 

But what he is responsible for is offense. He plays every single snap on offense, calling plays and touching the ball each time. I am a big fan of evaluating quarterbacks by the points they score: taken over a longer term, it boils away most bullshit in quarterback evaluations to their very simple job - score, score again, and score some more. Since points per game have been stable-ish with time, it allows us to compare across eras too.   

And Herbert / the Chargers are just… not amazing at scoring points. Since joining the league, his teams have ranked 18th, 5th, and 12th in the league in that category. That’s one great season and two mediocre ones. But let’s keep an open mind. Why might this not be his fault?

Option 1: His team sucks around him

As Tom Brady’s wife, put it: “he can’t ****ing throw and catch the ball at the same time”.

Well said. But I don’t think this applies to Herbert. His running back, Austin Ekeler, is easily among the top 10  in the league (ESPN’s poll says 10th). He has two elite wide receivers in Mike Williams and Keenan Allen. His offensive line has elite talent, including two 2021 All-Pros. Pro Football Focus ranked it the 17th best unit last year, about average. 

Option 2: They got unlucky

A lot of NFL results can be explained by bits of random luck. Balls bouncing the wrong way on a fumble. Penalties at the wrong time. Bang-bang calls that go the other way. Those will show up more in points scored than other metrics because it is more inherently variable. But looking at more all-encompassing metrics of offense tells a similar story. The ultimate catch-all public advanced offensive stat is DVOA - and the Chargers ranked 19th, 4th, and 15th. 

Option 3: It’s normal for his age

You hear this a lot around Herbert, that he’s the first quarterback to compile whatever stat by his age. Second-most touchdowns ever in first three seasons! Most yards ever! So, the logic goes, he could be a top-5 quarterback because of what you think he will do, not what he’s actually done. This argument is really moving the goalposts, but I was curious and checked it out anyway. 

The below graph plots all similarly drafted quarterbacks² since 2011 on their rookie-contract performance for two metrics: points per game, and QBR, ESPN’s all-in-one measure of quarterback performance. Herbert is in red.

Three things come out pretty clearly in this graph:

• QBR (or quarterback play) has a very close relationship³ to points per game. Quarterbacks drive offenses! 

• Second, Herbert’s accomplishments are not extraordinary (pretty close to the pack)

• You best believe that’s Top Right Patty separating himself from the pack again

Here’s that upper right quadrant in a nice table form:

Among his most elite peers, Herbert is just above-average. His extraordinary number of games played give him elite total stats but the per-game ones are par for the group and only Kyler won less. Even his elite passing yards per game is largely explained by his 40.1 passes per game, the highest ever and 20%+ more than most here. All the rankings with him ahead of Lamar Jackson - whose whole career is just putting up points (29.0 ppg) and winning football games (76%) - is crazy to me.

I don’t say any of this to imply Herbert sucks. He’s one of the best quarterbacks drafted in the last 10 years, and he could - and will - get better from here. With LA’s options, I’d have given him the same contract. 

But, to date, he’s just not that guy, pal.   

1. I count the 2021 Week 17 game against Vegas. It was win-or-go-home for both teams, on Sunday Night Football.

2. In the first two rounds of the draft. I used 2011 as the endpoint because it was the first year of the new rookie wage scale, which massively shifted team ability to build around quarterbacks.

3. The single-variable regression has an R² of 75%, which is nerd-speak for saying the QBR of your quarterback explains roughly 75% of the difference between your actual points per game and league average.