The best QBs to have for the next three seasons

A recent ESPN Insider article by their illustrious panel of experts ranked the 32 teams in terms of which NFL franchises are in the best shape for the next three seasons. One of the categories was quarterback, with teams ranked #1-32 based on their QB situation for the next three seasons. It was no surprise that Aaron Rodgers ranked #1, but following behind him at #2 was Derek Carr. That’s right: these experts would rather have Carr for the next three years than players like Russell Wilson, Drew Brees, Andrew Luck, Ben Roethlisberger, and the last two MVP winners, Matt Ryan and Cam Newton. However, I’d argue that none of these is the craziest omission, which is the inexplicable choice of Derek Carr over Dak Prescott.

Disclaimers

• Why is this on a Seahawks website? (a) Because all of the bad arguments that were used against Wilson early in his career (e.g., he only looks good because of his teammates, he’s only efficient because he throws at low volume, the plays he creates with his legs outside of the pocket don’t count) that have proven to be bad arguments as Wilson has maintained a high level for five seasons are now being applied to Prescott; and (b) Derek Carr is lame.
• This might look silly in a year if Carr improves as much as he did between 2015 and 2016.

The Statistics

The film comes later in this piece, but before anyone accuses me of cherry-picking individual plays, statistics measure the result of every play of the season. In Derek Carr’s three seasons, he has never played better than in 2016, and Prescott’s 2016 rookie season was substantially more efficient than Carr’s 2016 (and no, it is not easier to be more efficient when throwing at a lower volume):

The difference between the two in 2016 was enormous. For example, QBR is scaled so that a player’s QBR represents expected win percentage based on QB play. Taking each player’s 2016 QBR at face value, a team headed by Carr would be expected to win about 62% of their games, compared to 82% for Prescott (and yes, QBR is predictive of team success - possibly more than any other measure of QB play).

In 2016, Prescott was better and younger, and going forward he carries a substantially smaller cap hit. We can’t even point to Prescott enjoying the benefits of Dallas’ excellent offensive line as an advantage over Carr’s situation because the Raiders also have a great offensive line. In fact, in my recent study of offensive line play, the Raiders appear to be better at pass protection than the Cowboys.

Before I get shouted down by the “WATCH THE GAMES!!!” crowd:

Let’s move to the film.

The Film

Short summary: Dak can do everything Carr can do, but Carr cannot do everything Dak can do. Dak’s mobility offers advantages in play calling diversity (bootlegs, moving the pocket, read option, etc.) as well as the opportunity to pick up 1st downs by scrambling when no one is open downfield. Below I will highlight some of what Dak offers that explains his large statistical advantage over Carr even though they play with similarly talented teammates.

Moving the pocket

One of the reasons people underestimate Prescott is because so many of his throws look easy. Here are a few plays from one game (the last play shown is a 4th and 1):

Now watch how unnatural Carr looks when moving the pocket:

Carr can’t make plays like this with the consistency that Prescott can:

A mobile QB opens up a lot of plays and puts extra stress on the defense that QBs who are not rushing threats simply cannot. Does this make some of the mobile QB’s throws easier? Yes. Is this a reason to discount the mobile QB? No! Yards, first downs, and touchdowns gained by taking advantage of a QB’s mobility count just the same as plays from the pocket. This is one of the advantages of a statistic like QBR, where a QB is rewarded for the plays he makes (including those with his legs) rather than how he looks doing it.

Rushing value

One of the most frustrating things about popular discourse surrounding quarterbacks is the complete disregard of a QB’s rushing value. In terms of expected points added through the ground, Prescott was #3 in the NFL and Carr was dead last out of 30 qualified QBs. Carr gained 5 first downs (0 TDs) on 38 rushing attempts (13% of rushes for 1st downs), while Prescott gained 21 first downs (6 TDs) on 57 rushing attempts (37%). This is somewhat puzzling given that Carr ran a faster 40 and had a higher vertical leap at the combine than Prescott, so he’s not exactly at an athletic disadvantage.

Here are five plays in which Dak rushed for first downs, the same number of first downs Carr rushed for all season. These five plays gained 81 yards, or, in other words, more yards than Carr rushed for all season.

Accuracy

I leave the reader with a cutup of some Carr throws from 2016, where while watching Carr’s tape I was so shocked at the number of passes Carr missed while sitting in a clean pocket (the inner Seahawks fan cries at these pockets) that I was compelled to put together a compilation (warning: extremely long):

Every QB misses throws, but the number of throws Carr misses from a clean pocket is astounding. It is also noteworthy how many missed opportunities to create positive plays by scrambling there are in the montage above. The Raiders passing offense was still very effective (#4 in passing DVOA) because they almost never take sacks, Carr almost never gets pressured (all QBs perform better when not pressured), and he gets to take a lot of deep shots to talented WRs from a clean pocket, but there are a great deal of QBs who could perform just as well if not better if given the pass protection that Carr has.

Conclusion

Carr’s career stats are bad. Very bad:

Whenever I’ve pointed this out, people tell me to watch the film. Well, now I’ve watched the film. He’s not a bad quarterback (anymore), but he’s so completely useless as a rushing threat and misses so many unpressured throws that Prescott is far more impressive on tape. In this case, the stats don’t lie.