How Well Do New Quarterbacks Play?

Chicago Bears fans are going to need to be patient when Justin Fields starts playing, whether that’s during the first game of 2021 or some later date. While it would be amazing if he stepped into the game and gave us a perfect indication of the quarterback he would turn into, that’s highly unlikely.

How unlikely? Well, I was curious what typical first-round quarterback play looked like, but I am simply done talking about the 2017 class. It’s still too early to pass judgment on anything more recent, however. Fortunately, from 2009 to 2016, twenty-five different quarterbacks have been taken with a pick inside of the top 40. That’s a pretty robust sample to work with.

Interestingly, they are nearly perfectly split into two groups. 12 of these quarterbacks have a career ANY/A higher than Jay Cutler’s and either a median indexed passer rating above average for the league or at least one Pro Bowl. Most have all three. These are good quarterbacks by anything but the most demanding of standards.

By contrast, another twelve have a median indexed passer rating below average for the league, a career ANY/A lower than Jay Cutler’s, and have never earned a Pro Bowl. These are bad quarterbacks, essentially. Wow. That is an amazingly well-divided sample.

Overview

How did these twenty-four players look in their first eight starts? They clearly separated themselves, even if it was not always a dramatic separation. Across their first eight starts, the good group had an average passer rating of 83.13 and an average ANY/A of 5.88. In other words, they looked like pretty competent quarterbacks, but they did not set the world on fire. The best was Colin Kaepernick, who had a 98.7 passer rating and a 7.60 ANY/A. Still, all six players north of the 83.13 passer rating mark (and the 5.88 ANY/A mark) ended up on the “good” quarterback list, even if EJ Manuel (rating 82.66 and 5.63 ANY/A) came close, and he clearly didn’t turn out to be a good quarterback.

Meanwhile, the bad quarterback group has an average passer rating of 72.3 and an ANY/A of 4.67. Of the eight quarterbacks who played below those marks in their first eight games, only Stafford and Goff ended up on the “good QB” list. This absolutely does not mean that the first eight starts are destiny, however. No matter how I worked the numbers, the strongest correlation I could build between any single stat for the first eight starts of a player’s career and his eventual aggregate performance landed at a 0.42 (0.60 is usually the threshold for significance), and that’s largely because of how much variation is outside of the quarterback’s control. There is still one thing that jumps out at me.

Good players start. Eight of the twelve quarterbacks on the “bad” list did not see their first start until after the second week. Only three of the good quarterbacks (Kaepernick, Goff, and Bridgewater) waited. In the case of Kaepernick, Alex Smith was on the team and playing adequately. Bridgewater was taken with the last pick of the first round and he replaced Christian Ponder, another recent 1st-round pick. Goff clearly needed work, clearly came in still underprepared, and needed a coaching change in order to rehabilitate his career (even now it’s unclear how much of his success was him and how much was simply a phenomenal coaching effort put into place around him). While there are exceptions, if a player sat instead of playing, it was usually not a good sign.

Sacks Matter

The bad quarterbacks in this group were sacked on 7.1% of pass attempts, whereas the good quarterbacks in this group had an average sack rate of 5.8%. However, the actual results are more interesting than simply saying “good quarterbacks get sacked less, even early on.” There’s not really a correlation between early sack rate and later performance, but broad categories hold up, and those two categories are “concerning” and inconclusive”.

In order, the quarterbacks sacked more than 8% of the time are Lynch (12.3%), Goff, Sanchez, Ponder, Tebow, Manziel, Mariota, Bortles, and Gabbert (8.4%). Most of those are bad quarterbacks, or quarterbacks who still struggle with problems being sacked, and Goff is a bit of an outlier because while his accomplishments consistently show that he became one of the “good” QBs, there’s no doubt that a very different sort of player took the field in those first few games.

On the other hand, the lowest sack rates (all better than 5.8%, again in order) belonged to Carr (3%), Weedon, Dalton, Locker, Luck, Winston, Bradford, and Newton (5.6%). That is a very mixed group, even if the next three names are also “good” quarterbacks—Stafford, Wentz, and Tannehill. So, sacks are not a perfect predictor. It seems like it might matter if a quarterback is sacked at a high rate early on, but a lack of sacks does not prove that a quarterback is going to turn out just fine.

Interceptions Matter, too

The worst interception rates across the first eight starts belong to Freeman (6.3%), Ponder, Sanchez, Bortles, Stafford, and Locker (3.6%). That seems right. However, the lowest interception rates belong to Kaepernick (1.4%), Manuel, Wentz, Tebow, Gabbert, and Mariota (2.3%). As we saw with low sack rates, that is a very mixed group. Once again, while it’s probably a bad sign if a quarterback throws a lot of interceptions early on, it’s not necessarily evidence he is going to play well in the future just because he avoids these errors.

There is a really simple explanation here. At first, many new quarterbacks are set up for success at the level of the drive or the play, as opposed to being asked to win the game. Additionally, there’s not a lot of tape on what these new players are going to do, and they have every reason to play within the system. Even flawed quarterbacks can play moderately well under those circumstances, and if a player cannot keep the interceptions (or sacks) down under those circumstances, it’s cause for at least mild concern. However, playing well under these controlled circumstances is far from proof that they will continue playing well when the training wheels are off.

Completion Percentage is a Little More Consistent

A very familiar pattern emerges with completion percentage. The bad quarterbacks group had an average completion percentage just under 56% in their first eight starts. Meanwhile, the good quarterbacks had a completion percentage just over 60% in their first eight starts. The eight players with a completion rate worse than 56% were (in order from best to worst): Josh Freeman, Brandon Weeden, Matthew Stafford, Jarod Goff, Christian Ponder, Mark Sanchez, Blaine Gabbert, and (with a comical 46.9% completion rate) Tim Tebow. A low completion percentage is clearly a bad sign, and only Stafford and Goff — who seem immune to being pigeon-holed by these early metrics — managed to recover.

However, this time things are a little different on the positive side. The nine players with completion rates above 60% were (from low to high): Cam Newton, Derek Carr, Teddy Bridgewater, Andy Dalton, Paxton Lynch, Colin Kaepernick, Blake Bortles, Carson Wentz, and Marcus Mariota. With the exception of Bortles and Lynch (both of whom struggled in other areas), those were all good quarterbacks. At this point, I also have to admit that I don’t have a full eight starts for Lynch, because his career functionally ended after five starts.

F8 =/= Fate*

Ultimately, the first eight starts of a quarterback’s career do not really show who he is. There are quarterbacks like Stafford and Goff who struggled early and who then went on to find success. Then there are quarterbacks who avoided the easy mistakes when everything was made simple for them who never found that extra gear once the training wheels were off. Ultimately, the first eight games seem to serve as an early warning sign for bad performances and not much more.

However, there’s an asterisk. Running through permutations, I did come up with a single combination of traits that had a notable correlation with a quarterback’s eventual stat line. Looking for quarterbacks with a sack rate under 8%, an interception rate below 3%, and a completion percentage above 60% had a positive correlation (0.66) with a quarterback’s eventual median passer rating, career ANY/A, and ability to reach the Pro Bowl.

This is far from a predictive metric, and the measures involved are a little crude. Nonetheless, while the first eight starts are not destiny, those are the marks I will be looking for when Fields finally gets the nod.