Measuring the unquantifiable traits of successful quarterback prospects

We don't have direct measurements of key quarterback traits like pocket management and processing, but we can gather hints in the charting data if you know where to look

I’ll cop to fact that this analysis could be fairly viewed as an exercise in data mining. The question is whether it’s a reasonable method of finding insights, or more appropriately labeled data fishing. When you’re looking for the numbers and traits that produce elite quarterback play in the NFL, you always have to leverage a small sample of data and an even smaller sample of successful players to that become your targets. These limitations are heavily reflected in my last piece that critiqued being too confident in what the results of a handful of quarterback classes tell us. Now I’m going to turn around and use that same limited data to make my own conclusions about which quarterback stats can predictive a successful transition to the NFL.

I don’t think it’s hypocritical to urge caution in reading too much into recent quarterback prospects successes, and then use those same successes in my analysis. There’s a difference between being open to the larger possibilities and throwing up your hands and giving up. We can only make do with the data we have, and in this case I have access to nine years of college performances (2014-2022) via PFF charting data. The right path, in my humble opinion, is to try and pull insights from what we have to work with, while being open to changing and updating our opinions with new data, and acknowledge the spectrum of potentially successful quarterback prospects is much larger than what we’ve seen to date.

If we confine prospects “hits” to only the young, elite quarterbacks that every team is looking for, we’re ultimately restricting the definitive successes to seven quarterback prospects (2017: Patrick Mahomes and Deshaun Watson; 2018: Josh Allen and Lamar Jackson; 2020: Joe Burrow and Justin Herbert; 2021: Trevor Lawrence). You can expand the definition of success a bit further to include prospects like Jalen Hurts and Tua Tagovailoa, but I don’t think we should be any more confident in their ability to sustain last year’s performance than we were with Carson Wentz and Jared Goff going into their contract extensions.

With the necessary context and caveats clearly stated, we still can find significant, and perhaps predictive, qualities in recent quarterback prospects that signal future NFL success. I already focused on the stability of accuracy going into the NFL, and there are other, less straightforward ways of identifying traits. Specifically, I wanted to find which statistical charting stats could reflect two of the most commonly cited characteristics in quarterback scouting reports: pocket management and processing. These aren’t easy to quantify, but looking back through the results of the last several quarterbacks classes, I found in the past that that prospects’ outperformance of relative efficiency for pressured passing and the ability to scramble and avoid sacks under pressure to forecast NFL success. This analysis is based around four key metrics, all looking at quarterback prospects’ final seasons, all as a percentile of college quarterback seasons (FBS & FCS):

• Pressured scramble rate

• Pressured sack rate

• Pressured yards per attempt (YPA) relative to clean-pocket

• Blitzed yards per attempt (YPA) relative to non-blitzed

How a quarterback responds to pressure in college is the closest thing to replicating what NFL play will be like on a down-to-down basis. The level of talent is more standardized in the NFL, there are no true walkover opponents, or massive talent discrepancies like the top quarterback from major college programs enjoy for much of the season. By digging into how a quarterback plays under pressure, we get a better sense of their inherent skill level that will differentiate them in the NFL, beyond what can be taught and schemed.

EXECUTING UNDER PRESSURE

One of my favorite metrics to look at for college quarterbacks¹ is the percentage of pressured snaps that don’t end in an attempted pass. There are two main outcomes that fall into this category: scrambles and sacks. Both of these stats are fairly stable going from college to the NFL (roughly 0.3-0.4 correlation), and they nicely separate the non-attempt outcomes into positive and negative. The two sacks are also highly correlated to each other, at around 0.5.

I like to plot the percentiles for pressured scramble and sack rates against each other, highlighting the 50th percentile for both, dividing the results into four categories. Below is just this for the final-year seasons from first- and second-round quarterbacks from the draft classes from 2015 to this spring. You’ll quickly notice that many of the elite young quarterbacks in the NFL fall into the upper-right quadrant, or high relative scramble rate and low relative sack rate (when pressured).