In pursuit of finding the secret to predicting an NFL player’s success, I took a look at the most analyzed position in all of football—the quarterback. To do this, I utilized Pro Football Focus’s unique quarterback statistics. Pro Football Focus, better known as PFF, is a media outlet that prioritizes the usage of analytics in the football world. PFF tracks and grades every football game from low-level D1 college football, all the way into the NFL. This in-depth analysis offers grades on every single play of every game, and the entire season as a whole. For the quarterback position, their unique statistics include:
BTT(Big Time Throws): a pass with excellent ball location and timing, generally thrown further down the field and/or into a tighter window
BTT%: the percentage of attempts that are BTTs
TWP(Turnover Worthy Plays): a pass that has a high percentage chance to be intercepted or a poor job of taking care of the ball and fumbling
TWP%: the percentage of attempts that are TWP
ADJ%(Adjusted completion percentage): Adjusted Completion Percentage – the % of aimed passes thrown on target (completions + drops / aimed)
Score: Each statistical category rank added together. Similar to golf, the lower score is better. (BTT%rank + TWP%rank + ADJ%rank)
By doing this PFF eliminates the “Oh, well, he played against weaker competition in college,” argument. These stats are specifically based on aspects of the game that are completely in the quarterbacks control that they did well, and not so well.
For the pool of players, I chose quarterbacks that were first-round draft selections or that had received significant playing time in the NFL. First sorting it out by year, I ranked each quarterback by their performance in BTT, TWP, and ADJ%. After that, I moved every quarterback from the years analyzed into a group as a whole and ranked them best to worst in the same categories. I then added each ranking together for every player to come to a combined score. For example, Russell Wilson finished second in BTT%, third in TWP%, and second in ADJ% to come to a final score of seven (2+3+2=7).
Are PFF’s unique stats an indicator of overall career success?
Defining success in the NFL is a rather difficult thing to do because of the subjective nature of the question. Where do you draw the line between failure and success? Is it winning a Super Bowl? Winning an MVP? Or simply sustaining a long career? These different levels all mean “success” to some degree, and of the player pool nearly all of them fill in somewhere between these different levels of success.
The original thought was that PFF’s unique stats would be easily translatable to NFL success. To my discovery, this was not the case. My findings indicate that there is no clear correlation between where a player ranks in these statistics and success in the NFL. When players were put into a grouping as a whole, there were no trends. They could’ve been put in a random order and you wouldn’t be able to tell the difference.
In his first three years, Patrick Mahomes quickly became the face of the NFL—winning a Super Bowl and the MVP award. Oddly enough, he did not fare too well in these rankings. Of the 30 players that were analyzed in this study, Patrick Mahomes finished 16th with a score of 50. This was worse than Brandon Weeden, Dwayne Haskins, Paxton Lynch, and Blake Bortles.
Another example is with current Raven’s quarterback Lamar Jackson. Even with winning the Heisman trophy at Louisville, Jackson faced harsh criticism when preparing for the NFL draft. He was an afterthought in the quarterback-packed draft of 2018. Starting the final seven games of his rookie season, Jackson went 6-1 and gained the momentum needed to propel himself to a MVP trophy in 2019. If you were to only base your opinion of Jackson off of my formula you would be shocked at his performance in the NFL. Jackson ranked 29 out of 30, and 19 or lower in the three statistical categories.
Can PFF’s unique stats be helpful in deciphering between players in a specific draft class? Finding a “sleeper”?
While determining specific levels of NFL success cannot be determined from these statistics, I did find that in some years, there is a trend of lower draft picks
The first year we can see this in is the 2012 quarterback draft class.
Andrew Luck and Robert Griffin III were the unanimous number one and two draft selections; there is no disputing that. But aside from them, why not Russell Wilson? Across these three statistical categories, Wilson performed extremely well in comparison to his draft counterparts. This is not to say that it was clear and obvious that Wilson would become the player he is today, however, it is worth noting that he should’ve had much more hype around him during the draft process.
Sure there was a lot to question with Wilson, but to be as proficient as he was in these statistics meant that he possessed elite-tier accuracy, decision making, and playmaking ability. By the numbers there is no explanation for why Wilson was selected behind Brock Osweiler and Brandon Weeden.
In fact, it wasn’t even close. Aside from scoring the best, Wilson distanced himself from these two by seven and 14 points. Proving he was far better than the both of them in every category. This is the perfect example of why scouts need to close their mouths, stop drooling over height and physicality, and start paying attention to the numbers.
The second year I want to look at is the 2017 draft class.
This topic of conversation has gained plenty of traction over recent years due to the success of both Patrick Mahomes and Deshaun Watson, honestly, it’s been exhausted to this point. The central point of discussion has been, “Why was Mitch Trubisky drafted over Deshaun Watson and Patrick Mahomes?”
Statistically, there is not an argument to be made for Mahomes as he performed shockingly close to Trubkisy in this formula. However, for Watson, there is a clear argument to be made. As if winning a national championship wasn’t enough, the quarterback from Clemson finished first in his draft class for BTT%, TWP% and ADJ% and was still the third quarterback selected.
If a quarterback is notably proficient in all three of the statistical categories in comparison to the rest of the prospects in the given class, it is reasonable to assume that he should be drafted higher than what he was projected. While this example isn’t as drastic as the Russell Wilson scenario, it still provides a similar trend and suggests that Watson should have been taken ahead of Trubisky.
The third and final is the 2019 quarterback draft class.
While this quarterback draft class has only had three seasons to prove their worth in the NFL, one player has already proved to be better than anyone could have imagined. Gardner Minshew, has been nothing short of electric with his time in the NFL. The sixth-round draft selection spent his time bouncing from community college, to Eastern Carolina University, and finally arriving at Washington State University as a graduate transfer.
Gardner Minshew has primarily been known for his infectious energy and mullett, when in reality he should have been known for his stellar play on the field. Of the draft class, Minshew finished first in both TWP% and ADJ%. While his BTT% is conceringly low, Minshew possessed the attractive qualities that make a sleeper in the NFL draft. Only being excellent in two of the three categories, his numbers in the two are enough to suggest he likely should have been taken earlier than he was drafted.
To this point, Minshew is still brutally underrated for where he was selected. In 21 NFL starts, Minshew has thrown for 5,530 yards, 37 touchdowns, and has a quarterback rating of 93.1%. He ranks better than Drew Lock, Dawyne Haskins, and Daniel Jones in nearly every single statistical category. Again, this is not to say blindly follow statistics when selecting a quarterback, but if they performed better in comparison to the other draft prospects, there is a high chance he is much better than assumed.
After crunching the numbers, it became clear that there is no correlation of PFF’s unique stats being a predictor of greatness. For lack of better terms, it’s just too difficult. Coaching, teammates, city, and player development can all play a factor in a players ability to find success in the NFL.
However, through these findings I discovered that it may be useful as a tool to find sleepers or to compare two players amongst one another in the draft. Russell Wilson, Deshaun Watson, and Gardner Minshew are all evidence of this, and provide examples of how to identify similar players in years to come.
*all stats pulled from PFF.com*