Until now, using college statistics to predict the NFL performance of quarterback prospects has been a largely futile endeavor. Previous research indicates that except for completion percentage, conventional college statistics, such as passer rating, don’t predict how well prospects will do in the NFL.
The best explanation is that different levels of competition and different types of offenses distort these statistics. For example, Baker Mayfield has better statistics than Josh Allen in part because Oklahoma’s system is more quarterback-friendly than Wyoming’s.
One good solution is to adjust each quarterback’s statistics for the quality of his teammates and opponents, as Football Outsiders does for their QBASE projections.
An alternative approach, which we use, is to classify quarterback prospects by their attributes. First, we investigated whether prospects with experience running a pro-style offense in college are more likely to succeed in the NFL. In the 1980s and 1990s, pro-style experience was essential.
Most of the successes, such as Peyton Manning, had pro-style experience, whereas most of the busts, such as Dan McGwire, did not. But today, many good young NFL quarterbacks, such as Jared Goff and Dak Prescott, played in spread offenses in college. As a result, pro-style experience is no longer a good predictor of NFL success.
In light of that, we wondered if there is another measurable characteristic that should translate to the NFL. We found one in a Doug Farrar article on SI.com about functional mobility. Farrar explains that some elite NFL quarterbacks, such as Aaron Rodgers and Russell Wilson, are mobile but use their mobility to make better throws rather than running to run. Alternatively, other elite quarterbacks, such as Tom Brady, can move in the pocket and read defenses well enough to handle pressure effectively.
Therefore, we look for prospects who don’t run often but are efficient when they choose to run. We also look for prospects who can avoid sacks. Sacks are especially bad because they signal immobility, a slow release, trouble reading defenses, or any of the above. Ultimately, we find that functionally mobile prospects are more likely to succeed in the NFL.
That isn’t to say that intangibles, such as character and the eye test, aren’t also important. But functional mobility, in addition to scouts’ opinions, can help NFL teams do a better job drafting quarterbacks.
Below are our projections for the 2018 quarterback prospects expected to be drafted in the first three rounds. We make these projections, which originally appeared in a Football Outsiders column, by using a statistical model to adjust ESPN’s Scouts Inc.’s grades for functional mobility and accuracy.
We project each prospect’s NFL Adjusted Net Yards per Passing Attempt (ANY/A), which is a Pro Football Reference statistic that’s similar to passer rating. An ANY/A of 3.50 is poor, and 7.50 is elite.
Quarterback, Our ANY/A
Sam Darnold, 5.93
Josh Rosen, 5.32
Baker Mayfield, 5.27
Josh Allen, 5.02
Lamar Jackson, 4.97
Kyle Lauletta, 4.91
Mason Rudolph, 4.80
Luke Falk, 4.65
Like NFLDraftScout.com, we believe Sam Darnold is this year’s best prospect. We also find it fascinating that in recent years, our projections have increasingly agreed with those of scouts. If this trend continues, it will no longer be necessary to supplement scouting grades with functional mobility. Instead, scouting and analytics will have come to an agreement.
–Jeremy Rosen is a doctoral student of economics at Georgetown University. Alexandre Olbrecht is an associate professor of economics at Ramapo College of New Jersey and the Executive Director of the Eastern Economic Association.