There's a chasm between Mac Jones' stellar college numbers and the lukewarm forecast that qualitative evaluators give for his future in the NFL. And it forces a question that quarterback-needy teams have to answer: Which will best predict his pro career?
Statistically, Jones is one of the greatest college quarterbacks of his generation. At Alabama in 2020, he recorded the single-highest QBR season in the metric's history, dating back to 2004. He sits atop a leaderboard of best seasons littered with top draft picks and elite NFL QBs; Kyler Murray, Joe Burrow, Tua Tagovailoa, Russell Wilson and Andrew Luck trail Jones. Clemson's Trevor Lawrence, the consensus top-ranked quarterback in the 2021 NFL draft, never finished above fifth in QBR in a single season. Again, Jones' 2020 ranked No. 1 among all seasons.
Jones is considered by many evaluators -- including Todd McShay and ESPN's Scouts Inc. -- to be the fifth-best signal-caller in this class, after Lawrence, Ohio State's Justin Fields, BYU's Zach Wilson and North Dakota State's Trey Lance. Though Mel Kiper Jr. predicted in his latest mock draft that the 49ers would select Jones with the No. 3 overall pick, he wrote that "I'm not sure I see a superstar when I watch him on tape. He's solid -- and extremely accurate -- but not spectacular."
Jones was not considered a candidate to be selected in the top three until the 49ers traded up nine spots. Suddenly Jones, who many believe would thrive most in Kyle Shanahan's system, became the favorite to land in San Francisco. But whether he is best off with the 49ers and whether the 49ers are best off with him are not the same question. And while expectations for his draft slot have shifted in recent weeks, the vexing disparity between his statistics and the consensus scouting evaluation remains.
There have been several criticisms levied against Jones, including that he's the least athletic of the first-round quarterbacks and less of a running threat than his peers, that he doesn't have high-end arm strength, and that he played with elite talent around him that would elevate any quarterback. Is it possible evaluators are underrating Jones? Let's separate fact from fiction and put numbers to those criticisms:
Mobility
Jones did use his legs substantially less than any of the other top four quarterbacks. He rushed -- either by design or on a scramble -- on 8% of his career action plays, well below Lawrence (16%), Wilson (19%), Fields (27%) and Lance (34%). And if we focus on scrambles, those made up 3.5% of his career dropbacks, which again was behind Fields (11%), Lance (6%) and Wilson (6%) but was on par with Lawrence. He did scramble less frequently in 2020, however.
Those rates are below average for rushing and average for scrambling for quarterbacks drafted in the previous 10 years, but they're not totally unheard of. They are similar to what Jared Goff and Gardner Minshew posted in college. Jones' career scramble rate is higher than those of Josh Rosen, Drew Lock and even Ryan Tannehill, who went on to be a productive runner in the NFL (Tannehill did have a higher designed-run rate, though).
Also, not all mobility involves crossing the line of scrimmage. Jones was sacked on just 3% of his dropbacks last season and only 13% of pressures against him were converted into sacks. Both of those numbers were lower than any recorded by Lawrence, Fields or Wilson, though each was only barely better than Wilson's rates. It could be a sign of Jones' ability to move in the pocket, his offensive line's quality or his ability to get the ball out quickly. No matter what, if Jones is avoiding sacks -- even when under pressure -- that is a good sign for his NFL prospects.
Overall, the numbers suggest Jones' running (or lack thereof) wouldn't make him an outlier, but it's unlikely to be an asset the way it will be for other quarterbacks in this class. He can succeed without being a threat to run -- there are plenty of examples of that in the current NFL (including Matt Ryan, whom Kyle Shanahan has had success coaching), but it just means that he'll have to make up for that relative weakness with strengths elsewhere.
Quality of teammates and situation
I'm not sure just anyone could put up the numbers Jones did in 2020 -- even his former teammate Tua Tagovailoa, who went No. 5 overall last year despite injury concerns, didn't record quite the same production Jones did in a similar offense. The point certainly has some merit: Jones played with a Heisman Trophy-winning receiver, another receiver who will go early in the first round and a stellar offensive line.
And while QBR adjusts for opponents, it does not adjust for teammates. So Jones is getting credit for playing against the SEC, but is not being debited for his offensive line or receivers.
However, Real Plus-Minus -- which incorporates adjusted on/off metrics and advanced box score stats -- does account for both the quality of opponent and teammates on every play. Jones' 2020 season ranked No. 1 in RPM last season and trailed only Burrow's 2019 for the highest RPM season in the metric's three-year history. While the quarterback of an offense as successful as Alabama's is always going to look good in any metric, Jones' RPM was elite despite the fact that he was playing with the metric's No. 1-ranked wide receiver in DeVonta Smith and its No. 1-ranked offensive line. In other words, even accounting for the exceptional talent around him, Jones significantly elevated the offense.
Situationally, Jones was the beneficiary of play-action, which he used on a massive 52% of his dropbacks, which was more than any of the other top FBS quarterbacks in this class -- the FBS average for qualifying QBs was 32% -- presumably because his team was so often winning and/or a threat to run. Play-action is well-documented as an advantage for the offense, and that might have inflated his statistics though he was also ranked second in QBR without play-action.
Making sense of the qualitative and quantitative
No matter how deep you dig, there are significant elements of quarterback evaluation that cannot be captured through current college statistics. The qualitative is necessary.
How do we reconcile the differences between the scouting and production? By marrying them. In our newest projection model of college quarterbacks we put Scouts Inc. grades and age -- along with several statistical measures including a player's best QBR, most recent QBR, and production split out between running and passing -- into the model, therefore weighting each component by its predictive ability (and interaction with the other variables). It's the stats and the scouting, blended together.
The result? Our model projects Jones as a tier below Lawrence, Fields and Wilson, who all have either a 63% or 64% chance of outperforming Jones in expected points added per play in 2021. The projections put Jones neck-and-neck with Lance, with Jones slightly ahead of Lance in Year 1 but with a slight preference for Lance by Year 4, a shift due to the fact that Jones is nearly two years older. (Surprisingly, the model narrowly prefers Texas A&M's Kellen Mond, who is projected to be a second- or third-round pick, to both.)
In other words, of the five quarterbacks likely to go in the first round, the model considers Jones the fourth best in the short term and fifth best in the long term, though the percentages express the significant uncertainty surrounding the projection of quarterbacks from college to the pros.
There are three main reasons for the relatively tepid forecast for Jones:
His Scouts Inc. grade of 90 trails the other top four quarterbacks in the class. Even with the knowledge of his college statistics, the qualitative carries significant weight.
His incredible collegiate play mostly came in one season (with a few strong games in 2019, too). What we've found is that multiple seasons of excellent play -- like Lawrence finishing in the top 10 in QBR but never to Jones' level in 2020 -- is a better predictor of success at the next level than a single elite season like Jones had.
His age. Jones is the oldest among the top five quarterbacks -- the 22-year-old is about six months older than Fields, roughly a year older than Lawrence and Wilson and almost two years older than Lance -- and that does matter. It's not everything -- Burrow, for example, was 23 when he was drafted -- but all four of the other major quarterbacks in this class achieved their accomplishments at a younger age.
Altogether, the model paints a more modest picture of Jones compared to Lawrence, Wilson or Fields, the last of whom will likely be available for the 49ers at No. 3.
QBR was a big fan of Jones in 2019 and 2020, and there's still a great chance he's able to have a strong career in the NFL. It's just less of a chance than for the other high-end QBs in the 2021 class.
Paul Sabin contributed to this article.