Skip to main content
ABC News
Patrick Mahomes’s Numbers Have Regressed. But That Doesn’t Mean His Skills Have.

No one who has watched Patrick Mahomes over the past two seasons will dispute that he’s a special player. In his first season as a starting quarterback, he passed for 50 touchdowns (tied for the second-most all time) and won the MVP. But just in case we needed another reminder of his greatness, Mahomes gave us further proof that he’s part alien by passing for 340 yards in a blizzard against Denver in Week 15 — the fourth-most passing yards in a snow game in at least the past 20 seasons.1 Yet some consider 2019 to be a setback for Mahomes.

Heading into the season, a debate raged over Mahomes’s chances to repeat his historic 2018 production. Some argued that mean reversion was inevitable, while others were optimistic that Mahomes could defy history and turn in another MVP-caliber season.

Who was right? Definitive answers are typically fraught, but with two weeks left in the season, we can answer the question with decent confidence: Regression won (as usual!). But what exactly does it mean when we say a player regressed? And what does 2019 tell us about Mahomes’s future?

In football, statistical regression describes situations where an outlier performance in one year is followed by production closer to league average in the following season. Mahomes’s touchdown rate is a prime example. In 2018, he threw a touchdown on 8.6 percent of his passing attempts. This season, Mahomes has thrown a TD on 5.4 percent of his pass attempts, a number closer to the league average of 4.8 percent.2 While he has still performed better than average, it’s clear that even a superstar like Mahomes is not immune to the gravity well that is regression to the mean.

It’s not as if Mahomes has had a bad season. He’s behind only Lamar Jackson in Total QBR on the year. If the season ended today, his season would rank as the 25th-best since 2001, in terms of QBR. Moreover, a deeper look at Mahomes’s QBR splits shows that he’s been basically the same QB this year as he was last season. According to measures that are predictive of future performance — like QBR from inside the pocket — he hasn’t changed much year-over-year (80.9 QBR in ’18 vs. 76.3 QBR in ’19).

Instead, it’s been Mahomes’s performance in the unstable things like red-zone passing that have vacillated. But that’s to be expected. Brian Burke was one of the first analysts to show that red-zone passing performance is probably best thought of as simply a subsample of a player’s overall passing skill. And when we examine subsamples — also called splits — randomness tends to take center stage. Mahomes’s QBR inside the 20 has dropped from 89.7 in 2018 all the way to 40.5 in ’19 — below league average.

This mirrors a similar but less precipitous drop in production outside the pocket — another facet of QB play that is highly unstable year-to-year. In 2018, Mahomes crushed the competition when he rolled out, throwing across his body and using no-look passes to incredible effect. Mahomes’s QBR was a stellar 86.7 last season on those rollouts, but this year, it’s a more pedestrian 60.5.

A final area where he’s struggled relative to his MVP season is against man coverage, yet another split where QB performance is often unstable. Mahomes’s QBR is nearly 20 points higher against zone vs. man this season, whereas in 2018, the gap was just 3 points. Reading too much into these splits would be a mistake. Mahomes’s ability to create out of structure probably hasn’t worsened dramatically in a year, and he didn’t suddenly forget how to throw close to the end zone. Man coverage didn’t somehow become his kryptonite. He’s just been less lucky than he was last year.

This should bring some comfort to worried Chiefs fans (if there are any actually out there). Mahomes’s play has regressed, yes, but the chances of sustaining an almost 9 percent TD rate were always vanishingly small. The areas that he has done well in this season are facets of QB play that we would expect to continue — both into the playoffs and into next season. Conversely, the areas where he’s struggled are noisy and infected with randomness. With evidence-based reasons to be hopeful for a deep playoff run, Mahomes truthers who dismissed his potential regression to the mean as too pessimistic back in August can now embrace it as a source of optimism.

Check out our latest NFL predictions.


  1. Data from Armchair Analysis. 2000 is the earliest year for which the database has weather data.

  2. Calculated over the past 10 years.

Josh Hermsmeyer was a football writer and analyst.