DVOA means efficiency per play and DYAR is total volume. So by football outsiders metrics, Stroud is the 2nd best performing qb this year on a play by play basis and the most valuable overall.
His DYAR is 1031. on pace for 1964 DYAR over this season.
His DVOA is 36.8%, meaning he is 36.8% above a league average QB.
To give context to how impressive this is, I will list a few QB seasons with similar metrics.
Aaron Rodgers in 2020 in his 48 TD 5 INT season had a 33.7% DVOA and 1649 DVOA. Granted this was a 16 game season.
Mahomes 50 TD season was 39.9% DVOA and 2031 DYAR.
So Stroud is playing at a similar level to 2 MVP, HOF QBs best seasons
If you are skeptical about these advanced metrics, Stroud also grades well in traditional ones.
He is 3rd in ANY/A (and the 2 ahead of him have much better offensive rosters), 6th in passer rating and 13th in pass success rate.
He should be in serious MVP consideration, especially with the defenses he has faced so far.
Link to new Football Outsiders
Image of DVOA stats
I really don’t think that’s completely true. Here’s the last ten MVPs.
Jackson’s significant rushing value is not accounted for in passer rating or yardage. Should be considered 1st outright in wins, since he sat out an easy victory with the #1 seed locked up.
Mahomes was a narrow second to Brees in both EPA and rating, however Brees missed a game and had much lower volume (>5000 yards vs <4000 yards).
Newton’s significant rushing value is not accounted for in passer rating or yardage.
All of these ranks are based on players who started at least 9 games. The 2013 cutoff gave me a round decade and avoided having to deal with RB-as-MVP in the form of Peterson '12. Before 2012 I don’t have EPA handy.
It’s pretty clear at a glance that volume isn’t actually that big a deal to voters.
“Advanced” and “Traditional” efficiency stats tend toward agreement, but when they disagree the “Advanced” EPA is a better predictor of the MVP. To be fair to “traditional” efficiency, Manning '13, Jackson '19, and Mahomes '22 were behind only players who missed significant time. The two significant disagreements are:
The case against wins is made most strongly by Ryan winning over four players with at least as many wins and a better record (Brady, Dak, Alex Smith, and Derek Carr) in '16. Mahomes '18 and Rodgers '20 both failed to lead the league in wins but did secure the #1 seed in their conference. The case for wins is made best by Newton '15; I suspect the rather odd Cam Newton MVP award had more to do with voters over-valuing his rushing ability than Carolina’s 15-1 record, but I could easily be wrong.
Overall, being really good at quarterbacking is the best way to win MVP. It’s also the best way to win lots of games and get high ranks in both old and new efficiency stats, so it’s hard to say definitively which of those the voters are actually pinging on; those three measures usually agree. My personal guess is that EPA is the best proxy for what they are thinking; if Stroud finishes up at the top of the efficiency boards but someone else, say Mahomes, has four or so more wins the MVP race might give us a very interesting data point this year.