Of course, neither Brady nor Manning play defense, so a "head-to-head" comparison is a bit short-sighted to begin with. Jake Plummer was 2-0 against Brady, and Jay Fiedler put up better numbers in his four head-to-head matchups against Manning. Somehow, I doubt a "Fiedler vs. Manning" post would generate much debate.
So how do we begin to truly evaluate which quarterback has performed better in 13 Manning vs. Brady games? Well, let's start by looking at their WPA, EPA and passing success rate (SR) during each of their games:
We can see that Brady has a fairly significant edge in EPA, and a decent one in WPA. He's also a bit ahead in success rate, but both have put up excellent rates in that category—for reference, those marks would rank fourth and sixth among 2013 quarterbacks. So case closed—Brady is the winner, right?
As you might expect, it's not that simple. First of all, plotting their EPAs graphically reveals a fairly eye-opening trend:
Brady and Manning have switched places, with Peyton getting the upper-hand in EPA in five of the past six matchups, after Brady took five of the first six. Unsurprisingly, that nearly mirrors the trend of Brady dominating early in the win column, before Peyton surged back in the mid-to-late 2000s.
Brady's overall EPA is much higher because of a few wild negative-EPA outliers in Manning's history. Manning's standard deviation in EPA is 12.36 from a 5.70 average, while Brady's is a much neater 6.81 from a 8.08 average. With a garish cumlative minus-25.57 EPA in the two playoff games against the Pats, it's no wonder that Peyton developed a reputation as a postseason choker.
Speaking of those two games, they point us towards an important factor to consider when comparing production. Manning's playoff clunkers were indisputably atrocious, but it's hard to imagine any quarterback in the league experiencing much success against those suffocating early-00s Patriots defenses in a Foxboro snowstorm.
Thus, it might be instructive to measure each quarterback's production against the defenses they had to face. The graphs below overlap each quarterback's EPA vs. the defense's average passing EPA per game from that season, playoffs included. So the bigger the gap, the more the quarterback outperformed against that defense's season average:
Obviously the comparison is imperfect, as Brady and Manning do not play each other every game of the season, so the defense's numbers are almost entirely based off performance against other teams. Moreover, they do not consider the context of a specific game—perhaps Manning's recent EPAs are higher because his team has generally fallen behind early, necessitating a pass-heavy approach that pumps up his numbers.
However, they do provide some standard to compare their stats against, rather than accepting them in a vacuum. And when we do use this standard, we find that Brady's edge in average EPA is not really an edge at all. Manning has faced better defenses, as the Patriots have on average produced an EPA per game of about 2.5 points lower than the Colts and 2012 Broncos. Thus, Brady's average EPA is roughly 6.82 points higher than the defensive average, while Manning is actually a smidgen better, at 6.92 points above average.
That's basically a wash, and debunks both the surface stats that portray Manning as having put up much better numbers than Brady in their head-to-head meetings, and the initial EPA reading that gives Brady the advantage.
One other factor to consider is "clutchness," which we can measure through both WPA and context-specific stats. Again, Brady has the edge in WPA, which fits in with his reputation as the better crunch-time quarterback.
However, we can take their fourth-quarter head-to-head stats (per Pro-Football-Reference's Game Play Finder) and find an argument to the contrary. Though Manning's 10 to 6 touchdown edge is nullified by virtue of having thrown 59 more passes, he still has a comparable YPA and completion percentage. Moreover, you could also point to the fact that Brady has thrown the same number of interceptions:
This isn't particularly surprising when you consider that Manning has had a couple signature fourth-quarter comebacks (2006 AFC Championship, "4th-and-2"), as well as a couple near misses (the last two meetings, the 2003 "goal-line stand" regular season game). Indeed, if you eliminate his minus-0.42 WPA clunker from the 2003 AFC Championship, Manning's average WPA rises to 0.22. Brady's average WPA is...0.22.
It's not really fair to eliminate Peyton's worst performances and anoint him superior to Brady, because those stinkers did happen, and they are forever a part of the rivalry's legacy. But for the most part, we can see that at least statistically, Brady vs. Manning is essentially a wash.
And really, it's fitting that we don't have a definitive answer, because these two quarterbacks are equals in almost every way. We've debunked the notion that Brady is somehow more "clutch" and that Manning owns the numbers, which is really just a convenient mainstream narrative to try and flesh out some difference between the two. There are only a small handful of meetings left between the two, so instead of fruitlessly arguing until you are blue in the face, perhaps this Sunday night would be better spent simply enjoying this wonderful matchup.