I recently posted the adjusted QB ratings of the Super Bowl winners going back to 1976. Let’s start by finishing off the list. (As a reminder, 100 is average; 120 means the QB rating is 20% above average.)
1975 Bradshaw 122
1974 Bradshaw 90
1973 Griese 122
1972 Morrall 129
1971 Stabauch 144
1970 Unitas 99 (Morrall 1 regular season start, 133; previous 2 years 91, 128)
1969 Dawson 106 (this is misleading, as Dawson was injured in the regular season, but led the league in adjusted passer rating in six of the previous 7 seasons, and performed at an elite level in the playoffs. He was a superstar, not the decent game manager his 1969 stats made him look like.)
1968 Joe Willie Namath 110
1967 Starr 97
1966 Starr 135
One upshot is that the NFL has been dominated by the passing game for a long time. Even going back to the beginning of the Super Bowl era, when the running game was somewhat more important and passing games were less efficient, the best teams have generally had star QBs and have almost always had at least good ones. Stabauch had a short career but was an unbelievable player at his peak. Starr in 1967 is like Dawson ’69 — his 1966 performance is much more representative of his abilities. Starr was a tremendously good player. Griese is a dubious Hall of Famer but was very good at his best. Morrall was inconsistent but a very good player overall. Namath, as is well known, is one of the most overrated players in NFL history, but he was a good QB at his best. (I should say Namath is also underrated by the NFL rating system. It actually works very well in the contemporary NFL, in which most good QBs complete a high percentage of passes, but it underrates and old-fashioned vertical passer like Namath.)
Now, let’s look at this year’s matchup:
Manning 67/Osweiler 95
Newton, as we’ve discussed, is a better player than his stats suggest given his surrounding skill talent, but even at that the Panthers passing game is well within the historic range of a credible Super Bowl champion. The Broncos, OTOH, are trying to do something without precedent. Osweiler would be one of the 2 or 3 worst QBs of a championship team ever. But he’s not starting the game in any case. The starting QB would shatter the record for the worst regular season QB to win the Super Bowl by a YOOOOOOOOGE margin. Nobody has ever won the Super Bowl with a starter remotely as bad as Peyton has been in 2015, not even when MEN were MEN and real teams GROUNDED and POUNDED.
One might be tempted to say that the stats underrate the performance of Denver’s passing game — we know they have an outstanding defense, but how can you go 12-4 with a passing game that inept in the modern NFL? The answer to this is that the Broncos aren’t really a 12-4 team in terms of quality. Point differential and DVOA predict future W-L much better than win-loss records themselves, and both show the Broncos as more like a 10-6 or 9-7 team. (DVOA puts them right between the Steelers and Jets — that seems about right.) And, just to preempt inevitable arguments that Denver winning a bunch of coin-flip games represents a repeatable skill, no there’s no reason to believe that great defenses tend to overperform their point differential. (The 2013 Seahawks had 13 expected and real wins. The 2008 Steelers were 12/12. The 2002 Bucs were 13/12. The 2000 Ravens were 13.5/12. The 1991 Eagles were 9.5/10. The ’85 Bears were 14/15.) The factors that have caused the Broncos to have a better record than their underlying performance are luck, luck, luck, and luck. And this is pretty obvious if you actually look at the games. I don’t see how Denver’s defense caused Reid and Belichick to make critical high downside-low upside blunders at key moments, caused one of the best kickers in Super Bowl history to miss an extra point, etc. etc. It’s a tribute to Denver’s defense that they dragged a passing game this bad to above-average performance, but above-average is all this team is. And the regular season overperformance was crucial to their playoff wins; a team with margins this narrow really needed the bye and home field in the conference championship game.
Carolina were also not as good as their 15-1 regular season record. But since then they’ve played arguably the two best regular season teams in the league and beaten one by a much greater margin than the 31-24 score suggested and beaten the other 49-15. DVOA now sees them as the genuinely elite team the Broncos aren’t, and I agree. The line has already moved a point towards Carolina, and I think 5 1/2 is still too low.
There is, of course, a caveat here. Manning has been absolutely horrible this year, but he’s also one of the greatest NFL players ever. It’s possible that with two weeks off and knowing that it’s probably the last game of his career, Manning can have one last good game, like he did against Green Bay. But it’s hard for me to see that. The Broncos had one sustained drive each in their two playoff wins, and Carolina is a step up on defense and they’re moving to a neutral field. It’s possible that Manning has one good game left in him, but I’m not sure he’s a lot better than his replacement-level regular season performance at this point.
Is it possible for Denver to win anyway? Sure. As the Tom Coughlin Giants proved, just because you’ve hit tails several times in a row doesn’t mean it can’t happen again. Probably the closest historical precedent is the 1970 Colts, who eked out a 3-point win in a famously blunder-filled Super Bowl with Earl Morall coming in to relieve a washed-up Johnny Unitas. (Evidently, it wouldn’t be surprising to see Osweiler at some point.) Unitas at that point was merely mediocre rather than actually bad like Manning was this year, but Manning is certainly capable of at least an OK game. It’s possible that Denver can keep the game close and again get the breaks and make the plays at the end. But unless Manning can step up his game significantly, I think it’s more likely that Carolina wins with relative comfort.