By now you’ve had about all the Super Bowl coverage you can probably stand. Some of what I read included Bill Simmons’ pick (as always), Bill Barnwell’s breakdown and prediction, and a summary of previous offense-defense Super Bowls. So I’m not going to pile on much and instead just add a couple of my reactions along with the pick.
I think the idea of looking at previous match-ups that seem similar to the current game is an appealing one. So when Mays ran through previous top-offense versus top-defense games and found that the better defensive team has a slight edge, 5-3, that feels like it should mean something. But the downside to this particular method is that by using rank, you lose information about exactly how good each team is at what they do. There’s also some context that has to come into play. For example, one defensive win was Tampa over Oakland, but that game rather infamously seemed to swing to Tampa because the Raiders never changed some of their calls after Gruden switched teams. But more importantly, by just labeling Tampa as ‘#1 defense’, we’ve lost that they were a really good defense in general, not just that year. On the other hand, the Raiders were the #1 offense that year, but not on any all-time lists. So even though the game was put on the list because the offense and defense were ‘matched’, that doesn’t really mean they were equal. It also obviously ignores half the game, with the match-up between the great offense’s defense and the great defense’ offense. We know that the Broncos have a great offense and the Seahawks have a great defense, but it matters that Seattle is amongst the best while Denver may be the best.
In Barnwell’s post, he mentions injuries as something that could hold Denver back. They could, but the Broncos seem to have been getting along just fine so far. Von Miller has missed time already, and as Barnwell points out the Broncos actually seem to do better without him than with. Rahim Moore has been out since mid-November, so the team is used to playing without him. The list goes on, which makes it sound like there are a lot of injuries, and there are, but evidently it hasn’t mattered too much. Obviously if there are in-game injuries, those could be critical. But I don’t think we’ve seen anything that makes the injury report especially important.
Bill Simmons is taking the Broncos by 3. That’s enough to almost scare you off them entirely. But even with the terrible year he’s had picking games, you have to figure it could go that way.
As Barnwell points out, Seattle was ranked ahead of Denver by a few different systems. For whatever reason, my model disagrees. The Broncos have been number 1 there since I put up my first ranking in week 3. In those other rankings, Denver always drops off because of strength of schedule adjustments. My model takes opponent performance into account, so I think I have that covered, although it could obviously be off. Maybe more importantly, my model inherently takes home field into account; the stats get fed in as coming from the home or away team. So when I predict the Super Bowl, I run it once with Denver at home and then again with Seattle at home, and I average the two to make it neutral field. If either team actually had home field advantage, they would be favored. But Denver is predicted to win by more at home than Seattle would at home. As it turns out, the three models are pretty much in agreement that Denver is about a 1.5 point favorite. The line I see at SBR is -2.5, which makes it kind of a close call, but maybe you would see that as a reason to take Seattle to cover. I personally am leaning a bit towards Denver to win, and winners typically cover, but it’s more of a preference than a strong feeling. The models do like the game to go over 48 points, and I think taking Seattle on the moneyline +115 would be reasonable. They would have to win 46.5% of the time for that to be worthwhile.
So I’ll be rooting a bit for the Broncos, but mostly I just hope the players get to decide the game without any injuries or iffy calls making a big difference. This should be a pretty great game, so I want to enjoy it.