Super Bowl predictions

Seattle won, which was expected. 

As the playoffs progressed, I went back and looked at the model I developed. I streamlined it a lot, and went back to 2015 to see if it was true then, too.   It was!

Here's what I did:

• In week 10, I looked at the 4 teams from each conference who had the highest point differential at that point in the season. 
• in all but one year, one of those 8 teams won the Super Bowl. The exception was in 2018, when New England won; they were the fifth best team in the AFC. (I should note that the 5th best did make the superbowl a few times, but this was the only win). 
• Taking the 5 best in each conference solves this problem

• At season's end, I took those teams and looked at the average points per game that were allowed on defense (for the entire year)
• some of the teams didn't make the playoffs.  That's okay, any of the remaining teams are all we need to consider 
• other teams made the playoffs, but because they weren't in the first 10, they weren't considered - None of them have ever won the Super Bowl. 

• Last, I sorted them by PPG allowed, and the team with the lowest PPG allowed is the expected winner. 
• The playoffs are a little weird, and its hard to call a winner.  In the 11 years I considered:
• 2 times, the superbowl champion was the best team (so we knew at the start of the playoffs)
• 3 times, the next best after the wildcard weekend won
• 3 times, the next best after the divisional playoffs won
• 3 times, it went to the conference championship
• There was only one year where the best team didn't win; the second best did.  That was in 2022 when Kansas City beat Philadelphia on the last drive. Everything else seemed right to that point.

So, essentially, we always know that the superbowl winner will come from that group, and we can assess the likelihood of a team winning as the playoffs progress...