How stupid can I be?

Am I stupid because my highly-touted (mostly by me) Super Bowl prediction system has predicted the wrong Super Bowl winner in each of the last two years?

Answer: Probably.

On the other hand, am I stupid because I don't learn from my own system?

Answer: I want to believe this is the somewhat better answer.

Last week I predicted that New England would be the Super Bowl winner, but surprisingly, only seven of the twelve predictors in the system favored New England, with the other five going to the Giants.

Before this year’s Super Bowl the system predicted the last 14 of 17 flat-out winners.

Over the years we've had a lot of fun with the Super Bowl prediction system, and it has worked quite well. I'm sure there's a luck element involved. But I did some research after this year's miss, and the bottom line is I was too stupid to see something the system was telling me.

I looked back at all the times when I predicted the winner based on winning only six or seven categories. That could be 7-5 like this year's Super Bowl, or 7-4-1, 6-5-1, or even 6-6 with a tiebreaker. It turns out that when a team manages to win only six or seven of the twelve categories, it is basically a toss up. The system is 50-50 on all of those games, going back 18 years or even when I go all the way back to the first Super Bowl. Over the last 18 years, the system has correctly predicted the winner twice and missed the winner twice when a team is favored by the system but only has six or seven categories. It has a record of six wins and seven losses when you go all the way back to the beginning of the Super Bowl.

What I should have said, if I were smart enough, is that this game is a toss up. The New England Patriots win seven categories and the Giants win five, but this is a toss up.

And I should have pointed out that if I were a betting man, which I am not, that the system is clearly saying take the Giants and the 12 points.

That's how stupid I can be.