I just came across the article “A Statistician Shares How To Pick Your Super Bowl Pool Like A Champ” at businessinsider.com.  The author of the article, Jill Krasny, asked edgehogs.com statistician, William Briggs, for some advice:

“You want to pick the scores that are most likely to happen, and look at historical information about how score differentials (i.e., pairings) are most realized,” Briggs said. “You shouldn’t pick squares out of the blue that happen infrequently.”

Then she offers this note:

Note: Some people pick the labels on the rows and columns only after all the boxes have been bought, making the game more random. If your office does it this way, and not all do, these statistics will still help you figure your chance of winning.

I would argue that a fundamental rule of the Super Bowl squares game is that you pick a square BEFORE the numbers have been placed on the grid.  Instead of saying “Some people” in her note, she should say “Almost all people.”  (Am I wrong about this?  I’ve never, ever seen the numbers on the board before the squares are filled in.)

The article is still of some use, though, as you get some idea of what your chances of winning are after you get your numbers.  Of course, the whole premise that the article was written on (you get to choose your numbers) is almost never true.

Finally, they looked at the last 2,822 NFL games, but if you’re interested in complete results for over 14,000 games in a pretty heat map grid format, I’ve compiled that here.

Go Pats.

Cheers.