9nop(This is assuming no ties.)

The odds that a 17 choice sequence would be chosen: 100%. (Not that particular 17 game sequence, just any 17 game sequence).

The odds that the first 9 games would line up to the first 7 games of that same sequence: (0.5^9) (Small)

The odds that all 17 games would match up to the sequence: (0.5^17). (256x Smaller)

The odds that the remaining wins turn out to be wins: (0.5^5 = 1/64)

From here on out, this is where binomal probablitity needs to get used. It’s “Odds for Succeeding X number of times within Y trials given these odds for each attempt”.

I used a binomal calculator I found online, so it gives me decimals instead.

5 or more wins out of the 7 games (just in general): 22.656% The exact coin sequence playing out: (0.5)^7

Exactly 5 games matching up (Win Or Lose): 0.16406. (Parsing 9-8 vs 10-7 vs 11-6 vs 12-5 is absurd, just know the better records are less likely).

If we assume coin is wrong for Lions, Giants, and Falcons (2nd time), that means the coin is right 4 out of the remaining 7 times, or 0.5 probability in our favor.

That would mean the coin would be right 13/17 times, probability: 0.1816.

What does this mean? Who knows. All I know is that winning the Super Bowl in Vegas would result in absolute hilarity.