Uhh, this guy's probability equation is right, but the probabilities he uses in it are ridiculous. Does he keep a database of every 2-minute drill situation accounting for field position or amount of timeouts? Even if he does, the sample size of teams getting the ball at their own 30-35 with one timeout left is probably way too small to produce a meaningful probability. So even if that 53% is accurate, he might as well of pulled it out of his a**.
And second of all, there is no way the real probability of the Colts scoring from the 28 was anywhere close to 53%. Personally, I think that their chances at that point were much closer to 80% and plugging that number into the equation it gives the Pats a better chance to win if they punt. I also think that the 60% conversion rate for the 4th down in that situation is also debatable.
So the bottomline is that if you plug more reasonable numbers into the equation, the Pats probably would have had a slightly better chance to win if they had punted. However, Belicheck isn't the moron of the year for going for it, since the probabilities are debatable. If they got the first down no one would be criticizing him right now, and I think everone agrees that the Pats had a decent chance of converting.
And second of all, there is no way the real probability of the Colts scoring from the 28 was anywhere close to 53%. Personally, I think that their chances at that point were much closer to 80% and plugging that number into the equation it gives the Pats a better chance to win if they punt. I also think that the 60% conversion rate for the 4th down in that situation is also debatable.
So the bottomline is that if you plug more reasonable numbers into the equation, the Pats probably would have had a slightly better chance to win if they had punted. However, Belicheck isn't the moron of the year for going for it, since the probabilities are debatable. If they got the first down no one would be criticizing him right now, and I think everone agrees that the Pats had a decent chance of converting.