Annals of efficient market theory, NFL coaches edition
By 
OK kids, here’s a football stats problem.
Situation: You’re trailing 27-24, four and a half minutes to go, fourth and two at your own 33. You have one time out left.
If you go for it and make it, your win probability becomes 38%.
If you go for it and get stopped, your win probability becomes 20%.
If you punt, your win probability becomes 21%.
Is it statistically correct to conclude that, if your probability of converting the first down is higher than 5.55%, you should go for it?
What if your quarterback is Aaron Rodgers?
Mike McCarthy is paid $6 million per year to figure this one out.