I don’t know for sure that the numbers favored Jax going for two there, but assuming they did, it would be a good example of how to gain an edge using analytics: to overcome cognitive biases.
Most often, the bias we see is loss aversion, but last night, it was the bias that creeps in when a team is trailing by a “round” number (3, 7, 10, etc). The goal becomes getting the game back to a tie score, even though that only gives you a 50% chance of winning (maybe a tick higher in Jacksonville’s case since they were at home).
But at the point where the 2PC became much easier due to the penalty, Pederson correctly recognized that the Jags had a chance to win the game outright, so he took it. The fact that everything worked out perfectly doesn’t prove he was right, but it does show why you make decisions like that