Go for 2 down 14?
On Saturday, Winnipeg scored to take the deficit from 14 points to 8 points (before the convert). There was 2:27 left on the clock and they trotted out their kicker (who hadn't been kicking well!) to kick the extra point and make it a 7 point game. I would have much preferred Winnipeg to go for 2, purely as a mathematical advantage. Going for 2 in this situation makes it possible to win the game in regulation, as opposed to kicking two extra points, and taking the game to overtime.
Let's assume Winnipeg gets 2 point conversions at a rate of 56% (league average).
Let's also assume Winnipeg hits PATs at a rate of 89% (league average).
Lastly, let's also also assume Winnipeg wins 50% of the games that go to OT.
There are 4 possible scenarios in a situation like this.
1) Wpg gets the 2 point convert & kicks the PAT to go ahead by 1. 49.8% (0.56 * 0.89)
2) Wpg gets the 2 point convert & misses the PAT, and thus are tied. 6.1% (0.56 * 0.11)
3) Wpg misses the 2 point convert but gets the 2nd 2 point convert, and thus are tied. 24.6% (0.44 * 0.56)
4) Wpg misses the 2 point convert twice, and loses. 19.4% (0.44 * 0.44)
Here, Winnipeg wins in regulation 49.8% of the time, loses in regulation 19.4% of the time, and goes to overtime 30.7% of the time. Therefore, they win the game 65.15% of the time if they choose this strategy. Let's look at the strategy they chose.
1) Wpg makes the PAT & makes the PAT, and thus are tied. 79.2% (0.89 * 0.89)
2) Wpg makes the PAT & misses the 2nd PAT, and loses. 9.8% (0.89 * 0.11)
3) Wpg misses the PAT & makes the 2 point convert, and thus are tied. 6.2% (0.11 * 0.56)
4) Wpg misses the PAT & misses the 2 point convert, and loses. 4.8% (0.11 * 0.44)
In this case, Winnipeg wins in regulation 0% of the time, loses in regulation 14.6% of the time, and goes to overtime 85.4% of time. Therefore, they win the game 42.7% of the time, a stark drop-off from the 65% if they had chosen the first strategy.