Go For Two
When a team is down by 7 and scores a touchdown late in a game, coaches will almost always kick the extra point instead of “rolling the dice” and going for two. Going for two is considered risky but it is actually statistically more favorable to go for two when the clock shows two minutes or less left in the game. When a team attempts a two -point conversion, their odds of winning are 36.9% compared to 32.5% after making an extra point.
There is a 48% chance of success on a two-point conversion in the NFL. Conventional wisdom says that because the odds are below 50% it’s smarter to go for one than to go for two because the expected amount of points from kicking (1 point) is more than the expected amount of points when going for two (.96.) However, more information needs to be factored in late during games.
The assumptions made before calculating the probabilities are:
A) Both teams are of equal skill level
B) The possession after the TD is potentially the final point scoring possession of regulation.
C) There is approximately two minutes left in the game.
D) If the kickoff following the score is not an onside kick the team will start at the 27-yard line (NFL average.)
E) Onside kicks are recovered at the 50
F) All extra points are successful
The decision tree below shows the probabilities associated with going for two:
Tree Explained: There are two possible outcomes after going for two, it either fails (52% chance) or it is successful (48%.)
In the case the two point conversion is successful there is a 65% chance the game will end without the opponent scoring resulting in a win.
In the scenario where the two-point conversion is not successful, the team may attempt an onside kick. If the onside kick is not successful (80%) a loss is resulted; if it is successful (20%) there are two possible outcomes: the team scores (65%) resulting in a win, or the team does not score (35%) resulting in a loss.
Win probability = (.52*.20*.55)+(.48*.65) = .369
Here is the decision tree for the extra point:
The opposing team can score on the next drive which would result in a loss (35%) or they will not score and the game will go into overtime which would give the team a 50/50 chance of winning.
Win Probability= .65*=.5= .325
Going for two late in games is not risky; it is actually statistically safer. While the difference may only be 4.4%, it would be foolish to not take the best odds no matter how small the difference is.