Two-point conversions

[RXB]

You're John Fox. You're down by 19 early in the 4th quarter. Touchdown. You go for two. Not smart.

The thinking, of course, is that 8+8+3=19 which ties the game. The problem is: the likelihood of converting both two-pointers, and getting a FG, and stopping the other team from scoring so that you outscore them by exactly 19 is so low that it's not worth the risk.

So, you fail on the two-pointer. Then immediately get a turnover and score another TD which you convert normally. Instead of being down by five points you're down by six. Which allows an Indy FG to become a much greater blow to your chances of winning the game than it would otherwise have been.

[RXB]

And right on cue... a Bronco turnover leads to an Indy FG.

[tucker6]

Early in the 4th, you really can't expect to get the ball back more than 3 times in the time remaining. So if you're Fox, they need to get 19 pts in those possessions. It is statistically easier to make two 8 point TD's and a FG than it is to score 3 TD's. Using your argument, all the failed conversion did is put them down by 6 points instead of five after the second TD.

[RXB]

Early in the 4th, you really can't expect to get the ball back more than 3 times in the time remaining. So if you're Fox, they need to get 19 pts in those possessions. It is statistically easier to make two 8 point TD's and a FG than it is to score 3 TD's. Using your argument, all the failed conversion did is put them down by 6 points instead of five after the second TD.

8+8+3=19, which only ties the game. You still have to score yet again to win. Accomplishing all of that is clearly less likely than just scoring three TD's. (The chances of converting a pair of consecutive two-point conversions is ~20-25%.)

Regarding "all the failed conversion did is put them down by 6 points instead of five after the second TD"-- when Indy got their FG it was a nine-point game late in the 4th quarter instead of an eight-point game. Which is huge.

The way to play it is to just kick the extra points on the first two TD's. You go for the two-pointer after the third TD, either to try to take a three-point lead or to try to tie the game if the other team has kicked an intervening FG, which Indy did last night. More than 12 minutes remained when Denver scored their first 4th-quarter TD and they ended up getting three more possessions. The only time a team should consider going for a pair of two-point conversions is when they have almost no chance of getting the ball back more than once.

The timing of that Denver turnover that set up the Indy FG was amusing to me as I was just finishing the last paragraph about that possibility when it happened.

[RXB]

I'll give another example of poor game theory by football coaches regarding late-game conversions. A team is down by 14 points quite late. They score a TD. What should they do now?

Well, what they almost invariably do is kick the point-- which is wrong. They have a better chance of winning the game by going for a two-pointer. If they make it, they can win outright with another TD followed by a simple kick. If they miss, they still can try another two-pointer after a subsequent TD for a chance to tie. The math is clearly in favour of this strategy but good luck trying to get coaches to heed it.

[Quagmire]

I agree with you, 2 pt conversions should attempted more often. Interesting article here:
http://www.boydsbets.com/nfl-two-point-conversion-success-rate/

So what does this tell us? Well, NFL teams should start going for 2 more often. The problem doesn’t lie with the performance of the team or the confidence of the coach in his players, the problem lies with job security. The NFL is a league that prides itself on winning and time-honored traditions. Any coach who did something this drastic would instantly be subjected to scrutiny. Any mistake would be blown out of proportion and his job would be on the hot seat. This is the same reason more coaches don’t go for it on 4th down and short.

[Valuist]

I agree with you, 2 pt conversions should attempted more often. Interesting article here:
http://www.boydsbets.com/nfl-two-point-conversion-success-rate/

So what does this tell us? Well, NFL teams should start going for 2 more often. The problem doesn’t lie with the performance of the team or the confidence of the coach in his players, the problem lies with job security. The NFL is a league that prides itself on winning and time-honored traditions. Any coach who did something this drastic would instantly be subjected to scrutiny. Any mistake would be blown out of proportion and his job would be on the hot seat. This is the same reason more coaches don’t go for it on 4th down and short.

I interpret those stats as good reasons to NOT go for two, unless absolutely necessary.

There's been a total of 543 attempted two point conversions, and according to those percentages, only 239 were successful. That's 0.88 pts per attempt. If any NFL team had a kicker that could only convert 88% of extra point kicks, they wouldn't be around very long.

[RXB]

I interpret those stats as good reasons to NOT go for two, unless absolutely necessary.

There's been a total of 543 attempted two point conversions, and according to those percentages, only 239 were successful. That's 0.88 pts per attempt. If any NFL team had a kicker that could only convert 88% of extra point kicks, they wouldn't be around very long.

I agree generally.

What should happen is that teams should be aware of the probabilities and use that awareness to their advantage. Fox's two-point decision yesterday was a poor one that goes against the probabilities; it was nice coincidence for the purposes of this thread that one of the potential risks of his gamble came back to bite him.

But in the dying-minutes 14-point deficit scenario that I mentioned earlier, the probabilities favour going for two. If you get the second TD, based on the percentages you'll lead outright almost 43% of the time (get the two-pointer on the first TD, then kick the go-ahead point after the second TD), tie about 25% (fail on the first two-point try, make the second one), remain behind in about 32% (fail on both).