Don't we need a large sample size to be able to accurately assign a percentage here? Aren't there countless factors in the game that affect this call?

My God is it hard to get people to understand this. I had a half hour conversation with my brother last night and he refuses to fully grasp the concept here. The crazy thing is that he's read the Advanced NFL Stats breakdown and doesn't dispute any of the numbers.

His argument is that it's a risk/reward scenerio. He's saying that it's riskier to go for it so it just depends on how much risk you are willing to take on. My counter was "riskier from what perspective". To me, if the risk is losing the game, it's clearly riskier to punt.

Anyone else run into this with anyone?

Not sure if it was correct or not, but it is insane that people are labelling this one of the biggest coaching blunders in boston sports history and comparing it to Grady Little leaving Pedro in the ALCS...

It is so funny that people say, "You NEVER go for it 4th and 2 on your own 28!" Isn't it so obvious the correct answer to "do you go for it, 4th and 2 on your own 28?" is IT DEPENDS!

It has already been well outlined what the factors are so I won't go through them again, but i'll just say i'm sure the BB knew the factors and knew that the decision wasn't automatic. He had a play he had enough confidence in given the game situation that would satisfy the EV equation. He probably wasn't even thinking about league averages, he was thinking about the game situation and probably thought the conversion play would work a much higher percentage of the time than any kind of league average.

It turned out it didn't, and so you could criticize his play call or his evaluation of his team's abitlities to execute the play or the Colt's abilities to stop it, but you CANNOT say that the decision to go for it is wrong no matter what.

It tilts me to hear all the stupid people talk about it. Lucky for him, he doesn't seem to care. I have little doubt in my mind he knows how to arrive at the correct decision and simply made an error in his evaluation of his teams ability to execute. Someday, he will probably write a book that will change football forever (unless he decides to pass down his knowledge to someone else and keep it secret). It's funny that a play that failed made me realize he is even more of a genius than I had thought...

if we are basing on a decision being correct or not on the reaction of the opposing team's fanbase then i'd absolutely agree that Indy and the Pats should go for 2 after every TD. (not sure about Minn).

and i'd love to see the Pats score a TD on the Bills and go for 2 to go up 8-0. then have the Bills score a TD (lol, i know) and see if Jauron goes for 1 or 2.

I am just surprised that all the analysts that agree with BB's decision are not making statements like this. Truly, this is the case. Yes, the call on 4th was the correct call, and I even agree with the 4th down play call, but if your plan was to go for it on 4th, then why not attempt to run for it on 3rd down first. I mean if you get 1 yard even, it does 2 things, makes IND use it's final TO (think they would let it run to the 2 min warning, which also leaves NE with a TO instead of using that TO after 3rd because of incomplete pass stops it from getting to the 2 min warning) Then on 4th down after the 2 min warning you could run again needing only 1 yrd or a PA pass (same one you would do at 4th and goal @ the 1)

But did you even PLAY high school football? (I was at a bar after this and actually got that question )

We have a huge sample size on certain things, like the chances an average offense gets the first, or the likelihood that an average offense scores after being given the ball on the 30 or returning a punt. We can make reasonable adjustments given what we know about the teams that were playing.

The thing is, it wasn't a close enough call for that stuff to matter. Assuming the Patriots and Colts both have average offenses going for it was a good call. Once you try to adjust for reality it becomes a better call.

If there is some in-game thing that greatly changes the odds, is there anyone on the planet better suited to make that call than Belichick?

Someone had a good post about this earlier. Most people are really risk averse. Going for it makes one play really important. If you punt the eventual outcome will be decided over a number of less important plays. That is a lot easier for people to stomach, even if the EV is lower.

My cousin said it best last night. We were talking about the play and he said it was obviously the right call because even if NE punted, it's not like Manning can't throw 2 20 yard passes to get to the 30 yard line in 10 seconds.

This would require a much more substantive model to figure out, maybe something like a monte carlo model to chain together tons of randomly generated drive outcomes to see how exactly it all works out. That would require a ton of work though, whereas the down by two touchdowns late situation is easy to understand and conceptualize.

Note that if you are down by 15 and score a touchdown, nobody ever goes for the 2 point conversion on the first TD, they always want to push the "moment of truth" off into the future.

Not surprisingly, TMQ says good call:

http://sports.espn.go.com/espn/page2...7&sportCat=nfl

I was watching this in a bar and thus ran into every fallacy possible and maybe ran into some new ones that haven't been cataloged yet

I thought I had shown that they were inconsistent when i pointed out that as the Colts you want them to kick so kicking for the Pats can't be that bad of a decision. At the least, I thought this showed that either they are wrong to think it is awful for your team to go for it in NE's spot or wrong for being happy to field the punt in Indys spot rather than having to stop them on 4th. I was wrong. The response I got from a local expert was that "Football is not a zero some game. Things can be good for you and good for your opponent and bad for you and bad for your opponent."

Clearly he's referring to some weird situation, maybe late in overtime where both teams need a win a make the playoffs and a team just kneels out the clock to play for the tie. Genius, really.

Then he ruins it by saying the MJD kneel at the 1 wasn't correct because "what if you miss the field goal?"

I never understood that either. Seems to me that we can make better decisions with more information, and going for 2 on the first TD, make it or not, gives us more information for the rest of the game. Not doing that seems insane.

That isn't what he said. He said the kneel was good, but that after wasting the first 40 seconds he would then try to run it in for the score. I don't think I agree (although doing it after a second kneel down makes sense).

Odds are much better after the 2nd TD ldo.

if they ran on 3rd down, clearly, with 2:08 left, indy would let the clock run to the 2min warning. then if they don't get it on 4th there's like 1:54 left...so running takes 6 seconds off the clock, which actually (very very) slightly HURTS NE.

so the clock is not an issue at all when choosing the 3rd down play. (and he couldn't have predicted that he would need the TO. the fact that he called TO itself indicates that something unexpected came up or that he changed his mind. and besides, the TO is relatively unimportant at that point. it only becomes important in hindsight.)

so it's almost totally - what gets the first down more often, run/x or pass/x? for NE, i'm not at all convinced that run/x gets it more often.

Oh, ok. Still disagree though.

Bold is exactly right, but then wrong conclusion.

Waiting is probably the smarter strategy since you'll have more info when you actually need to go for 2 or not. What happens if a safety somehow occurs due to a bad punt snap after you failed on a 2-pt conversion that left you down 9? Now you're down 7, whereas if you waited to go for it, you'd now be down 6 and don't even need a 2-pt conversion if you score.

No point in even continuing the conversation , since this guy is clearly not capable of rational thought. The best thing to do is just invite him to your next poker game.

which is ******ed because your giving yourself less information to work with this way

Okay, I can also see the case why this is true. It tells you whether you still need 2 scores or not.

If you are down 15, then score and go for 2 and miss the information gain for the opposing team is more significant than it is for you.