QUOTE (samaroo @ May 22 2015, 09:37 PM)

Okay, more math. Here's a counter-point to the "teams should go for two" thinking posted in a discussion on Football Outsiders.

So, maybe coaches won't go for two more? I know I've put forth data for both sides of the argument, but it's the offseason, and there's nothing else to do. Personally, I think (and hope) that teams do go for two more often. But that's just because I think FG's are boring, especially PATs.

If I understand his math, I think he doesn't understand his math!

To make it easier, let's just look at three PAT's in 1 game, and let's use 94% success rate for XP and 47% for 2 pt conv.

Team A kicks it all three times. Their expected value is .94x1 ptx 3 times = 2.82 pts

Team B goes for 2 three times. Their expected value is .47 x 2 pts x 3 times = 2.82 pts.

In a three TD game, the "variance" has a chance to correct itself. Sure a bigger sample would help smooth it out, but the guy act as if each game is going to be decided by a point. A 2 point miss on try one and you can still get it on try two and be where you would be if you kicked it twice. The variance corrects itself. (I know that's not the correct stat term but that is to put it in the same lingo) I don't know how many are, but I would guess that 90%+ of the NFL games are decided by more than 2 points. This cancels out the variance difference 90% of the time.

I don't think this changes much at all really.

The real takeaway, IMO, is this.

Last year, the Eagles scored 54 TD's. Parkey made all of his XP, but if he was avg and only made 99.% of them, he would have made 53.5 of them, so either 54 or 53, so maybe one missed XP.

This year, if we kick it every time again, our expected made XP is 54 x .94 = 50.7, or roughly three more missed XP's for us all to talk about after the games. Realistically maybe only one costs us, but it sure will put more focus on the kickers.

I don't think that is the NFL's intention but that will be the result.