where a TD puts you up by 1 point then kicking the XP increases your win % by 1.8%. So not converting there obviously costs you at least that much. As far as the change in win % it's not necessarily impacted at all by where the try occurs from. It's still going to be 2 points, 1 point, or 0 points. But yes you would want to factor in the likelihood of success as well.

The issue with this specific scenario (TD puts you up by 1) is that the difference between getting 1 point versus 2 points is overwhelmingly weighted towards going for 2. So much so that even a "bad" 2 point conversion team would usually be better served going for 2 there. It's more a function of how useless 1 point is in that scenario than anything.

Of course in our game that 1 point came back to matter... I even said to my wife in the stands at the time... "The only way that kicking the XP helps us at all here is if we score another TD and then we would be up 9 instead of 8". At the time I didn't think there was any way our offense was putting another TD on the board with our banged up offense but we did. The ND game was full of things that had to break a specific way in order to line up for a ND win. And because we're snakebit (kidding... sort of) those things happened.

If you're interested in reading more about two point analytics I highly recommend this article. It's the one I'm getting my numbers from.