(09-18-2016, 12:00 PM)stupac2 link Wrote:[quote author=Roberton3 link=topic=15645.msg171724#msg171724 date=1474221800]
[quote author=stupac2 link=topic=15645.msg171720#msg171720 date=1474221172]
You should convert on ~half of them if you don't suck at it, but apparently Oregon sucks at it because 20% is abysmal.
Or they were just unlucky. It's a small sample. If you have a 50% chance of converting and you have five attempts every game, then in roughly one in five games (18.75% of the time) you'll do as badly as the Ducks did or worse.
[/quote]
Well, right, but if you're going to go for it literally every time then you probably don't think your true chance of making it is ~50%, because then it's barely better on average but the variance is unacceptably high. I don't really know what the true percentage should be, but you'd probably want something where the odds of losing points compared to PATs on ~4 TDs are <10%. The Ducks still could've gotten unlucky, it's really impossible to ever rule it out, but I think going 1/5 is bad enough to make it seem like a terrible idea.
Also, wasn't Freeman injured early in the game? You'd think with him out the odds would change significantly enough to reconsider the idea.
[/quote]
Depends. IF you are supposed to be the better team, then you will want to decrease the variance. If you are not supposed to beat your opponent, then you will want to increase the variance.
Also, theoretically, if you always go for 2, over time you should get better at it with practice (while your opponents will not be especially good at defending the 2 pt conversion), so you ought to be better than an average expected number of points.
BC