2-point conversion rates in expected situations

In the Indiana game column, Brian appears to be stating that 2-point conversions have a 47% chance of success.  That's a little higher than I thought it was (I had read that it was around 42-43%), but regardless, I'm curious to know if there is any data out there for 2-point attempts that are a surprise to the defense (i.e., the swinging gate attempts that teams like Oregon like to run) compared to the conventional ones, where the offense lines up normally so the defense knows it's coming.  I would assume that the conventional rate is lower than the surprise ones (it seems like Oregon always converts on those swinging-gate tries), but I'd be curious to see what the numbers actually say. 

Purely anecdotally, it seems to me like teams fail significantly more than they succeed on conventional attempts.  Is the conversion rate of 40-whatever percent being inflated by those swinging-gate attempts?   If the numbers on conventional attempts are significantly lower than surprise attempts, maybe coaches are going for 2 (conventionally) too often - or maybe the swinging gate is underused.