Suppose you took the top 10 live cash pros and the top 10 online cash pros in the world. Let's not debate who they are, but assume there is a reasonable particular benchmark for assigning who the top players actually are. Let's assume their best game is NLHE and that NLHE is the game of choice for this question. Let's say you analyzed the most recent 100,000 hands of each and every one of these individual pros (and let's just exclude those hands where these pros happened to have played any of the other top pros in these particular hands). You looked at every one of these hands, whether they limped, raised, 3-bet, folded, won, or lost - so long as they made it to the river; they could have folded by the river, so long as they got there.

First question: what percent of these hands do you think they - collectively, as a group of players - were bluffing (let's include semi-bluffs)?

Second question: do you think there are noticeable differences between the percent of time that the live cash pros were bluffing versus the percent of time that the online cash pros were bluffing?

I'd guess the pros as a whole are bluffing about a third of the time on the river (more on the earlier streets), maybe a little less.
Who would they be playing against then? There aren't many whales playing nosebleeds online, whereas my understanding of live games is that games form around a donator. You beat that player according to whatever his tendencies are (i.e. "don't bluff the unbluffable").

If you exclude the the top players they're playing against, I'm sure your results are much more scattered as they'll be playing much more exploitative, whereas vs other "topplayers" their frequencies will be more balanced. (basically what Arty Said )

I am talking about WHOEVER they played against in their last 100,000 hands, excluding the other top 10 pros.

Oh sorry, I didn't see your answer when I first glanced at the response. Thanks. Just seeing what others might guess compared to what I might guess at.