[AaronBrown]

It's all the same mathematical theory, but it plays out quite differently in the different games.

In draw poker, the most common bluff is not to draw at all. You can do this with any hand. It can never be betrayed by the cards, no one has any information about your hand other than what they infer from your draw and betting. Therefore, you can decide to bluff from the beginning of the hand, and never have to change your mind.

In 7-card stud, you can try to bluff from the beginning, but your upcards may not support it. You can bet strong early, then get cards that couldn't help you if you started with a high pair in the hole, three of a kind or three to a straight flush. Or other player's upcards could prove you were lying. But if you decide to bluff later in the hand based on your four up-cards, your last hole card will not betray you. However, it's possible that other people have cards in the hole, or noticed folded upcards that you forgot, that give away your bluff.

In Hold'em you have the seven card stud problems, plus the problem that the last card is exposed. You have to think carefully to put together a consistent bluff. Hold'em is a game that rewards semi-bluffs much more than pure bluffs.

If you decided to do a pure bluff before the hand started, you would normally just call preflop. You don't want to advertise a monster hand, you want to represent that you caught a good flop. Plus this gives you the option of switching gears and pretending you were slow-playing.

Ideally, there's a small pair on the flop, say 7s. You can go crazy, acting like you hold A7. That works unless someone holds 7s. Two of a suit is a dangerous bluff, because you might not get the third match on the turn or river. That only makes sense if there are several straight possibilities as well.