You're ahead of any hole cards he could have. Where is the semi-bluff part? Only if you catch higher than 7 on 6th will you be a dog... and the pot is big enough to cover you regardless.

QJ3 are gone, so you have at worst about a 33% to catch a lock on 6th- including the two 3's that he doesn't know are a brick. Even if he would put you on the nut wheel on 3rd street, can he continue with was it most likely a rough 98.. in case he's wrong?

He has about the same to catch a card that "improves" him to an 87 with a 7-draw. He's more likely to be stuck with a rough 9 on 6th street.

Did you mean this to only represent what the pro MIGHT be putting you on, as far as the "semi-bluff"?

Otherwise, I'm confused by your semi-bluff motivation for check-raising. This seems like a pure value-raise to me.

isn't villain's hand better than Hero's if villain has two cards less than 7 in the hole? 987xx rather than J832A ?

i think he's calling it a "semi-bluff" because he would prefer the opponent folds.

anyway, i think leading out is better than the c/r unless you have a monster read that your opponent is stupid enough to bet in his spot -- which apparently this "pro" was.

No, actually it isn't. the Jack with the 8-smooth draw is about a 6% favorite on 5th (ignoring the board display advantages that add to your lead imo). If the Jack catches bad and villan catches good on 6th, you're only about a 3:1 dog.

How are you going to play even 2978(3A) against J82A (xx) when you were check-raised on 5th? How aggressive will you play as villian on 7th if you catch less than 7-smooth, if hero is rammin' and jammin' from 5th-river?

Your analysis here is correct, and the same point is made later in the article: So it's never claimed that the c/r is a true semibluff, but it's clearly not a value bet either.

According to Wikipedia, a value bet is "A bet made by a player who wants it to be called (as opposed to a bluff or protection bet). This is typically because he has a superior hand that he expects to win at showdown, or a very good draw for which he can increase his pot equity by more than the amount of his bet." IMHO this definition is a little too narrow, since many value bets are made in a spot where the hero would prefer the villain(s) to fold, but isn't likely to get that result. To be a "value bet" it has to be a bet hero would make without any fold equity or advantage to be gained on future streets (such as free cards, free showdown, etc.) In this article, the hero is attempting to gain the greatest fold equity and would not make the bet purely for value given the tournament factors.

Really? I think that's an idiosyncratic definition. This was not a value bet exactly because we prefer a fold. It's true that there's a sort of inbetween bet you can make, where you are rooting for a fold but you think you probably have value if you're called, but it really needs another name.

Really? I think that's an idiosyncratic definition. This was not a value bet exactly because we prefer a fold. It's true that there's a sort of inbetween bet you can make, where you are rooting for a fold but you think you probably have value if you're called, but it really needs another name.

[/quote]
What you're talking about sounds like a protection bet when you're ahead but vulnerable so betting charges the draw. Here we are on the draw, but it's got so many outs that it's a statistical favorite. This situation comes up all the time in razz, and there are times when it's really a value bet if hero's brick is showing. In this case, betting a (A3)82J would have been a value bet.