good time for a combo bet v. a TAG? or do those bets not exist....

[whitelime]

The bet 100% exists as you define it ("sometimes worse will call, sometimes better will fold"). It's just your understanding of poker/game theory is weak and it makes it seem like you don't know what you're talking about:

"Okay, but let's say for instance that I bet this river. I have a decent approximation of his range, but I really don't know for sure. However, what I DO know, is that in this situaiton (or other similar situations), he will sometimes call with worse hands (Ax), and sometimes fold better hands (99, JJ-AA).

I can only guess at his hand range, however, I am sure that this bet will function in two-ways always in this situation.

How then, can this not be a two-way bet in this situation?

I don't know for sure ever if I am value-betting or bluffing in this scenario. See what I'm saying?"

When you say stuff like this, it seems like you understand what you're talking about from a definition standpoint, but are confused about how to apply it to a poker game.

You're only thinking half the equation here, the good half. OMG this bet has two good functions b/c sometimes worse hands call and sometimes better hands fold. Yeah, what about when better hands call? Just because a bet serves both functions of sometimes making worse hands call and sometimes making better hands fold, doesn't make it a +EV bet.

Basically, you need to understand that based on the frequencies of the possible outcomes that can happen, that a bet is either +EV or -EV in a vacuum (let's disregard future metagame EV right now). Try and just focus on this. Value bets, Bluffs, and "value bluffs" can each be +EV or -EV. This should be the deciding factor in the end on whether or not you make the bet (disregard potentially more +EV check/bets).

If you want to start factoring in future metagame, see strassa/jman's posts for the 2 biggest metagame advantages of making a "value bluff".

The way to determine what to do in your situation is to figure out rough frequencies of what he'll do with hands that beat/tie/lose to you, when you bet $x and when you check. Then calculate what the best play is. In 99% of situations, it's pretty obvious what the best play is, and every poker player I know intuitively makes his decisions (no one actually works out the math, though you should try it out once b/c it's something you should be able to do and I think it would help your poker/game theory a lot). In the situations where it's close, you can look to potential future metagame implications as a deciding factor in making a certain decision.