With some hands you want to try and end up HU preflop after a large raise -- AA for example plays much better against one opponent, and you want to charge that opponent to see the flop.

Other hands you want to play cheaply and ideally you would have several opponents in the hand with you. Low pocket pairs for example. You're hoping to limp in with these guys from late position or a blind. 98s. A3s. You want lots of people in the hand in case you hit your lotto ticket.

Sure, you can sometimes mix it up and open raise with A3s or limp with AK, but you're still thinking in terms of these two categories.

But do ALL playable starting hands fall into one of these two categories? Are there any where you want to juice the pot just a little?

Are there any starting hands where you want to put in a 2BB raise with the expectation of having 5 or 6 callers? (or raise to $5 in 1-2.) Or, to put it another way, at a very passive table if you have 77, who not bloat the pot a little if you are pretty sure you can get away with it? If you're happy to have 6 people limp in, aren't you happier to have 6 people pay $5?

Do not limp.

Just no. I keep repeating this until I die.

In general, open limping is a bad idea. Limping in position after limpers is debatable but often OK.

Never and always are bad words in poker.

While you should almost never open limp, I can think of one scenario where you would.

In a tournament, you're UTG at a full table. A couple of players to your left are loose passive and have shown a tendency to over limp after an open limp, the CO and Button are hyper-LAGs and love to punish limpers.

You have AA. Guess what - YOU LIMP - because you want the 2 LPPs to call and at least one of the LAGs to raise big so you can reraise.

Now mind you, if what you want to happen doesn't happen, you're going to have to play appropriately post flop, but occasionally, you can do something non-standard when a specific set of circumstance arises.

But yes, in general, don't limp.

In theory, all hands play better against one opponent. The old theory books that recommended limping/calling with suited hands and small pairs multiway were pretty much wrong.
It only makes sense to overcall/overlimp with hands if your opponents are spectacularly bad at playing in multiway pots and you know how to exploit the situation. Fortunately, most opponents at low stakes are spectacularly bad at playing in multiway pots, but that's not a great reason to get into a bad habit imo. If you play weak hands in multiway pots, you'll just find yourself in really weird spots, and the variance is crazy. If your hand isn't strong enough to raise, and isn't an obvious call when facing a raise (stuff like 99), you're usually better off folding.

I think in 1/2 if everybody is limping and you're on the button, and the blinds are very passive, limping is the right play often. Boom. Just said it.

The question is, would you ever intentionally make a small raise. 3BB in a limpy 1/2 game is basically a small raise. $5 is a small raise. If everyone limps to you and you bet $5, every single limper is going to throw a $5 chip out there. You haven't reduced the field, you've just increased the pot. Is that ever a smart move?

This ^^^. My game changed noticeably just taking this advice. My red line showed a marked raise with the number of limpers that would fold to a raise and I stopped getting myself in to awful spots when they hit with hands they shouldn't be in with.

I went through the next phase of getting annoyed raising a strong hand against limpers makes them fold so maybe I should limp along to try and induce some play but predictably I opened myself up to all sorts of crap again and quickly learnt about variance. Sacked that idea off pretty quickly. If I raise a strong hand and a limper folds so be it. Saying that there is the odd time I'll limp, but in very specific circumstances, not as a general option and I'm still learning about that.

If you use a HUD like PokerTracker you can spot their VPIP/PFR combined with them limping you can play them like a fiddle.

Open limping is bad and should be avoided almost always (I agree with Kurn about 'always and never'), overlimping can be OK.

Say you're on BTN with 65ss, 3 loose-passives limped in ahead, more loose passives in the blinds, we're 100bb effective. Folding would be terrible with your sexy implied odds but iso-ing a bunch of loose-passives with 6 high is bad as well. Overlimping is clearly the best action here.

Onto the question at hand, this sounds like a sub-section of the question 'can we vary raise size with hand strength?'. The answer is 'yes, vs. bad players'. You want to be pretty sure they're bad though.

Change your hand to 65o, same situation. Do you still want to play it? The hand loses a little bit of equity, and we won't get as many semi-bluffing opportunities, but 1) would we really want to be semi-bluffing against these calling stations and 2) would we really want to hit a 6-high flush in a 6-way limped pot?

I think our implied odds are huge enough that overlimping 65ss is the best choice.

We get a made straight or flush something like 1 in 50? and we're getting 100:1 implied several times over. Really we should be able to play those kind of odds and come out ahead vs. passive fish.

I agree domination is a thing though and yeah I'd fold 65o, so maybe I'm contradicting myself here.

EDIT

Generally I agree we play SCs for the semi-bluff power vs. people with a fold button, so we need to take a more aggressive from the get-go.

I think it depends on the opponents, though I haven't analyzed this enough preflop.
I think the answer is yes. For instance with 33 in that situation, I think a raise to $7 is often higher EV than limping even if no one will fold. It often increases your implied odds, mainly because bigger bets will be made/called from the flop onward since the flop pot is bigger to begin with. But if it doesn't much increase your implied odds (ie if they're likely to pay you just as much when the flop pot starts at $10), it's better to limp. Ofc the likelihood of getting 3bet (slim in live 1/2) also affects your decision, and so should the likelihood of someone raising behind you if you limp.

But when that play is better than limping, there's still the question of whether it's better than a bigger (and partial field-thinning) raise. At a fish barrel of a table, how do we engineer our raise so as to extract every last penny of EV? I don't think that's a beginner question.

this is has to be most common misconception ever

yes but not with AK

If the opponents are "decent", then poker is just a game of stealing the blinds. You want to play as few players as possible so your share of the 1.5bb is maximised.
If players are terrible (most are), then poker is a game of winning money from idiots post-flop. You expect to win a lot more than a share of 1.5bb when you play an implied odds hand against idiots.
Open limping 65s in EP is bad, because there's only a low chance of seeing a cheap flop with lots of implied odds, and you'll be OOP, making it harder to win the max. Over-limping 65s otb vs 3 limpers is more attractive, as a multiway pot is looking very likely, and you'll be in prime position to benefit from your opponents' mistakes.

At certain tables, open limping, even from early position can be fine.Sometimes you have to adjust as well.