When you three-bet, you’re gearing up to play a big pot, so you ought to come strapped.

“What are the best hands to three-bet?” This is one of the most common questions I get from students of no-limit hold ’em. To the extent that the folks asking this question are looking for a simple answer like, “Always three-bet suited connectors,” they’re out of luck.

If I’m feeling cheeky, I tell them, “Pocket Aces!”, but of course what they are really asking is which hands are good for three-betting besides the obviously strong hands. Still, there’s some truth behind the glibness of my answer. The best hands for three-betting are generally the best hands: big pairs, broadway cards, suited connectors, and suited Aces.

It’s not often correct to three-bet all of these hands, but your three-betting range should generally consist of some subset of them. There aren’t a lot of situations where you should three-bet 72o, or K5o for that matter. If your opponents fold, your cards won’t matter, and if you believe your opponents will fold often enough, then by all means three-bet any old hand you feel like. But you probably wouldn’t be asking this question if you believed that.

When you three-bet, you’re gearing up to play a big pot, so you ought to come strapped. That doesn’t have to mean Aces, but it should mean a hand with a reasonable chance of winning the pot if you go to the flop. In most cases, the stronger your hand, the higher the Expected Value (EV) of your three-bet will be.

It does not follow from this that you should therefore three-bet a strictly linear range, starting from Aces and working your way down. The complication is that many hands that will have a high EV when they three-bet will also have a high EV when they call. Your job is to decide, when you have a reasonably good hand, whether it will play better as a three-bet or a call in this particular situation.

Here are some questions you can ask yourself to help make that determination:

Is your hand strong enough to call? Suppose three-betting were off the table, and you could only call or fold. Which would you prefer to do with this hand? Somewhat counter-intuitively, your lighter three-bets should generally come from the hands that end up in the “fold” bucket after you ask this question.

This is because, when considering a three-bet with a hand that would otherwise fold, you are comparing the EV of three-betting to zero, which is what folding will net you. Thus, any three-bet with a positive EV will be preferable and should be taken.

When considering a three-bet with a hand that would otherwise call, however, it’s not so simple; the threshold for a raise is higher. You must believe that three-betting will offer an even higher return than the already-positive-EV call.

How do you determine that? Asking the remaining questions may help.

What will the Stack-to-Pot Ratio (SPR) be if your three-bet is called? With a hand consisting of two broadway cards, you’ll generally flop either nothing or top pair with a strong kicker. In the latter case, you’d prefer to have an SPR of 5 or lower, so that you can comfortably stack off with your pair. If you think your hand won’t warrant playing for stacks in this scenario even if you flop top pair, then that’s a reason why your hand may not warrant a three-bet. One potentially surprising implication of this is that AKo is not an automatic three-bet in a deep-stacked game where the SPR in a three-bet pot will be well over 5.

Connectors and other suited hands, however, prefer high-SPR situations. If you make a flush or straight, you’ll generally be happy to put in eight or nine times the size of the pot. Even when you flop a draw, the deep stacks will play to your advantage by enabling you to semi-bluff and put pressure on the one-pair hands in your opponent’s range. That’s not to say that you should always “empty the clip” when you flop a flush draw, but the mere threat of such a large pot generates fold equity on early streets that you wouldn’t have in a low-SPR situation.

Of course your suited hands will not uncommonly flop pairs after three-betting, and with a high SPR, you’ll need to proceed cautiously in such cases. You should not generally aim to play for stacks in these situations, which means you’ll probably end up checking either the flop or turn.

One of the reasons why the biggest pairs play well as three-bets at a variety of stack depths is that they are flexible. You’ll frequently have an over-pair to the flop, in which case you can confidently stack off in low-SPR situations. In high-SPR situations, you must be more cautious when you hold one-pair, but you have the potential to make a set that can happily play stack off for many times the pot.

Will you dominate your opponent’s calling range? This is a question whose answer depends on both your and your opponent’s position, and to some extent on anything you may know about how your opponent responds to three-bets. People do tend to focus too much on the risk of being dominated when three-betting non-premium hands and not enough on the possibility of dominating other hands in an opponent’s range. For instance, although raising AQ into AK is unpleasant, situations where your opponent has incentive to call with AJ, AT, or KQ more than compensate you for this risk.

Dominating hands in an opponent’s calling range contributes a lot to your pre-flop equity, which makes putting more money into the pot immediately appealing, but it also makes your post-flop play easier and more lucrative. Of course holding AQ vs AJ on an A-high flop is appealing, but even on a 742 flop, your chances of getting to showdown with the best hand are better in this case than if your opponent held 98s, which might be more tempted to bluff you out.

Do you mind playing a multi-way pot? Note that this is not, or at least shouldn’t be, a matter of personal preference, as in “I don’t like taking multi-way pots with Aces, they always get drawn out on.” Some hands simply play better than others in multi-way pots in ways that are relatively determinable.

This is partly a function of SPR, but the bottom line is that you need to make stronger hands to win multi-way pots, and so hands with the potential to make sets, straights, and flushes are happier to go multi-way than offsuit broadways hands.

Position is also relevant here. If you are on the button facing a CO raise, there may be other reasons to three-bet AKo, but a multi-way pot is less of a concern because you are guaranteed to have position, and only two players remain to act behind you. Facing a raise with AKo in early position, though, you should be more inclined to three-bet in order to avoid playing a five- or six-way pot in bad position. If five players see the flop, even on an Ace-high or King-high board you may face reverse implied odds, as others will be reluctant to put money into the pot if they can’t beat top pair.

Hopefully that gives you a lot to think about the next time you’re facing a raise. The most important thing is to think, period. If you have in mind that such-and-such is “a three-betting hand”, regardless of the situation, while something else is “a calling hand”, you’re doing it wrong. Next month, I’ll look at how to put all of this together with a couple of examples.