Set mining is calling a raise with a small pocket pair (2-2 through 6-6, sometimes up to 8-8) with one plan: flop a set and get paid by someone holding a big hand who never sees it coming. It is one of the most profitable plays in poker when the conditions line up, and one of the quietest leaks when they don't. The difference is in what you check before the chips go in.
First, the number that runs the hand
With a pair in your hand, two cards of your rank are left in the deck. The chance that at least one lands on the flop is 11.8%, roughly one time in eight and a half. The other seven and a half times you hold a low pair against a range that is almost always ahead, and the right play is usually to fold to the first bet.
That defines the hand: you pay seven and a half times to win once. If the raise is 3 big blinds, those seven losses cost 22.5 blinds. The time you hit, you need to win back at least that much just to break even. In other words, 7.5 times what you paid just to end up at zero.
The 15x rule (and why it isn't 7.5x)
If break-even sits at 7.5x, why does the standard ask for an effective stack of at least 15 times the raise? Because hitting doesn't guarantee getting paid. Sometimes the raiser has air and gives up on the flop. Sometimes you get one bet. And sometimes you hit and lose: to a bigger set, to a straight or a flush on the river. The half of the money that never arrives is the reason for the margin.
Three things to apply it well:
- Effective stack is the smaller of the two stacks. If you have 100 blinds and the raiser has 30, the number that matters is 30. You can't win what they don't have.
- Against players who pay off, 15x is enough. Against players who fold fast when the board turns ugly, ask for 20x or more. If you can't yet tell the types of poker players at your table apart, that read is worth more than the rule.
- Strong ranges pay better. An early-position raiser holds big pairs and broadways: when you hit, they have something to pay you with. A button opening anything folds far more often.
All of this is the cleanest application of implied odds: money that isn't in the pot yet but that you can win if you complete your hand.
When to do it and when not to
At 100 blinds in a cash game, facing a 3-blind open with even stacks, calling with 4-4 in position is routine: 33 times the raise. If instead you face a 3-bet (a re-raise) to 10 blinds and need to put in 7 more with 100 behind, the ratio drops to 14x, right on the edge. It's worth more against 3-bet ranges that pay to the end than against players who only re-raise the top and then slow down. The logic of the 3-bet tells you what to expect.
In tournaments the math changes fast. With 25 blinds facing a 2.2 open, calling to hit a set gives you 11x: not enough. At 15 or fewer, small pairs stop being set-mining hands and become push or fold hands. And never call out of position with a short stack "to see what comes": you're buying an expensive lottery ticket.
You hit: now get paid
A hidden set is strong because nobody sees it. How to get paid depends on the board:
- Dry board (K-7-2 with no draws): you can call the flop bet and let the raiser keep firing; their top pair pays you two streets. It's the exception where slow play makes sense.
- Board with draws (9-8-5 with two of a suit): raise now. Every card that completes a straight or a flush takes value away, and flat-calling gives free cards.
- Big pot, short stack: the SPR (stack divided by pot) is low, so the set goes all-in without a second thought.
And when you lose set over set, or to a straight that arrived on the river, it wasn't a mistake: it was a cooler. Log it, review it, move on.
You missed: the other half of the plan
Seven times out of eight the flop doesn't bring your card. The discipline that makes set mining profitable is folding without pain. Two nuances:
- If the board is lower than your pair (6-6 on 5-3-2), you don't have a set, but you probably have the best hand. Play it as a pair, carefully, not as failed set mining.
- If the raiser checks instead of betting, a small bet with your low pair can take the pot from two high cards that missed. Never the plan, but value that shows up.
Mistakes we clean up every week in reviews
- Calling with a small pair against a 20-blind stack "because I have 100." The effective stack was 20.
- Calling a 3-bet out of position with exactly 15x against someone who folds to the first bad flop.
- Hitting on a wet board and flat-calling: the flush arrives and the set becomes the second-best hand.
- Continuing with 3-3 on A-K-9 after a bet "because the card might come on the turn." It might: two cards out of 47, about 4% of the time.
If starting hands are still a question mark for you, start there: small pairs are only one piece of the chart.
The lesson
Set mining is a math decision dressed up as hope. Before you call, check three things: an effective stack of at least 15 times the raise, an opponent with something to pay you with, and position to collect. If all three say yes, you call and wait for your 12%. If any fails, you fold the pair: the next hand is a minute away.
Got a hand where you called with a small pair and aren't sure it was right? Bring it to the community with stacks and positions and we'll run the numbers together. Join free.