Pokeriad / Learn / Small pairs
Updated November 2026
Small pairs, twos through sixes, make their money by turning into three of a kind. You flop that set 11.8% of the time, about 1 time in 8.5, so call a raise only when the money you can still win is at least 15 to 20 times the price of the call.
The plan is called set mining: pay a small price, see the flop, and give up on the 88 flops in 100 that miss. The fun part is the other 12, when a set of fives hides under a king-high flop and can win a whole stack.
| What happened | What you do |
|---|---|
| Folded to you, late seat | Raise to 3 big blinds |
| Someone raised, deep stacks | Call if the smaller stack is 15 times the call or more |
| Smaller stack under 15 times | Fold |
| Re-raised before you act | Fold |
| You flop a set | Bet or raise, and get the money in |
| You miss | Check, and fold to a bet |
Only 2 cards in the deck match your pair. The chance that neither lands in three flop cards is 48/50 × 47/49 × 46/48, which is 88.2%, so you hit 11.8% of the time.
That works out to about 7.5 misses for every hit. Keep that number, because the whole price question comes down to it.
You miss 7.5 times for each hit, so every set has to earn 7.5 times what the call costs just to break even. On a $6 call, that's $45.
Real hands pay you less than that. Your opponent won't always have enough to call you down, and a set sometimes loses. So players use 15 to 20 times the call, measured against the smaller of the two stacks.
Because almost every flop beats them. Hold fives and at least one card higher than a five lands on 98.1% of flops. A pair that small can't stand much betting after that.
Against one random hand fives win 60.3 times in 100, which sounds fine, and against eight they win 13.1. Every extra player makes it more likely someone paired something bigger.
Chance of winning if everyone stays to the river
Against two big cards you're a small favorite. Against a bigger pair you're a heavy underdog, winning about 19 times in 100, and that's why you don't call big raises with these hands.
So an all-in before the flop is only for short stacks. Deeper than that, you'd be risking your whole stack for a hand that's already behind anything with a bigger pair.
Get the money in. A set is hidden, so most players never put you on one, and it's ahead of nearly every hand that could call.
Bet two-thirds of the pot or more, and raise if they bet first. Bet hardest when draws are around: a flush draw still wins 26 times in 100 against you. Another player holding a bigger set is rare, 3 hands of 1,081 on king-five-two.
Most of the time, and quickly. The money you save on the misses is what makes the plan work.
A $1/$2 game, with the pot shown after each round of betting. The stacks are 33 times the $6 call, so the price is easy.
The hijack raises to $6 and you call. Both blinds fold, leaving $1 and $2 in the pot.
You've hit. The hijack bets $10 and you just call, because there's no draw to worry about and you want them to keep betting.
The hijack bets $25. You raise to $75 and they call.
The hijack checks. You push in your last $109 and they call with ace-king.
You take the $403 pot, and you put in $200 of it: $6, $10, $75 and $109.
Now count the cost of the misses. At $6 a try, about 7.5 misses cost $45, and one hand like this paid $203. A hand that big won't come every time, which is why the rule asks for 15 to 20 times the call.
Twelve spots in three rounds: the numbers, the price, and what to do after the flop.
When you're first in from a late seat, yes. Our trainer's 6-player charts raise fives and sixes from every seat, and twos through fours from every seat except the first.
Calling a small raise with a pocket pair, hoping to flop three of a kind and win a big pot, and folding when you miss.
Rarely. Against a random hand on king-five-two, a set of fives wins about 96 times in 100.
A habit many players use: don't call a raise with a small pair unless the smaller stack is at least 15 times the call. It's built from the 7.5 misses per hit, doubled, because a set rarely wins the whole stack.
They're worth less once stacks get short, because set mining needs deep stacks. With 20 big blinds or less, it's closer to an all-in or a fold.
Win rates against 1 and 8 opponents come from 60,000 random deals at each table size, with ties shared. Head-to-head numbers count every possible board and every suit pairing.
A set or better on the flop is 1 minus C(48,3) / C(50,3), and 19,600 flops include 2,304 hits. That's 17,296 misses to 2,304 hits, or 7.5 to 1. The tool's average is 11.76% times the amount won, minus 88.24% times the call. It leaves out the pot already there and the times a set loses.
Send them this page, or save the cheat sheet as an image.