What is a pocket pair?

Two hole cards of the same rank, like two sevens or two aces. It is the only starting hand that already holds a made hand before the flop. You are dealt one about once in 17 hands, and any particular pair, aces included, once in 221.

Players split them into three groups, and the groups play differently:

  • Big pairs, aces down to queens or jacks: hands you raise and re-raise for value.
  • Middle pairs, roughly tens down to sevens: strong before the flop, awkward after it, because overcards arrive on most flops.
  • Small pairs, sixes down to deuces: hands that rarely win unimproved and earn their keep by flopping a set.

How often does a pocket pair flop a set?

11.8% of the time, about 1 in 8.5. That number is the same for every pair, deuces or aces. By the river the chance of making a set or better climbs to 19.2%.

The other half of the picture is how often the flop brings a card above your pair, which is what makes middle and small pairs hard to play:

Your pairFlop has at least one overcard
K♠K♥22.6%
Q♠Q♥41.4%
J♠J♥57.0%
10♠10♥69.5%
8♠8♥86.7%
6♠6♥95.8%

These are exact counts over all 19,600 possible flops. With sixes or lower, an unimproved pair almost never has the flop to itself.

Are pocket pairs good hands?

Before the flop, mostly yes, and better than they look against high cards. Measured with PlusEV’s equity calculator engine, 40,000 simulated runouts for every combination of the pair:

MatchupPair wins (equity)
2♠2♥ vs a random hand50.1%
7♠7♥ vs a random hand66.2%
J♠J♥ vs a random hand77.5%
2♠2♥ vs A♣K♦52.6%
J♠J♥ vs A♣K♦57.1%
7♠7♥ vs A♣2♦70.3%
Any pair vs a bigger pair: 22 vs AA, 77 vs 88, JJ vs QQabout 18%

That is the famous coin flip: any pair against two higher cards is a small favourite. The catch is the last row. A pair that runs into a bigger pair wins less than one time in five, whichever pair it is.

Which pocket pairs should you open?

All of them from late position, and only the bigger ones early. In the 6-max cash ranges PlusEV drills, at 100 big blinds, every pair from sixes up is a raise from under the gun through the button. The small ones depend on where you sit:

PairUnder the gunHijackCutoff
5♠5♥50%100%100%
4♠4♥33%60%100%
3♠3♥ 2♠2♥20-28%37-39%74-100%

Under the gun, five players still act behind you, and a small pair that gets re-raised has to fold. One seat later it is half the problem, and from the cutoff on it is no problem at all. The small blind, which plays every later street out of position, mixes raising and limping with every pair. The full opening ranges are in our Texas Hold’em charts guide.

What do you do with a pocket pair when someone raises?

Re-raise the big pairs. With the small ones, call from the big blind and fold from most other seats. Averaged over the 15 facing-a-raise spots in PlusEV’s ranges, aces and kings re-raise every time, queens 96.5%, jacks 81%, tens 63% and nines about half the time. Below that, the seat decides.

From the big blind, call. The big blind calls a single raise with every pair from deuces to fives in all five of its spots, and folds no pair at all. It has a blind in the pot already, closes the action and gets the best price at the table, which is the whole case for defending the big blind wide:

Big blind facing an under-the-gun open, 6-max cash, 100bb. Every pair from tens down is a call.
AAAKsAQsAJsATsA9sA8sA7sA6sA5sA4sA3sA2s
AKoKKKQsKJsKTsK9sK8sK7sK6sK5sK4sK3sK2s
AQoKQoQQQJsQTsQ9sQ8sQ7sQ6sQ5sQ4sQ3sQ2s
AJoKJoQJoJJJTsJ9sJ8sJ7sJ6sJ5sJ4sJ3sJ2s
AToKToQToJToTTT9sT8sT7sT6sT5sT4sT3sT2s
A9oK9oQ9oJ9oT9o9998s97s96s95s94s93s92s
A8oK8oQ8oJ8oT8o98o8887s86s85s84s83s82s
A7oK7oQ7oJ7oT7o97o87o7776s75s74s73s72s
A6oK6oQ6oJ6oT6o96o86o76o6665s64s63s62s
A5oK5oQ5oJ5oT5o95o85o75o65o5554s53s52s
A4oK4oQ4oJ4oT4o94o84o74o64o54o4443s42s
A3oK3oQ3oJ3oT3o93o83o73o63o53o43o3332s
A2oK2oQ2oJ2oT2o92o82o72o62o52o42o32o22
FoldCallRaise 12.5

From any other seat, small pairs mostly fold. Sitting in the hijack, cutoff or button when someone has raised, deuces fold 89% of the time and fives 72%. The hijack facing an under-the-gun open folds 22 through 55 every time and sixes 81%:

Hijack facing an under-the-gun open, 6-max cash, 100bb. Small pairs fold; big pairs re-raise.
AAAKsAQsAJsATsA9sA8sA7sA6sA5sA4sA3sA2s
AKoKKKQsKJsKTsK9sK8sK7sK6sK5sK4sK3sK2s
AQoKQoQQQJsQTsQ9sQ8sQ7sQ6sQ5sQ4sQ3sQ2s
AJoKJoQJoJJJTsJ9sJ8sJ7sJ6sJ5sJ4sJ3sJ2s
AToKToQToJToTTT9sT8sT7sT6sT5sT4sT3sT2s
A9oK9oQ9oJ9oT9o9998s97s96s95s94s93s92s
A8oK8oQ8oJ8oT8o98o8887s86s85s84s83s82s
A7oK7oQ7oJ7oT7o97o87o7776s75s74s73s72s
A6oK6oQ6oJ6oT6o96o86o76o6665s64s63s62s
A5oK5oQ5oJ5oT5o95o85o75o65o5554s53s52s
A4oK4oQ4oJ4oT4o94o84o74o64o54o4443s42s
A3oK3oQ3oJ3oT3o93o83o73o63o53o43o3332s
A2oK2oQ2oJ2oT2o92o82o72o62o52o42o32o22
FoldCallRaise 7.5

The difference is who is left behind. The big blind closes the action. A cold call from the hijack still has four players to get through, any of whom can re-raise and take the set-mining price away. The re-raise itself is covered in our 3-bet guide.

When is set mining profitable?

When the stacks are deep enough that the occasional set pays for all the misses. That is standard pot-odds arithmetic, and it is what the calculator below does:

Set mining calculator all amounts in big blinds
14.8bb a set must win after the flop, on average
40x stack to call, passes the 15x rule
Call. A set has to win about 15% of what is left behind, which is realistic.

Rule of thumb behind the verdict: a set usually wins up to about a third of the stack behind, sometimes two-thirds, rarely more.

At 7.5 to 1 against, a call of 2.5 big blinds has to bring back about 18.8 big blinds every time the set comes in. The pot already holds some of that, and the rest has to come from your opponent after the flop. The calculator’s defaults are calling a 2.5bb open with 4bb in the middle and 100bb stacks: a set then needs to win about 15bb more on average, around 15% of what is left, which is a realistic ask.

Shrink the stacks and the same call turns bad. At 30bb deep the set has to win more than half the remaining stack, and most opponents will not pay that off. Most guides quote a multiple instead: the effective stack should be 15 to 20 times the call, and the 15/25/35 rule sets 15 times as the floor for a small pair. It is the arithmetic above with a margin for the sets that do not get paid, and the sets that lose: a pair that hits still runs into a bigger set or a flush now and then. Like any rule of thumb, it is a starting point. The pot odds guide covers the same maths for draws.

What do you do with a pocket pair facing a 3-bet?

Big pairs raise again, everything else mostly calls. After opening and facing a 3-bet, aces 4-bet 94% of the time and kings 91% in PlusEV’s ranges, while queens split 56% 4-bet and jacks 42%. Every pair from nines down calls between 58% and 87% of the time and almost never re-raises.

That runs against common advice, which says to fold small pairs to a 3-bet unless very deep. At 100bb the price still works: calling a 3-bet to 7.5bb after opening to 2.5bb costs 5bb more, with about 92bb behind, over 18 times the call. Out of position or against a bigger 3-bet, the multiple drops and the fold comes back. When to raise again instead is the subject of the 4-bet guide.

Which pocket pairs do players get wrong?

Jacks, more than any other pair. Across PlusEV’s 1M+ recorded training decisions, counting only spots where the solver has one clear answer and weighting every difficulty level equally, players pick the right action with pocket jacks 74.1% of the time (3,890 answers from 797 players). That makes jacks the sixth most misplayed of all 169 starting hands. Eight of the 13 pairs are misplayed more often than the median hand, which players get right 86.5% of the time.

PairRight answer chosenRank of 169 hands (1st = most misplayed)
J♠J♥74.1%6th
2♠2♥77.0%joint 18th
3♠3♥77.0%joint 18th
9♠9♥79.8%34th
10♠10♥81.9%45th
4♠4♥82.3%46th
5♠5♥84.5%65th
8♠8♥84.9%joint 68th
7♠7♥88.1%joint 97th
A♠A♥89.3%111th
6♠6♥89.6%114th
Q♠Q♥91.3%121st
K♠K♥93.8%joint 131st

Each pair has between 2,693 and 13,932 answers behind it.

The ranges suggest why the order is not simply big to small. Jacks re-raise an open 81% of the time and 4-bet 42%, nines and tens split between re-raising and calling, and deuces and threes are a call in one seat and a fold in the next: the right answer keeps changing. Kings and queens almost always do the same thing. Knowing that is one thing; picking it in two seconds is another. That second skill comes from repetition, which is what the PlusEV preflop trainer drills.

What do you do with a small pair when the flop misses?

Usually, give up cheaply. This part is standard theory rather than our data. A small pair on a board of three higher cards is beaten by any pair your opponent holds, and with deuces through fives there is an overcard on 98% or more of flops.

  • Hit a set: play for stacks. That is the hand you called for, and it is usually best.
  • Miss, out of position: check, and fold to a bet on most boards.
  • Miss on a low, dry board, like 5-5 on 8-4-2 of three suits: your pair is often still best. Check or bet small, and do not give it up to one small bet.
  • Multiway: an unimproved small pair is almost never good with three or more players in, so it is set or fold.
  • Miss, in position, heads up: a small continuation bet can take the pot on dry boards, but it is a bluff with a weak hand, not a value bet. Firing again and again with an unimproved small pair is the leak the set-mining plan exists to avoid; the pair was never the point.

Does this change in live games and tournaments?

Yes, mostly through stack depth. The figures above are 6-max cash at 100 big blinds, and the adjustments are standard theory:

  • Live cash. Stacks are often deeper and opponents pay off sets more, so calling with small pairs gets better. Raises are also bigger, so check the price before you call.
  • Full ring. More players behind every early seat, so fold more small pairs up front.
  • Opponents. Set mining pays best against tight players who stack off with overpairs, and worst against players who fold to any aggression, because their sets never get paid.
  • Rake. At low stakes the rake eats into thin set-mining calls, so aim for the deeper end of 15 to 20 times.
  • Tournaments. Stacks shrink fast. Below about 30 big blinds, the 15x rule already rules out calling a normal raise to set mine, and middle pairs turn into raise-or-shove hands instead of calls.

The short version: raise big pairs, call small ones from the big blind, fold them from most other seats when someone has raised, and only set mine when the stacks are deep enough to pay for the 11.8%. For where every pair sits among all 169 hands, see the best starting hands.

Quick answers

What are the odds of flopping a set with a pocket pair?

11.8%, or about 1 in 8.5. It is the same for every pair. By the river the chance of making a set or better rises to 19.2%.

What is set mining in poker?

Calling a raise with a pocket pair, usually a small one, mainly to flop three of a kind and win a big pot. You give up the hand on most flops, so it only pays when the stacks are deep enough for the big hand to win a lot.

What is the 15/25/35 rule in poker?

A rule of thumb for calling a raise with a speculative hand: the effective stack should be at least 15 times the call for a small pocket pair, 25 times for a suited connector and 35 times for an offsuit connector or suited gapper. It is a revision of the older 10/20/30 rule.

How often are you dealt a pocket pair?

About one hand in 17, or 5.9% of the time. Any one specific pair, such as pocket aces, comes once in 221 hands.

Should you fold small pocket pairs before the flop?

Often, yes. In the 6-max cash ranges PlusEV drills, the hijack facing an under-the-gun raise folds 22 through 55 every time. The big blind is the exception: it calls a single raise with every pocket pair.

Which pocket pair is hardest to play?

Pocket jacks. Across PlusEV's 1M+ recorded training decisions, players choose the right action with jacks 74.1% of the time, the sixth worst of all 169 hands. Deuces and threes come next among the pairs, at 77.0%.

Is a pocket pair a coin flip against two overcards?

Close to one, with the pair slightly ahead. Pocket deuces win 52.6% against ace-king offsuit and pocket jacks 57.1%, measured with the site's equity engine.