Pocket pairs: the complete decay curve from A-A to 2-2
Preflop Edge guide · figures from the same 169-hand simulation data that powers every hand page (all-in win-rates vs random opponents)
In short: pocket pairs are the only hand class in Hold'em that arrives already made — you never flop "air" with a set, and you're never pure bluffing with a pair. But pairs are also the class whose chart value evaporates fastest as the crowd grows: between the vs-1 and vs-5 columns every pair drops by roughly 41.9 points of win-rate — A-A by 36.6, and 9-9 by 46.1, the steepest collapse of the thirteen. This guide shows where each pair actually stands, and why pairs are played for their drawing potential, not their current rank.
The pair ladder, measured
| Rank vs 1 / vs 5 | Pair | Win-rate vs 1 | Win-rate vs 5 | Win-rate vs 8 |
|---|
| Rank vs 1 | Rank vs 5 | Hand | WR vs 1 | WR vs 5 | WR vs 8 |
|---|---|---|---|---|---|
| 1 | 1 | AA | 85.2% | 48.6% | 34.8% |
| 2 | 2 | KK | 82.4% | 42.3% | 29.3% |
| 3 | 3 | 79.9% | 35.9% | 24.7% | |
| 4 | 4 | JJ | 77.5% | 33.9% | 21.0% |
| 5 | 6 | TT | 75.0% | 29.8% | 18.2% |
| 6 | 12 | 99 | 72.1% | 26.0% | 16.6% |
| 7 | 20 | 88 | 69.2% | 23.3% | 15.2% |
| 10 | 33 | 77 | 66.2% | 20.8% | 14.2% |
| 17 | 42 | 66 | 63.3% | 19.6% | 14.0% |
| 28 | 52 | 55 | 60.3% | 18.2% | 12.8% |
| 48 | 71 | 44 | 57.0% | 16.5% | 12.4% |
| 66 | 76 | 33 | 53.7% | 16.3% | 12.1% |
| 87 | 80 | 22 | 50.3% | 15.8% | 11.9% |
Three things to read out of that table before any strategy talk. One: on the vs-1 chart even deuces (87) outrank every two-card hand except the other pairs, and J-J sits at 4 — yet the vs-5 column reshuffles the ladder: deuces slip to 80, and hands like A-9 offsuit that *call* with a raise ahead of deuces heads-up overtake them in the crowd. "Already ahead" is not the same as "ahead for a whole hand." Two: that decay is why crowd size matters more than rank: the gap between a pair's vs-1 and vs-5 numbers is the purest measure of how much a hand leans on being "already ahead." Three: the ranking columns themselves shift — small pairs slide less than overpairs as the crowd grows, which is exactly what the two rank columns in each row record.
The math under the decay
Two mental models, both imperfect, both worth owning. The first counts set value: an under-overcard pair's favorite job is to improve. A middle pair like 9-9 (72.1% vs one, 26.0% vs five) has real flop equity in two ways: it can improve to a set or trips, and every straight or flush that beats a top-pair hand tends to run through the lower pairs — so a crowd size that shrinks your relative win-rate also raises how often you'll be "your set or nothing."
The second model counts overcard damage: A-A vs one random hand has few ways to lose, while A-A vs eight random hands faces eight independent chances that somebody's two random cards include a deuce, a seven, a flush draw that lands. The higher the pair, the shorter the list of hands that can honestly beat it heads-up — and the longer that list gets with every seat you add. That single fact explains most live-action pair mistakes: the same J-J that's 77.5% against any single random hand is 33.9% against five, and five strangers who choose to see a flop are not the same population as five cards dealt face-down.
Read the two crowd-size guides — this page and the rankings guide — together with the individual pair pages (J-J, T-T, 5-5) if you want the full per-hand tables.
Practical pair rules that survive the decay
- Big pairs (9-9 and up) are value hands, not traps. They're genuine favorites heads-up (T-T 75.0%, J-J 77.5%, Q-Q 79.9%, K-K 82.4%, A-A 85.2% vs a random hand), and they turn into coin-flip races the moment several opponents keep money in the pot. When the vs-5 column falls off the cliff, that's your tempo cue to stop building pots.
- Small pairs (2-2–6-6) earn their rent via the flop, not the chart. 5-5 wins 60.3% vs one random hand but 18.2% vs five: most of that gap is two overcards, which is precisely what a real raising range is full of. Their job is cheap 3-bet pots and short-stack shoves where a flopped set is the nuts, not "I called a raise with deuces to play a four-bet pot."
- Middle pairs (7-7–J-J) are the decision zone. Strong on the chart at every crowd size, thin in real games, because your "good" hand vs random opponents is a set-or-die hand vs opponents who flat-call only with overcards or bigger pairs. The vs-5 column on each pair's page is the honest way to see it.
- In PLO, pairs are half a hand. A middle pair in Omaha is two of your four cards and a liability unless it's connected to the rest — see the PLO guide before carrying Hold'em pair instincts across games.
Where to drill this
Every pair has its own page with the full 1–8 table — A-A, K-K, Q-Q, J-J, T-T through 2-2. Pair-vs-nonpair dilemmas are the subject of the 169-hand rankings guide, the suited tax is measured in the suited-vs-offsuit guide — and the Trainer, included in every Study Pack from $19, is where you rehearse these spots until the decay curve is instinct.