22 vs AK: Who Wins? Exact Odds
22 is a tiny favorite against AKo — 52.6% vs 47.4% — and effectively dead even with AKs at 50.1% vs 49.9%. The smallest pair still technically wins the race, by the thinnest margin in poker.
Win, tie, loss
Share of all possible runouts across every suit combination (96 combo pairs). Equity adds half the ties to each side: 22 52.0%, AK 48.0%.
Equity by suit configuration
22 vs AKo52.6% / 47.4%
| Suit configuration | Combos | 22 equity | AKo equity |
|---|---|---|---|
2♠2♥ vs A♦K♠ K♠ shares the deuce's suit | 24 | 52.65% | 47.35% |
2♠2♥ vs A♠K♦ A♠ shares the deuce's suit | 24 | 52.65% | 47.35% |
2♠2♥ vs A♦K♣ no shared suits — every backdoor flush is live | 12 | 53.04% | 46.96% |
2♠2♥ vs A♠K♥ A♠ shares the deuce's suit; K♥ shares the deuce's suit | 12 | 52.26% | 47.74% |
| All 22 vs AKo combos | 72 | 52.65% | 47.35% |
22 vs AKs50.1% / 49.9%
| Suit configuration | Combos | 22 equity | AKs equity |
|---|---|---|---|
2♠2♥ vs A♠K♠ 2♠ blocks one card of AK's flush suit | 12 | 50.30% | 49.70% |
2♠2♥ vs A♦K♦ no AK flush-suit card in the other hand — the flush draw is fully live | 12 | 49.92% | 50.08% |
| All 22 vs AKs combos | 24 | 50.11% | 49.89% |
Each row is a suit-isomorphism class: relabeling the four suits doesn't change equity, so every combo in a row has the identical, exactly-enumerated number. “Combos” counts the hand-vs-hand combinations the row represents.
Why the smallest pair barely hangs on
22 starts "made" but almost never improves, and it can even lose without AK pairing: when the board makes two bigger pairs, the deuces are counterfeited and AK's ace kicker plays. Add AK's straight and flush chances — all of its broadway cards are live against deuces — and the pair's head start nearly evaporates. It survives at 52.6% vs AKo only because "nearly" isn't "completely": on the 49.7% of boards with no ace and no king, an unimproved pair of twos is usually still the best hand.
- 22 vs AKs is the closest thing to a true coin flip in preflop hold'em: 50.1% vs 49.9%, closer to 50/50 than any famous matchup on this site.
- An ace or king reaches the board 50.3% of the time, and 22 improves to a set just 19.9% of the time — the deuces are literally betting that the board misses two overcards for five straight cards.
- Rank matters enormously on the pair side: QQ beats AKo 56.8% of the time, while 22 manages only 52.6%. Every step down in pair rank hands equity back to the overcards.
How it was calculated
Exact enumeration, not simulation: for every suit configuration above, all 1,712,304 possible five-card boards were dealt and both hands scored with app/lib/poker-eval.js — the same 7-card evaluator behind /tools/odds-calculator. Ties split the pot, so equity is win% plus half the tie%. Computed 2026-07-27.
Don't take our word for it — run 22 vs AK yourself in the odds calculator and watch the Monte Carlo estimate converge on these numbers.
Frequently asked questions
Who wins 22 vs AK all-in preflop?+
22 by a hair against AKo (52.6% vs 47.4%); against AKs it's a genuine 50/50 (50.1% vs 49.9%). Exact enumeration, every suit configuration.
Is 22 vs AK a coin flip?+
Yes — the purest one in poker. The pair's edge over AKo is just 5.3 points, and suited ace-king erases it almost entirely.
Why does a tiny pair beat AK at all?+
Because AK is unpaired: on the 49.7% of boards where no ace or king appears, 22 usually wins unimproved. AK's compensation is 6 pairing outs plus live straights and flushes — which is why the margin is so thin.