PLO starting hand rankings

All 16,432 hand classes, ranked two ways: exact equity against one random hand (400,000 trials per class), and full-game EV decoded from the 6-max 100bb solve. The two rankings disagree in instructive places.

The short answer

By raw equity, the best PLO starting hand is A-A-T-T double-suited at 71.5% against a random hand, and the top five are all double-suited aces with a pair: A-A-T-T (71.5%), A-A-J-J (71.2%), A-A-Q-Q (70.9%), A-A-5-5 (70.8%), A-A-9-9 (70.7%). By full-game EV in the solve, A-A-J-J double-suited takes the top spot instead. A-A-K-K double-suited, the hand most players would name, is 6th by equity and 5th by EV. At the other end, the worst hand in the game is quad deuces at 9.2%. The full ladder, both rankings, is in the CSV below.

Data download (CC BY 4.0, attribution: solveplo.app)

The top 100 by equity

Equity is exact to about ±0.1 (400,000 trials per class), so treat adjacent ranks within 0.2 points as ties. The EV# column is the same hand's rank in the solve's full-game EV ordering.

#HandSuitsEquity vs 1EV #Combos
1A♠A♥T♠T♥ds71.5%46
2A♠A♥J♠J♥ds71.2%16
3A♠A♥Q♠Q♥ds70.9%26
4A♠A♥5♠5♥ds70.8%36
5A♠A♥9♠9♥ds70.7%186
6A♠A♥K♠K♥ds70.7%56
7A♠A♥J♠T♥ds70.6%912
8A♠A♥8♠8♥ds70.5%106
9A♠A♥7♠7♥ds70.4%66
10A♠A♥T♠9♥ds70.4%712
11A♠A♥6♠6♥ds70.4%196
12A♠A♥4♠4♥ds70.3%86
13A♠A♥T♠8♥ds70.1%1112
14A♠A♥9♠8♥ds70.0%2312
15A♠A♥8♠7♥ds69.9%2012
16A♠A♥Q♠T♥ds69.8%1612
17A♠A♥7♠6♥ds69.8%1212
18A♠A♥Q♠J♥ds69.8%1712
19A♠A♥T♠7♥ds69.8%2712
20A♠A♥7♠5♥ds69.6%3312
21A♠A♥3♠3♥ds69.6%136
22A♠A♥J♠9♥ds69.6%2212
23A♠A♥9♠7♥ds69.5%2512
24A♠A♥6♠5♥ds69.4%2112
25A♠A♥J♠8♥ds69.4%1512
26A♠A♥8♠6♥ds69.3%2412
27A♠A♥T♠6♥ds69.2%3812
28A♠A♥5♠4♥ds69.1%2912
29A♠A♥K♠T♥ds69.1%3912
30A♠A♥8♠5♥ds69.1%2812
31A♠A♥K♠J♥ds69.0%3712
32A♠A♥K♠Q♥ds69.0%3112
33A♠A♥T♠5♥ds69.0%3012
34K♠K♥T♠T♥ds69.0%3916
35A♠A♥J♠7♥ds68.9%4312
36A♠A♥Q♠9♥ds68.9%4612
37A♠A♥2♠2♥ds68.8%146
38A♠A♥7♠4♥ds68.8%4012
39A♠A♥Q♠8♥ds68.8%5912
40A♠A♥J♠5♥ds68.8%4912
41A♠A♥T♠T♦ss68.8%5824
42A♠A♥K♠5♥ds68.7%3612
43A♠A♥9♠6♥ds68.7%5512
44A♠A♥Q♠5♥ds68.7%3512
45A♠A♥6♠4♥ds68.7%4512
46A♠A♥9♠5♥ds68.6%3212
47A♠A♥T♠4♥ds68.6%4112
48K♠K♥J♠J♥ds68.6%3826
49A♠A♥J♠4♥ds68.5%3412
50A♠A♥8♠4♥ds68.5%5012
51A♠A♥K♠4♥ds68.5%6612
52A♠A♥Q♠4♥ds68.5%5312
53A♠A♥J♠J♦ss68.4%4424
54A♠A♥5♠3♥ds68.4%2612
55A♠A♥T♠3♥ds68.4%4212
56K♠K♥Q♠Q♥ds68.4%2896
57A♠A♥J♠3♥ds68.4%5212
58A♠A♥K♠3♥ds68.3%5112
59A♠A♥J♠6♥ds68.3%6812
60A♠A♥Q♠6♥ds68.2%7312
61A♠A♥Q♠3♥ds68.2%6112
62A♠A♥K♠6♥ds68.2%6412
63A♠A♥K♠8♥ds68.2%6712
64A♠A♥K♠7♥ds68.2%4712
65A♠A♥Q♠7♥ds68.2%5612
66A♠A♥4♠3♥ds68.2%4812
67A♠A♥Q♠Q♦ss68.1%5724
68A♠A♥T♠2♥ds68.1%6912
69A♠A♥K♠9♥ds68.1%7812
70A♠A♥5♠5♦ss68.1%8824
71A♠A♥J♠2♥ds68.1%6312
72A♠A♥Q♠2♥ds68.1%6512
73K♠K♥9♠9♥ds68.0%4286
74A♠K♠K♥J♥ds68.0%37812
75A♠A♥K♠2♥ds68.0%8112
76A♠A♥7♠3♥ds68.0%7512
77A♠A♥9♠9♦ss68.0%7224
78A♠K♠K♥T♥ds68.0%38412
79A♠A♥J♦T♠ss68.0%5424
80A♠A♥J♠T♦ss68.0%7124
81A♠A♥6♠3♥ds67.9%6212
82A♠A♥K♠K♦ss67.9%6024
83A♠A♥T♦9♠ss67.8%8024
84A♠K♠K♥Q♥ds67.8%35412
85A♠A♥T♠9♦ss67.8%7624
86A♠A♥5♠2♥ds67.7%8512
87A♠A♥9♠4♥ds67.7%7412
88A♠A♥8♠8♦ss67.7%8724
89K♠K♥T♠9♥ds67.7%47612
90A♠K♠K♥5♥ds67.7%41612
91A♠A♥7♠7♦ss67.6%8924
92A♠A♥T♦8♠ss67.6%8424
93A♠A♥T♠8♦ss67.5%7924
94K♠K♥J♠T♥ds67.5%41412
95A♠A♥6♠6♦ss67.5%7724
96A♠A♥9♠3♥ds67.5%9712
97A♠K♠K♥4♥ds67.5%40712
98A♠A♥4♠4♦ss67.5%7024
99A♠A♥4♠2♥ds67.4%9312
100A♠A♥9♦8♠ss67.4%10524

89 of the top 100 hold at least two aces; the best non-AA hand is K-K-T-T double-suited at #34. Suit tags: ds = double-suited (2-2), ss = single-suited (2-1-1). Combos = concrete deals in the class (the suit arrangement shown is the class representative).

Where equity and the solve disagree

Equity vs one random hand is a clean physical measurement; full-game EV is what the hand actually earns across every seat and line in the 6-max 100bb solve. The gaps between the two rankings are the interesting part.

Kings fall

HandEquity #EV #
K-K-9-6 ds200618
K-K-J-7 ds199613
K-K-9-7 ds152557
K-K-Q-9 ds185586
K-K-T-8 ds133523

Bare aces rise

HandEquity #EV #
A-A-Q-9 3-suit353196
A-A-K-2 ss336183
A-A-Q-3 ss290144
A-A-7-3 ss335195
A-A-4-3 ss298171

The pattern is domination. K-K hands poll well against a random hand but run into A-A in exactly the pots that get big, so the solve discounts them by hundreds of ranks. Aces with weak side cards go the other way: the pair that can never be dominated preflop, plus ace blockers, earns more money than its raw equity suggests. Equity measures the deal; EV measures the game.

What suits are worth

Same ranks, different suit patterns. Where several arrangements share a tag, the best one is shown.

Ranksdsss3-suitmonorainbow
A-A-K-K70.7% (#6)67.9% (#82)--65.1% (#334)
A-A-J-T70.6% (#7)68.0% (#79)67.0% (#124)-65.4% (#268)
T-9-8-753.2% (#4,719)50.2% (#7,675)49.4% (#8,647)48.5% (#9,486)47.0% (#11,163)

Double-suit is worth 5-6 equity points over rainbow on premium ranks, and it moves A-A-K-K from rank 334 to rank 6. Note the monotone column: four of one suit ranks BELOW single-suited (you burn your own flush outs), and a rainbow T-9-8-7 rundown is a below-average hand at 47.0%.

What the top X% looks like

The hand at each percentile boundary, combo-weighted, with its exact equity.

TopBoundary handSuitsEquity vs 1
1%A♠A♥K♦J♠ss66.2%
2%A♠K♠K♥5♦ss65.1%
5%A♠A♥5♦2♣rainbow61.9%
10%J♠J♥9♦5♠ss58.4%
15%J♠J♥T♦6♣rainbow56.2%
20%K♠J♥T♠7♦ss54.7%
30%Q♠T♥9♥2♠ds52.8%
40%K♠T♥9♦5♥ss51.1%
50%A♠Q♥8♦2♦ss49.6%
75%K♠8♥6♦5♣rainbow45.7%
90%Q♠7♥5♦2♥ss42.4%

The spread is tight by design: the top 1% boundary sits at 66.2% and the median at 49.6%. PLO preflop equities cluster, which is why position and playability decide so much of the money.

16,432, and the bottom of the ladder

PLO4 deals 270,725 distinct four-card combinations. Merge the hands that differ only by suit relabeling and 16,432 classes remain - not 16,384, which circulates online as a misquoted power of two. The bottom of the ladder is a staircase of four-of-a-kind holdings: 7-2-2-2 (20.9%), 5-5-5-5 (19.0%), 4-4-4-4 (15.5%), 3-3-3-3 (12.3%), 2-2-2-2 (9.2%). Quad deuces can never make a set, never use more than two of its cards, and blocks its own outs - the purest bad hand in poker.

Rankings are not a preflop strategy

Two rankings on this page, and neither is a chart. Equity-vs-random tells you what the deal gave you; the EV column tells you what the solve earns with it across all six seats; what you should do with a specific hand in a specific seat is a third question, and it is the one the trainer drills. In the solve, only the bottom 4.1% of classes are pure folds from every seat - almost everything is playable somewhere, at some frequency, and almost nothing is playable everywhere. Position is the variable these tables hold still. Use them to calibrate hand strength, check exact matchups in the equity calculator, and let the charts carry the seat-by-seat decisions.

Methodology

Equity: each of the 16,432 suit-isomorphism classes was simulated against one random opponent hand, dealt to the river, 400,000 trials per class (standard error about 0.08 percentage points, so adjacent ranks within ~0.2 points are statistical ties). Table generated 2026-07-12.

EV: reach-weighted full-game preflop EV per class, in big blinds, decoded from the 6-max 100bb solve (2026-07-13). Each hand's figure averages every seat and preflop line the solve actually plays it in, weighted by how often - the same "Opens" ordering the trainer uses, sorted here by raw EV. It inherits the solve's tree and assumptions, so treat EV ranks as one solve's opinion, stated exactly.

Both source tables ship in the site's static data and agree on equity to the digit. Regenerate everything with node scripts/study-guides/gen-plo4-rankings.mjs.

Cite this page SolvePLO, "PLO Starting Hand Rankings", solveplo.app/plo-starting-hand-rankings, updated July 2026.

Questions

What is the best starting hand in PLO?
By raw equity against one random hand: A-A-T-T double-suited at 71.5% (400,000-trial simulation). By full-game EV in the 6-max 100bb solve: A-A-J-J double-suited. The famous A-A-K-K double-suited is 6th by equity at 70.7% and 5th by solver EV - a monster either way, just not the top.
How many starting hands are there in PLO?
270,725 four-card combinations, which collapse to 16,432 distinct classes once suits that play identically are merged. The "16,384" figure sometimes quoted online is a misreading (2 to the 14th power); the correct class count is 16,432.
How much is double-suited worth in PLO?
For premium ranks, about 5-6 points of raw equity over rainbow: A-A-K-K goes 70.7% double-suited, 67.9% single-suited, 65.1% rainbow - falling from rank 6 to rank 334. For a middle rundown like T-9-8-7 the double-suit is the difference between an above-average hand (53.2%) and a below-average one (47.0% rainbow).
What is the worst starting hand in PLO?
Quad deuces: 2-2-2-2 wins just 9.2% against one random hand. You hold every deuce, so no set is possible and every board misses you. The five worst hands are all four-of-a-kind holdings.
What percentage of PLO hands are playable?
In the 6-max 100bb solve, all but the bottom 4.1% of hand classes take at least one non-fold preflop action from some seat at some frequency (mixed strategies included). Playable is not the same as profitable everywhere: position decides most of it, which is why the rankings here are calibration, not a chart.