How the PLO rankings work

This is the honest baseline: every hand’s equity against a single random hand, run to the river over millions of Monte-Carlo trials. “Top 20% by equity” means exactly that — the 20% of all starting hands that win most often at showdown against a random holding. It is cross-checked against a reference corpus and is always available.

Raw equity hides something every PLO player knows: two hands can win equally often at showdown yet play completely differently after the flop. A♠K♠Q♥J♥ and A♣A♦7♥2♠ have similar equity, but one flops nut draws and clean straights while the other is a high pair that’s hard to realize. Balanced starts from equity and then re-orders hands of similar equity by how well they realize that equity after the flop.

It is an equity-first blend (half equity, half realization score) — never a standalone “playability” sort. We tested a pure playability axis and it was simply a worse version of the equity axis, so we don’t ship it.

The realization score was fitted against a deliberately simple postflop simulation, run on every hand class:

• Heads-up, 100bb deep, single-raised pot.
• Bets are 2/3 pot; one decision per street, flop and turn, then showdown.
• You continue when your equity beats the pot odds, and fold when it doesn’t — except in position, where you take the free card.
• Villain holds a frozen strong range (the top 15% of hands) and keeps betting when connected.
• Each hand is played both in and out of position and the two are averaged.

The score blends nine signals: rundown connectedness, nut-wrap potential, nut-flush potential (A-high suited is the nut), high-card/blocker strength, set value (with a discount for naked high pairs), a reverse-implied-odds danger penalty for low non-nut hands, and three board-sampled stats — how often the hand flops a nut/strong draw, a strong made hand, or a dry whiff. The weights were fitted to the simulation, not hand-picked.

When you enter a complete hand, the panel shows exactly how it connects with the flop. These numbers are exact, not sampled — Kairo enumerates every one of the ~17,000 possible flops for your hand and counts the outcomes. “Set 12%” means you flop trips or better on precisely 12% of all flops.

Use it to compare hands that look similar: a low rundown and a broadway rundown can flop straights equally often, but the panel will show the low one whiffing into second-best hands more — exactly the reverse-implied-odds risk the Balanced ranking penalizes.

This is a study aid, not a solver. The playability simulation behind these rankings is heads-up only, uses a single bet size, and the villain never bluffs or folds — so it under-credits hands that earn fold equity, and it says nothing about multiway pots or unusual stack depths. (That caveat is about the rankings — the equity calculator itself computes multiway spots.) It is calibrated against a simplified model of postflop play, not solver EV. Treat the Balanced ranking as a lens for “which hands realize their equity better,” and keep using equity and a real solver for decisions where precision matters.