PLO5 6-max — 100bb — technical details

Overview

Game

Pot-limit Omaha 5-card · 6-max

Stack depth

100bb effective

Hands

134,459 PLO5 hand classes (suit-isomorphic)

Solver

HydraSolver

Price

$100, one-time

Tree

Raise sizes

Pot-limit

Rake

5%, 1bb cap, 1.2bb ante

Streets

Preflop only

Solver run

Solver game

Omaha 5 Card Hi

Players

6

Stacks

99.8bb

Rake

5% up to 1bb · postflop pots only

Streets saved

1 street

Iterations

52,948,218

Solve time

71h 40m

Save format

2

Saved solve

plo5-6max-100bb-r5c1.hds · 3.9 GB · saved 2026-10-25 22:56

File

Format

Zip of per-node gzipped JSON

Download

PLO5-6max-100bb-ante1.2-rake5-1bbcap-json.zip · 3.7 GB

Opens with

Any JSON tooling — gunzip a node file and parse it

What's inside

README.txt

How the pack is laid out and read

index.json

Every preflop decision node and its actions

nodes/<line>.json.gz

One file per decision node, named by its action line

How strategies are stored

• —Each node file holds every action available at that node.
• —Each action lists its hands as hand:freq:ev — the hand class, how often it takes the action, and its EV.
• —Hands are suit-isomorphic classes, so one entry stands for every suit permutation of it.

Using the data

1. Unzip the download. Everything sits in one folder named after the pack.
2. Open index.json to find a spot: each node lists its action line (history), the player to act (actor), the actions available and its file under nodes/.
3. Gunzip that node file and parse the JSON. "#meta" holds the line, the actor and each action's share of the range (poolPct); every other key is one action.
4. Split an action's value on commas, then each entry on colons: hand, frequency (0 to 1) and EV in big blinds. EV is empty for hands that never reach the node.
5. One hand's frequencies across a node's actions sum to 1, so the node file alone gives the full strategy for every hand.

import gzip, json
from pathlib import Path

pack = Path("PLO5-6max-100bb-ante1.2-rake5-1bbcap-json")

# Every decision node: who acts, the actions, and the node's file.
index = json.loads((pack / "index.json").read_text())
print(len(index["nodes"]), "nodes")

# One node: the first decision, LJ to act.
node = json.loads(gzip.decompress((pack / "nodes/root.json.gz").read_bytes()))
print(node["#meta"]["actor"], "to act")

for action, rng in node.items():
    if action == "#meta":
        continue
    for entry in rng.split(","):
        hand, freq, *ev = entry.split(":")
        freq = float(freq)
        ev = float(ev[0]) if ev and ev[0] else None  # bb; None if unreached
        # ... your analysis here
    print(action, "parsed")