Starting Hands

Rankings are Monte Carlo equity against random opponents: 1 opponent for the heads-up ladder, 2 for 3-handed, 5 for 6-max. Each PLO4 hand class gets its own seeded run in a Rust evaluator - 400,000 trials per class in each field size, plus a 100-million-trial heads-up refinement on 15 top AA classes. NLHE uses 5 million trials heads-up and 2 million at 3-handed and 6-max. PLO5 uses 200,000 heads-up and 201,000 at 6-max; PLO6 uses 50,000 and 22,500. PLO5 and PLO6 have no 3-handed table. One row per strategically distinct hand: suits only matter through their pattern, so AsAhAdKs and AsAhAdKh are the same class and share one row. PLO5 and PLO6 use binned equity tables, so their absolute percentile is bin-level. Solver open ranges are read straight from our current 6-max 100bb solve (profile v20310_g1_fb100_tb75_rb75_tt1_rt1). They are stored in 0.5-percentage-point steps and do not sample Monte Carlo hands.

Monte Carlo estimates that are close in equity may change order between builds.

The category bars run on a square-root-stretched percentile axis, so the strong end gets more room. Each bar spans its category's 10th to 90th percentile, the bright core is the middle half (25th to 75th), and the tick is the median.

The hand categories are the same taxonomy used across SolvePoker - the solver's preflop study, the hand history Leaks tab, and this page all speak the same vocabulary. For paired hands, “+ Rundown” means connected side-card support; five- and six-card hands can still have a dangler.