What you get, exactly

The whole tree, one node at a time

Type any board. The tool shows the aggregate strategy at that node, then breaks the range into 97 hand categories so you can see which classes bet, which ones protect the checking range, and which mix. Click through any turn and any river; the breadcrumb rewinds the whole line.

Underneath, the tool tracks each player's actual arriving range: every action multiplies its frequencies through all 270,725 starting combos, so the river hand list shows the weights hands really arrive with, not their preflop weights. Hover a hand to see how much of the opponent's range it blocks. A runout grid shows the strategy on every possible next card at once, and a PPT-style filter (AA**, RR:$ds, suit variables) cuts the list to whatever you're studying.

Preflop is free. All of it.

Every opening, 3-bet, 4-bet, 5-bet, 6-bet, and vs-limp range, for every seat, at all nine depths, browsable by 31 hand categories in 11 groups with suitedness splits. The 6-max solves also carry per-hand EV in big blinds, so you can sort a range by what a hand actually earns instead of how often it gets played.

A preflop trainer quizzes the same lines, dealing hands weighted by how often they reach the spot, with fold and call buttons always present so 0%-frequency traps exist.

None of this sits behind the paywall. Preflop is where mistakes compound the longest, and giving the ranges away beats any ad we could run.

Leak reports that have to prove it

Most leak detection is population color-coding: your number differs from an average, so it turns red. Ours is stricter. Preflop verdicts come only from comparing your decisions against our own solves, matched to your table size and stack depth. PLO4 grades against nine solves (eight 6-max depths plus heads-up); NLHE grades against eight 6-max depths from 20 to 200bb. Both check opens the solver plays 10% of the time or less, folds of hands it opens at least 90% of the time, your 3-bet frequency against per-matchup solver targets weighted by the openers you actually faced, and how you defend against opens. PLO also checks how you continue your own opens facing a 3-bet. The river verdicts are exact pot math: your bluff ratio against the indifference frequency, your folding against the MDF your opponent's sizing sets, buffered and split by river texture so one weird node can't trigger anything.

Nothing gets flagged until a one-sided Wilson bound proves the gap at your sample size. Small sample, no verdict. The report grades discipline A to F on proven deviations per 100 hands and ranks what to fix first by the size of the proven gap.

Where the matching solve doesn't exist (PLO5, PLO6, heads-up NLHE, unusual table sizes or depths), you still get the exact-math river checks, but the solver comparisons are labeled observations, not verdicts.