Solver viewer

The viewer walks a solved tree: what the solver does with every hand at every node, its EV and equity, and how the strategy changes when you lock one player's play. It reads your own solves and the library boards on this machine.

Opening it

• Solve › Viewer: under Open a solve, your own solves sit under Your solves (or their folder) and library boards are grouped by spot, with a search box. A large save can take a while to open.
• When a solve finishes in Solve › Build, Open in Viewer goes straight to it.
• From the trainer, Open in viewer ↗ after you answer, and the ↗ on History and Biggest mistakes rows, open a separate viewer window on that exact node. So does Open node on a graded decision in Analyse. A note you wrote on a trainer decision shows beside the solve.

The trainer deals each hand with the suits renamed, so the board you played may not match the stored tree's suits. When that happens the viewer says so: it is the same spot with the suits renamed.

Reading a node

• The header shows the pot and each player's EV, equity and equity realisation. Explorer opens the range explorer panel.
• The Line across the top shows where you are, street by street. Click any action in it to go back to it or to see the actions not taken.
• Go to has one button per action at this node.
• View modes: strategy+EV (the default), ev, equity and range, the last three for OOP or IP. On a turn or river decision, runouts shows the same line on every card the street could have dealt, each with its strategy, EV or equity.
• round rounds the displayed frequencies to the nearest 25%. It changes the display only.

The hand grid

In strategy+EV each cell is a bar split by how often the solver takes each action, with the acting player's EV for that hand. In the other modes cells are shaded as a heat map. A hand not in the range is left empty.

• Hover a hand to see its individual combos below the grid, each with its own action frequencies, and EV and equity where the save has them. Click to pin it there; click again to unpin.
• suits (strategy view) splits each cell into per-suit rows. It is off by default and remembered.
• The suit filter (the four suit buttons) shows only combos holding at least one of the picked suits. Shares re-weight over what is left; the EV and equity in the header stay the whole range's.
• Click an action tile above the grid to show only the range that takes that action. A chip shows the filter (click its ✕ to clear it), and Continue down … → walks on to the node after that action.

The range explorer

Explorer in the header opens a panel for OOP or IP: the range split by made-hand class and by draws, as a share of the range, and both players' equity distribution on one chart. Click a row to highlight those hands on the grid. With an action filter on, the panel shows the range after that action.

When a node has no EV

Library boards are saved without their rivers to keep downloads small. On such a save the EV is missing wherever an unsolved street sits below the node, and the ev mode is marked ⚠. Walk to a river (its subtree solves on demand) or open a fully solved save. A missing EV reads as missing, never as 0. See Why some EV comparisons are missing.

Changing the solve

Lock & exploit

Under the grid, Lock & exploit (collapsed until you open it) fixes the acting player's strategy at this node and re-solves the other player's response:

• Lock current strategy, Lock pure on any one action, or Lock mixed… with your own frequencies; Unlock and Unlock all undo them.
• Re-solve response runs the re-solve. The exploit report then sets the baseline against the locked solve: each player's EV, how exploitable it is, and the difference.
• Remove donks (this street), on a turn or river: locks the out-of-position player to check at the start of this street on every card it could have come, so they never lead after the other player bet the last street. Tick + re-solve to run the re-solve straight after.

Best single size

In the range explorer panel on a turn or river, Best single size races sizes for one player's first bet. Pick OOP or IP, list sizes (% of pot, a for all-in, 2e for geometric; blank uses the trainer's Max EV candidates) and press Race sizes. Each row is the street re-solved from its root with that player betting only that size, with its EV, and how far behind the best it is, tied when within the margin, or dropped once its best case could not reach the leader. A race is several solves, so it runs only when you ask.

Subtree from here

On a turn or river, Subtree from here (and Re-solve this street with your sizes → in Lock & exploit) opens Solve › Build with this street's board, both ranges, the pot and the stack, from the start of the street, so you can re-solve it with other sizes. From a popout window it goes to the main window's Build.

Rules here

Where a library package has mined rules for the node you are on, Rules here lists them, each with what following it costs in bb, or ≈ free when the cost is below solve noise. See Texture buckets.

Keyboard

1–9 take that action along the Go to row, Backspace goes up to the parent, Alt+←/→ step back and forward through where you have been, and on runouts Ctrl+arrows change the card.

Colours

Fold is blue and check or call green by default. Bets and raises are coloured by their rank at the node: the smallest amber, each larger size a darker red, so the same colour can be a third-pot bet at one node and an overbet at the next. Settings › Appearance › Colours › Bet colours › Fixed by size (% of pot) gives each size band one colour everywhere instead; preflop sizes keep the ranked colours. The same section has presets, including a colour-blind safe one, and a picker for every action and hand class. The order actions are listed in follows Settings › Table & numbers › Action order.