The strategy library

Verdict Strategy Builder is built around a library of strategies. Each strategy owns its own builds, runs, results, and exports, and several strategies can be running searches at the same time — two MES ideas and an NQ idea side by side, if you like.

Home#

Home (sidebar → Home) is the hub:

• Get started — a checklist that ticks itself off as you work: Get your data → Load the sample → Your first build → Run the gauntlet → Export to NinjaTrader. Resume tour replays the guided tour over the real screens; Skip collapses the checklist.
• Strategies — a card per library strategy: instrument and timeframe, last-run state, an equity sparkline, and the best net PnL found so far. New strategy creates one; Import spec… brings in a saved spec.
• Active runs — every run currently queued, evolving, or validating, across all strategies.
• Favorites — a preview of your latest starred results.

The strategies list#

Strategies (sidebar → Strategies) lists the whole library with rollups per strategy: linked instrument, build and run counts, best fitness, best PnL, a Pass/Marginal/Kill verdict tally, and the last run's state. Click a row to enter that strategy's workspace.

A strategy usually links one dataset. The Data screen shows the reverse view — which strategies use each dataset — in its Used by column.

The workspace and the pipeline rail#

Opening a strategy enters its workspace, with a pipeline rail on the left:

Data → Build → Run → Results → Export

The rail is a live map of the strategy, not just navigation: each stage shows its state — done (with a summary sub-line, e.g. the dataset span or candidate count), active, ready, or locked. Results and Export stay locked until a run has finished, so you can't export something that was never evaluated. The rail's strategy switcher jumps between workspaces without losing your place.

Everything you learned in the stage guides applies per strategy: the Build screen, monitoring runs, reading the results, exporting.

Concurrent runs and the scheduler#

Every strategy can start runs independently; the Runs screen shows the whole fleet with live gauntlet stages. Runs execute in parallel up to the scheduler's slot count (Settings → scheduler); extra runs queue with a visible position and start automatically as slots free. Each live run shares the CPU evenly — more slots means more runs at once, each evolving more slowly.

Linking a dataset#

A strategy can be created with a dataset or link one later — the workspace's Build stage offers the link when it's missing, and changing the selected dataset re-links the strategy. Datasets themselves are shared: import once on the Data screen, use from any strategy.

Where things land#

• Results always live under the strategy whose run produced them — open them from the workspace's Results stage.
• Favorites cut across the library: one board of starred results from every strategy and ticker.
• Builds, runs, results, library strategies, and favorites all back up privately to your account (Account & sync).