Whispertrades

Heatmap

Updated Nov 13, 2026

A sweep of thirty backtests is thirty rows to read. The heatmap turns those same results into a grid: one axis per parameter you varied, one cell per combination, colored by the metric you care about. Patterns that are invisible in a list, such as a plateau of good values or a cliff edge you do not want to sit next to, are obvious on a grid.

Building one

  1. Run a sweep, usually with Sweep set to build separate backtests.

  2. In your backtest list, select the backtests you want on the grid and choose Heatmap. You can also open the heatmap straight from a tag, which is the easier route if you tagged the sweep when you ran it.

  3. Choose the metric to color by.

The axes are worked out from the backtests themselves. Whispertrades compares their configurations, finds the parameters that actually differ across the set, and offers those as axes. You do not have to tell it what you swept.

Views

The View control draws the same backtests four ways:

  • Heatmap: one cell per backtest, colored by your metric. Add metric puts up to two more beside it in each cell.

  • Chart: your metrics along one setting, each in a panel of its own, as columns or lines. It takes up to six metrics. On a sweep of two settings it shows one value of the second at a time, and Lines by draws a line for each value instead.

  • Panels: every metric as a small heatmap, to see whether the strong area sits in the same place from one metric to the next.

  • Table: every metric as a row of values.

Metrics you can color by

  • CAGR, Total Return, MAR

  • Max Drawdown, Ulcer Index

  • Sharpe, Sortino

  • Profit Factor, Win %

  • Annualized Volatility

  • Premium Retained %, Average Gain, Total Trades

Each metric knows which direction is good. Higher is better for CAGR and Sharpe, lower is better for drawdown, Ulcer Index, and volatility, and the color scale follows the metric rather than the number, so the strong cells always read as strong.

Reading the grid

Look for regions rather than winners. A single outstanding cell surrounded by poor ones is usually a coincidence of that particular history, and it is unlikely to repeat. A block of neighboring cells that all perform reasonably is a parameter range your strategy is genuinely comfortable in, and the middle of that block is a safer choice than the best cell on the grid. The figures above the grid show both: Best is the top cell, starred on the grid, and Best area is the setting whose neighbors score best on average with it.

The best-looking cell on a heatmap is the one most likely to be an accident. Prefer the center of a good neighborhood over the peak of a lone spike: a parameter that only works at exactly one value is a parameter that was fitted to this particular history.

Exporting

Export any view, the chart included, as a spreadsheet to do your own analysis on the numbers, or as an image to keep or to share.

Was this page helpful?