Skip to main content

Metrics

Metrics help you compare historical behavior. They do not predict future performance.

Common metrics:

  • orders: closed historical orders
  • wins and losses: positive and negative closed outcomes
  • winRate: win share among closed orders
  • netProfit: net result in the run output
  • maxDrawdown: largest historical decline from a previous peak
  • riskRewardRatio: average reward/risk relationship when available

How To Use Metrics​

Use metrics to compare:

  • strategy versions
  • config variants
  • symbols
  • time windows
  • runtime vs backtest behavior
  • AI/ML gate changes

Do not optimize only one number. A high net result on one short window can be overfit, data-dependent, or caused by unrealistic execution assumptions.

Recent Runtime Metrics​

The current app and runtime have more strategy-card and order-drawer metrics than early versions, including runtime strategy analytics, loss streak display, and replay/parity reports. Use those views as diagnostics, not as proof that a strategy is ready for live automation.