Insights · Beta
Thesis Performance Analytics
5D, 10D, 15D, and 30D accuracy, returns, excursions, calibration, and confirmation lift
Open in Enerlytics →How To Use It
On Pro or Team, or during an active Starter checkout preview, enable “Show full-workspace panels” in the builder, then search for Thesis Performance Analytics, add it to a wide canvas and save the dashboard. On Starter after the preview, the saved card shows an upgrade prompt instead of analytics.
What The Measures Mean
Directional accuracy is correct directional outcomes divided by mature directional outcomes at the selected 5D, 10D, 15D or 30D horizon. Directional return adjusts the underlying return for the thesis direction; forward return does not. MFE/MAE are average best/worst path excursions. Brier score uses only separately declared probability outcomes, not heuristic confidence. Creation-versus-confirmation lift uses paired mature cases, so always read the sample counts.
A Worked Example
Example: 3 mature 5D outcomes and 20 pending 30D windows require different denominators. Counterexample: a high percentage on three outcomes is not a validated trading edge.
Limits And Common Misreads
Thesis Engine is beta research. A proposed or active hypothesis is not a verified event, official Model Signal forecast or live trade. Outcome samples can be small.
Source And Freshness
Look for the source label and as-of or publication time on the live card. Different feeds refresh on different schedules. Read the freshness guide before comparing them.
- Enerlytics application state and linked evidence
Last reviewed 2026-11-10. Examples are illustrative or historical as labeled; open the dashboard for current observations.
Was this guide helpful?