A knowledge base earns its keep when it changes what gets tested. Many research teams keep literature in one place and experiments in another. In NOVAQ they are connected, and a note can become a test with its attribution intact.
A claim programme
217 claims from well-known authors and traders were put through 16 years of NQ and ES data with costs and detrending. Fourteen passed the statistical stage. An adversarial stage then asked whether each was market drift, too vague to be a claim, or overlapping with something else. Two survived, under 1%.
That funnel is a useful result by any measure. It gives a measured base rate for trading folklore and a repeatable way to apply it to new material.
Claim validation funnel
Setups from the vault
In October 2026, 15 setups from 7 authors and groups were pulled from the knowledge base into a pre-registered round, using five fixed criteria. One passed, and it proved to be a later version of a trade already under study, which is itself useful knowledge.
Lessons that carry
A probability edge is not a profit edge. And returns tend to gather in volatile years, so a result has to be read against its regime.
Limits
Each test covers one mechanical reading of a claim. Thirteen claims from one methodology had too few intraday observations and count as untested, not refuted. The survivors are not named here.