NOVI / RESEARCH INTELLIGENCE
Research Observatory
Open Prediction Lab · prospective model tournament →
Follow the evidence. Inspect the forecast. Review what changed.
01 / MEASURE
Prediction reliability
Current model version, asset and horizon. Descriptive results are separate from chronological validation.
Prospective regime breakdown
Method and evidence limits
02 / INSPECT
Decision replay
Only snapshots actually collected at or before your chosen time are shown. Outcome information is revealed only when it became available.
Frozen specialist evidence and forecast scenarios
03 / EXPLAIN
What changed
04 / WATCH
Derivatives stress
Observed liquidation sample
Dated-futures basis
Observed contracts and data coverage, with paused or stale contracts excluded. Estimated concentrations are never presented as observed liquidations.
05 / RESEARCH
Event & research timeline
External reporting is untrusted evidence. “First seen” is NOVI’s collection time. General market research is labeled separately.
06 / REVIEW
Evidence alerts
Experimental consensus threshold + momentum alerts require NOVI’s existing data and governance checks. Signals may be correlated. Alerts expire after 15 minutes; direction is reviewed at one hour.
Collection & coverage
Optional commercial datasets: labeled exchange-wallet flows and modeled liquidation concentrations are not connected. No subscription has been purchased. Forecasts, risk controls and execution remain governed by NOVI’s existing policies.