The record so far

We let the future grade us. Here's what we registered. Here's what happened. Every registered forecast leaves a record: the original prediction, the deadline, the rule that settles it, and what has been read so far. See what's still open, what's been resolved, and what remains unresolved. Registry snapshot Nov 21, 2026.

Registered68
Scored6
Machine-resolved0
Scored by people6
Ambiguous1
Unresolved61

What the counts mean

7 of 68 records have been through outcome review. 0 scored outcomes were resolved by two agreeing machine readers and are never admissible. No scored outcome has been reviewed by a person.

A machine resolution is reported and labelled, never counted (DOCTRINE §5, Amendment 7).

By reference class

A figure across a class appears only when the class has 15 admissible resolved forecasts. The largest admissible count in any class is 4, so no class figure is printed.

Registered forecasts and adjudications by reference class
Reference class Registered Scored Machine Ambiguous Unresolved Admissible
NCCN guideline reclassification 2 0 0 0 2 0
FDA label revision 2 0 0 0 2 0
FDA action by goal date 36 2 0 1 33 2
NICE guidance by scheduled date 28 4 0 0 24 4

The counts for each class are the canon's, sent with the export. A forecast that turns on the same decision counts once in the admissible column; that count is the one the reporting rule gates.

What has been read

Each finding is printed with the count for its class. No class has 15 admissible resolved forecasts, so no figure across a class is printed here.

FF-017 · National Institute for Health and Care Excellence (NICE) · registered 50% on 2026-09-26 · adjudicated 2026-11-10

NICE publishes technology appraisal guidance on Inavolisib with palbociclib and fulvestrant for treating recurrent hormone receptor-positive HER2-negative PIK3CA-positive advanced breast cancer after adjuvant endocrine treatment [ID6425] on or before 2026-11-03

FF-017: did not occur. NICE guidance by scheduled date: 4 resolved, 24 unresolved, 0 expired (28 registered). One forecast is one instance, not a rate.

How a forecast is settled

The rule is frozen with the forecast, before the outcome is known. After the deadline, the outcome is read against the rule by the founder or by two agreeing machine readers. A machine resolution is labelled and never counted; if readings diverge, the record is ambiguous and is never scored.

  • Occurred on time
  • Did not occur
  • Occurred after the deadline
  • Occurred in part; the rule was not met
  • Ambiguous, handled by the rule
  • Could not be resolved
  • Cancelled by an outside event

The seven statuses are the canon's. The full outcome review, including the individual findings, stays on each record page.