Sealed before. Counted after. No proof.
Every statement is sealed before the measurement and evaluated unchanged afterwards — this prevents a result from being relabelled after the fact as though it had been stated in advance. What stands here is a tally of that procedure, not proof of skill. A single SUPPORTED result is NOT a skill claim; pending seals are shown as INCONCLUSIVE_PENDING until their target time elapses.
Targets AGE UNDETERMINED
| Target | Subject | Verdict | Origin |
|---|---|---|---|
| T1 | SHAPE | SUGGESTIVE | AUTHORED |
| Beats baselines on log-loss (calibration, not accuracy). Not proof. | |||
| T2 | STRUCTURE_REGIME | TIES_CLIMATOLOGY_ACC · WINS_LOGLOSS | AUTHORED |
| Ties climatology on accuracy (baseline-easy) — the real win is calibration (log-loss). | |||
| T3 | ENERGY_STRESS | REFUTED_DEGENERATE | AUTHORED |
| Refuted as defined — near-constant target (HIGH band never realized); no model can express skill. | |||
| T4 | ELEMENTS | MAJORITY_BEATS_CLIMATOLOGY | AUTHORED |
| 22/30 elements beat climatology on log-loss (73.3%). Strongest signal in the set — still suggestive, not proof. | |||
SUGGESTIVE is not proof. Mechanism validation — historical walk-forward is SUGGESTIVE, not proof. Skill measured by proper scoring rule (log-loss / Brier) vs climatology & persistence baselines. Not a forecast, no edge, no direction. The numbers in the detail lines are read from the report; the verdicts are written as constants in the code and marked AUTHORED. Deriving them would require a threshold, and a threshold is a decision, not a measurement — so none is invented here. Were the underlying measurement to change, an authored verdict would not follow it.
Classification
| Classification | RESEARCH_SHADOW · MECHANISM_VALIDATION · kernel_state_hash=null |
predictive_claim | NONE |
| Open, pending | 0 |
A predictive claim would require many pre-registered, out-of-sample confirmations accumulated over real elapsed time. None exist yet.
Provenance
| Pre-registration | reproduce_leg/…/prereg_ledger/prereg_ledger.jsonl |
| Pre-registration | reproduce_leg/…/prereg_ledger/prereg_outcomes.jsonl |
| Evaluation | reproduce_leg/…/SEER_LEARNING_REPORT.json |
| Evaluation | reproduce_leg/…/LEARN_T2_REPORT.md |
| Evaluation | reproduce_leg/…/LEARN_T3_REPORT.md |
| Evaluation | reproduce_leg/…/LEARN_T4_REPORT.json |