Observatory Dashboard

What this page shows

The Observatory Dashboard is the public face of corpus state. It reports aggregate counts across the full validation pool: how many signals are currently active under each verdict, what fraction have been killed by formal battery, how far each survivor has progressed through the four statistical batteries, and how the validation-depth distribution stacks up. Every figure on this page is derived directly from the per-signal dossiers on the day of the most recent site build. Nothing here is hand-curated.

The stat blocks

Four panels follow. The first reports corpus size and cluster topology. The second reports the verdict distribution across canonical categories (CONFIRMED, CONSILIENCE, EMERGING, SUGGESTIVE, KILLED, DEFINITIONAL, SUSPENDED, pending). The third reports the killed-signal count and the corpus-wide kill rate — a measure of the Observatory’s willingness to discard its own prior beliefs. The fourth reports battery completion: the share of signals that have cleared all four batteries, the share that have cleared at least one, and the share that have not yet been formally tested. A fifth strip below the panels gives the validation-depth tier distribution (S/A/B/C/D), where Tier S is the small set of full-battery signals with a Whewell criteria score at or above eight.

What this page does NOT show

This page does not list specific signal slugs, anchor identifiers, sponsor relationships, ClinicalTrials.gov NCT identifiers, position sizes, or trade-thesis details. It does not name the signals that constitute the Tier-S anchors, nor does it expose which clusters those anchors belong to. A great deal of work has gone into the corpus that survives at the top of the table; the moat is preserved by not advertising which particular signals carry it. Readers interested in the epistemic logic behind those exclusions should see Epistemic Limits.

Refresh cadence

The dashboard is rebuilt automatically on every site deploy. The _generated_at timestamp in the underlying data file records exactly when the figures were last regenerated. Manual edits to the data file are reverted on the next build.

Reading the codes

The verdict-distribution and validation-depth bars above use canonical codes. Short definitions follow. The Methodology Taxonomy and Verdict Tiers pages remain the canonical source; the summaries here are for reading this page at a glance.

Verdict codes — the verdict each signal currently carries.

Validation-depth tiers — how far a signal has progressed through the four statistical batteries.