privacy · proofs · compute
v3.2 · checksummed

Stats

No dashboards, no trust — applied to this site.

This site measures itself the way the thesis asks a protocol to: in public, in aggregate, and with the method stated. Collection is cookieless. No identifier is set, stored, or shared across sites; no IP address is retained; browsers sending Do Not Track are not counted. What is recorded: page views, referrers, country, and the following named events.

EventMeaning
rung_clickA reader moved one rung down the ladder (start → banknote → essay → brief → record → review).
explainer_clickA reader followed a door out of an explainer essay, by name (record, essay, or brief) — which door an explainer sent the reader through.
pdf_downloadA PDF was fetched, by file name.
outboundA link to SSRN, Substack, or GitHub was followed.

The weekly digest built from these numbers, with the manual rows (subscribers, preprint downloads, replies) added by hand, is committed to the repository as telemetry/vitality/YYYY-WW.md and summarised in the updates feed each month.

Day-60 targets

Reported at day 60 from the first push; the baseline is the week before it.

MetricTarget
Hostile readers replying substantively≥ 3 of 5
Third-party reviews published on the site≥ 1
SSRN downloads, paper of record≥ 200
Citations or serious long-form engagement (post, podcast, paper)≥ 1
Banknote → record click-through≥ 5 %
Substack list growth vs. pre-push baselinereported, no target

If by day 60 no peer has engaged and record downloads are flat against baseline, the constraint is the artifact, not distribution: the track stops and work returns to the three papers.

The live view is not configured on this build. The numbers are published in the weekly digest instead.