Home/Research
Research library
Everything we measured. Filterable.
One library for findings, fact cards, common-error answers, method, and signal compares. Every number already lives on a free finding with controls. No invented figures. No price talk.
01 · Free
Read the finding
Full method, tables, controls, and honest limits. Nothing held back to force a purchase.
Findings →02 · Free
Verify the seal
Manifest, samples, root.hash, OpenTimestamps, and verify.py — before you pay.
Verify path →03 · Optional
Buy sealed tables
$19–$29 one-shot files. No account. No subscription. Card or sats.
Datasets →Lab status
Live /health, package sizes, seal ages from free manifests. No fake tip height.
BibTeX / RIS
Copy-paste citations on each free finding (#cite). Example: Sleeping Fleet.
1,704,022.66597442 BTC
The Sleeping Fleet
Early P2PK coinbases still unspent at outpoint level — and the homage trap that makes script-level counts silently wrong.
41 XT blocks, ever
The Scaling War
Governance history in one 4-byte field: XT, Classic, CSV, SegWit plateau, Taproot lock-in and decay.
1,082 luckiest nonce
Nonce Archaeology
Hardware history in winning nonces — plus equal-prominence retraction of a CAST-rounding false claim.
1,704,022.66597442 BTC still asleep
Sharable headline for threads. Links to full Sleeping Fleet method.
Bitcoin XT: 41 blocks ever
Peak month 0.2612% vs 75% threshold.
Classic peak 5.41% · era bounds
2,176 era-bounded vs ~64,928 naive bit-28 forever.
SegWit seven-month plateau
Roughly 24–36% for seven consecutive months before the break.
CSV: what agreement looks like
Short agreement shape vs SegWit stall — same mechanism field.
Taproot: 41-day lock-in, years of decay
Mean bit share still ~2.10% in 2026 — template hygiene.
Lowest winning nonce: 1,082
Height 67,071 · 14 July 2010.
Common claims. Measured answers.
Six claim cards that look right until you recompute — FAQPage schema.
Signal eras side by side
XT · Classic · CSV · SegWit · Taproot — published peaks only.
Measurement pack
Printable leave-behind of published headline figures. No email gate.
How we measure
Own node, measured twice, re-derivable, retractions equal.
Catch us if we lie
Four-step path: manifest → root → OTS → re-run queries.
Who this is for
Jobs, not personas
Journalists · pods
Quote free tables
Press kit, one-pager, no handler, no email form.
Independent researchers
Re-run the seals
verify.py + free kits before any purchase.
Data buyers
Sealed full tables
$19–$29 one-shot after free audit path.
Engineers · historians
Signalling eras
Era bounds, plateaus, controls — not vibes.