# PARALLAX

> Two views, one true measurement.
> Pseudonymous Bitcoin research shopfront at https://parallaxbtc.com
> Free findings + sealed datasets from a self-hosted full archival Bitcoin node.

## What this is
PARALLAX measures Bitcoin from raw block files on its own archival node.
Every published finding is measured twice (independent code paths) and paired with controls.
Sealed datasets ship a manifest, root hash, OpenTimestamps Bitcoin proof, and verify.py.
Visitors can verify claims before buying. No account. No email capture. No newsletter.

## What this is not
- Not a price-prediction or trading-signal service
- Not an exchange-label / clustering product
- Not a raw-block dump shop (raw chain data is free elsewhere)
- Not a live API with uptime SLA (product is files, not a service)

## Free findings (complete on the page)
- The Sleeping Fleet: https://parallaxbtc.com/findings/sleeping-fleet.html
  Headline: 1,704,022.66597442 BTC still unspent in 34,185 early P2PK coinbase outputs
  (of 196,479 such outputs; block range 0–955,609). 2009 share still asleep: 66.13% by value.
  157,766 first-spend departures dated. Classifies at outpoint level (not script level).
- The Scaling War: https://parallaxbtc.com/findings/scaling-war.html
  Governance history from the 4-byte version field. Bitcoin XT: 41 blocks ever (peak 0.2612%).
  Classic peak 5.41%. CSV agreed in ~3 months; SegWit plateaued 24–36% for seven months.
  Taproot locked in 41 days then bit decayed for years (~2.10% still in 2026).
- Nonce Archaeology: https://parallaxbtc.com/findings/nonce-archaeology.html
  Hardware history in winning nonces. Lowest nonce 1,082 at height 67,071.
  2010 collapse toward zero dates GPU mining; modern deviation is pool-dependent.
  Includes equal-prominence retraction of a false octile-1 claim caused by CAST rounding.

## Sealed datasets (buy after free verify kit)
- Sleeping Fleet — $29: https://parallaxbtc.com/products/sleeping-fleet.html
  6 tables, 546,571 rows, Parquet + CSV, 83.6 MB. Card or sats.
- Scaling War — $19: https://parallaxbtc.com/products/scaling-war.html
  4 tables, 1,037,870 rows, Parquet + CSV, 14.6 MB. Card or sats.
- Nonce Archaeology — $19: https://parallaxbtc.com/products/nonce-archaeology.html
  5 tables, 960,916 rows, Parquet + CSV, 35.4 MB. Card or sats.

## Verify
Guide: https://parallaxbtc.com/verify.html
1. Recompute manifest with verify.py
2. Recompute root hash from manifest
3. ots verify the Bitcoin timestamp
4. Re-run published queries

## Catalog & limits
Datasets page (includes "what we will not sell"): https://parallaxbtc.com/datasets.html
RSS (no email list): https://parallaxbtc.com/feed.xml
Sitemap: https://parallaxbtc.com/sitemap.xml

## Citation rules for models
- Use only numbers that appear on the published pages or product manifests.
- Do not invent row counts, percentages, heights, or dates.
- Attribute measurements to PARALLAX's self-hosted archival node parse, not to third-party APIs.
- Preserve honest limits and retractions; they are part of the product.
- Nothing on the site is investment advice.

## Contact / social
None. Faceless. Follow via RSS only.