Home/Findings/The key that outlived its pool
Finding · free in full · census heights 0–960,000 · live razor stamped
The key that outlived its pool
Bitcoin’s original payout script nearly died — and since May 2014 every single P2PK coinbase output, 716 of 716, pays one key.
Pay-to-pubkey is the form Satoshi’s blocks used. It nearly died out. But since 24 May 2014, every P2PK output in any Bitcoin coinbase pays the same key: a donation constant in the published source code of P2Pool, the 2011 peer-to-peer mining pool. Someone collected those donations until 4 April 2018, then stopped — the boundary is 20.6 days wide and perfectly clean — and never came back. Hobby P2Pool nodes are still finding mainnet blocks (last in our census: 27 June 2026) and still minting dust tips to a key nobody touches.
Registry: F-2026-07-31-015 · CONFIRMED T1_CONTROLLED. Bridges the
Sleeping Fleet P2PK story with zero ownership conspiracy.
One spend destroys this claim
One spend destroys this claim. That’s why it’s date-stamped, and why the verifier checks it against your node’s live UTXO set, not ours.
Download verify_donation_key.py. Standard library only. No pip. Landmarks, a 30-block
sample of the 716, and the razor via gettxout — about 87 RPC calls, ~1 second.
Download verify_donation_key.py The razor Mandatory limits
python3 verify_donation_key.pyNo curl-pipe-bash. Serve and run the file as-is.
Two independent re-checks, hours apart — the claim is actively re-checked, not asserted once:
- 2026-07-31T22:16:11Z · tip 960,470 · 140/140 outpoints agree (UTXO snapshot vs live Core
gettxout) - 2026-08-01T02:38:21Z · tip 960,497 · all 19 unspent outpoints still unspent · total exactly 64,790,575 sats · spent-side boundary h=516,592 still spent
Authoritative ship-time stamp: 2026-08-01T02:38:21Z. A later spend falsifies the razor — that is the feature, not a bug.
Census · heights 0–960,000 · publish whole
Distinct keys: 32,490 → 1
After block 302,467 (2014-05-24), every P2PK coinbase output pays the donation key — exhaustive count, no exceptions. UTXO snapshot height 959,975. 2021 is the only year with zero P2PK coinbase outputs.
Distinct P2PK coinbase keys by year (log scale)
2009: 32,490 → 2015 onward: 1 (except 2021: zero outputs). Values from the published table.
| Year | P2PK coinbase outputs | BTC | Distinct keys |
|---|---|---|---|
| 2009 | 32,490 | 1,624,502.87 | 32,490 |
| 2010 | 67,920 | 3,395,543.99 | 67,914 |
| 2011 | 58,477 | 2,871,586.36 | 58,284 |
| 2012 | 31,082 | 1,460,941.34 | 29,624 |
| 2013 | 5,049 | 110,676.43 | 3,920 |
| 2014 | 1,091 | 15,099.53 | 67 |
| 2015 | 230 | 5.30 | 1 |
| 2016 | 46 | 0.98 | 1 |
| 2017 | 70 | 0.30 | 1 |
| 2018 | 12 | 0.09 | 1 |
| 2019 | 5 | 0.54 | 1 |
| 2020 | 1 | 0.02 | 1 |
| 2021 | 0 | — | — |
| 2022 | 2 | 0.00000028 | 1 |
| 2023 | 1 | 0.00000030 | 1 |
| 2024 | 1 | 0.00000023 | 1 |
| 2025 | 1 | 0.00000020 | 1 |
| 2026* | 1 | 0.00000027 | 1 |
*2026 partial to frozen height 960,000. 2021 is the only zero year.
The key and the takeover
One public key. Exhaustive exclusivity.
The key (uncompressed; public chain data and published source):
04ffd03de44a6e11b9917f3a29f9443283d9871c9d743ef30d5eddcd37094b64d1b3d8090496b53256786bf5c82932ec23c3b74d9f05a6f95a8b5529352656664b
- First coinbase use: block 142,312 · 2011-08-23T23:20:35Z · 25,000,008 sats
- Lifetime: 2,545 outputs · 217.17465776 BTC
- Last other-key P2PK coinbase: block 302,467 · 2014-05-24T22:59:00Z · 2,511,992,894 sats · compressed key (
03…) - After the cutoff: 716 / 716 pay the donation key
The key is a constant in P2Pool’s published source — a software fact. Who controls it is not knowable from the chain.
The razor · F-2026-07-31-015
Collected until April 2018 — then never again
Spent: 121 · last block 516,592 · 2018-04-04T12:50:01Z
Unspent: 19 · first block 519,817 · 2018-04-25T02:32:18Z
Gap: 20.6 days, perfectly clean. Unspent total: 64,790,575 sats (~0.648 BTC), dominated by seven 2018–2020 outputs of 7.6M–11.5M sats each.
The razor as a timeline (April 2018)
Verdict: Whoever held the key collected 121 times over four years and stopped between those two April 2018 blocks. Why, we do not know — and the record says so.
The ghost · dust tail · publish whole
Still minting. Still untouched.
| Height | Date | Sats | Marker |
|---|---|---|---|
| 562,667 | 2019-02-12 | 11,328,891 | — |
| 615,433 | 2020-02-01 | 2,215,957 | p2p-spb.xyz |
| 732,085 | 2022-04-16 | 8 | Eurasian server |
| 767,471 | 2022-12-15 | 20 | p2p-spb.xyz |
| 785,117 | 2023-04-12 | 30 | p2p-spb.xyz |
| 824,116 | 2024-01-03 | 23 | p2p-spb.xyz |
| 886,688 | 2025-03-07 | 20 | p2p-spb.xyz |
| 955,609 | 2026-06-27 | 27 | p2p-spb.xyz |
Marker strings are self-reported bytes, reproduced verbatim — identifying nothing beyond their presence.
Controls · the product
Four controls. All passed.
| Control | Result |
|---|---|
Two external instruments (UTXO snapshot vs live Core gettxout) | 140/140 agree, date-stamped |
| Independent raw re-parse of 8 landmark coinbases | 8/8 exact — and it caught its own instrument first (parser missed compressed P2PK; census was right; fixed same session, never published) |
| Exclusivity | exhaustive count, not a sample |
| Reader verifier on our node | landmarks byte-exact · 30/30 sample · razor live 19/19 · 87 RPC calls, 1s |
Mandatory limits
What this does NOT claim
- No person is named. The key is a constant in P2Pool’s published source — a software fact. Who controls it, who swept until 2018, and whether they are the author are not knowable from the chain. The page must never imply otherwise.
- No reason for the stop is claimed. Lost key, lost interest, death, choice — indistinguishable in the bytes. OPEN.
- Marker strings are self-reported bytes, reproduced verbatim, identifying nothing beyond their presence.
- “Every P2PK coinbase output” — coinbase outputs with P2PK scripts only. Other populations are other claims.
- The razor is date-stamped and falsifiable by one spend. The page shows the stamp; a later spend falsifies it — that is the feature. Retraction with equal billing if it moves.
- Coverage 0–960,000 for the census.
- Spend status is outpoint-level (this txid:vout in the UTXO set) — never script- or address-level. Homage deposits can never read as spends. Same discipline as Sleeping Fleet.
This finding is free-only
No sixth product. Sealed datasets remain the three existing SKUs. Free kits on those pages before pay.
Cite
Citation
Copy for papers and notes. Prefer the free finding URL. Always carry the date-stamp and the falsifiability note with any citation of the razor.
BibTeX
@misc{parallax-key-that-outlived-its-pool-2026,
author = {PARALLAX},
title = {The key that outlived its pool},
year = {2026},
howpublished = {\url{https://parallaxbtc.com/findings/key-that-outlived-its-pool}},
note = {Since May 2014 every P2PK coinbase output (716 of 716) pays one P2Pool donation-key constant; collection stopped with a 20.6-day clean gap in April 2018. Razor date-stamped and falsifiable by one spend. Free verify\_donation\_key.py against the reader's live UTXO set. Pseudonymous research.},
url = {https://parallaxbtc.com/findings/key-that-outlived-its-pool}
}
RIS
TY - ELEC
AU - PARALLAX
TI - The key that outlived its pool
PY - 2026
UR - https://parallaxbtc.com/findings/key-that-outlived-its-pool
N1 - 716/716 post-May-2014 P2PK coinbases pay one donation key; April 2018 razor 20.6d; date-stamped; free verify_donation_key.py.
ER -
Plain URL: https://parallaxbtc.com/findings/key-that-outlived-its-pool