Archive status

How to read it

• Every percentage carries its denominator. The headline completion figure moved 79% → 73% the day the denominator got honest; a bar without one hides exactly that.
• SEARCHABLE is not TRANSCRIPT. A transcript existing on disk and a row existing in brain.db/FTS are different facts, and they part company badly — see Money & Tech below.
• Unknown is drawn as unknown, never as zero and never omitted. A show with no dates says so rather than plotting a gap at year zero.
• “n/a” is not a gap. A podcast feed's original form is audio: there is no video and none is missing.

Words: held vs distinct broadcasts

The 7.16 M-word difference is duplicate uploads of the same broadcast — a Bitcoin Group night exists as up to three uploads, so row totals count some broadcasts three times. Quote the distinct-broadcast figure; never quote a corpus word count without saying which of these two it is.

The loop flag is not a word-count correction. 714 of 9,386 rows (8%) carry repetition_loop, and those rows do hold roughly two fifths of the words — because they are roughly two fifths of the duration, not because they are inflated. Phase 0 measured the word rate inside flagged rows at 168 wpm against 157 wpm for every other row: statistically identical, with excess attributable to repetition coming out negative, about −140,000 words. One flagged 131,000-word transcript has 96 duplicate sentences, which is 1%. What the flag catches is faithful transcription of genuinely repeated audio — the sung outro, conversational filler, a sponsor read repeated eleven times across a four-hour stream. Treat it as a transcription-quality signal and as a reason to filter before per-episode term-frequency work, not as a discount on the word count.

The 100% goal — per show, per lane

Thomas's six own shows to 100% on video, audio and transcript across both lanes. This is the view that matters, because a single per-show bar hides a lane: audio is complete in the video lane and barely started in live. From _phase0/lanes.json (generated 2026-10-24 22:43); per show per lane; denominator is held + probed-unheld. See lanes.py docstring for what the denominator is not.

The shorts lane is listed for completeness and is not part of the goal: a Short is an auto-cut of a parent broadcast, is never an episode, and was never going to be given its own audio extraction.

“loop-free” is a transcription-quality column, not a holdings one — see the words panel above. It is not part of the 100% goal.

MB-WCN mirror — WCN, MB, TIB, TBG and BTS all live here

“surplus” is held here and no longer listed on either channel — this is the only copy of those.

Every other archive, by lane

External denominators — owner playlists

• SFBM — the SFBM channel’s own playlist “SF Bitcoin Meetups Aug 2014 - July 2015” has 22 members (0 unavailable). Of those we hold 22 as media on the drive and 22 as rows in brain.db (22 either way). Two independently-derived counts, printed side by side on purpose: where they disagree, the disagreement is the finding. Reported as a cross-check — not adopted as the denominator, and not baked into this page.
• SFBM — the WCN channel’s own playlist “Silicon Valley Bitcoin Meetup” has 3 members (0 unavailable). Of those we hold 0 as media on the drive and 0 as rows in brain.db (0 either way). Two independently-derived counts, printed side by side on purpose: where they disagree, the disagreement is the finding. Reported as a cross-check — not adopted as the denominator, and not baked into this page.
• SFBM — the WCN channel’s own playlist “San Francisco Bitcoin Meetup” has 74 members (0 unavailable). Of those we hold 40 as media on the drive and 40 as rows in brain.db (40 either way). Two independently-derived counts, printed side by side on purpose: where they disagree, the disagreement is the finding. Reported as a cross-check — not adopted as the denominator, and not baked into this page.

Do not add these to any total

They are already counted inside the rows above. Adding them double-counts.

Nested shows — no source of their own:

• YMB You, Me, and BTC — broadcast on world_crypto_network, identified by title, already inside those rows
• RR Read Rothbard, Use Bitcoin — broadcast on world_crypto_network, identified by title, already inside those rows
• TIB Today in Bitcoin — broadcast on world_crypto_network, mad_bitcoins, identified by title, already inside those rows
• TBG The Bitcoin Group — broadcast on world_crypto_network, mad_bitcoins, identified by title, already inside those rows
• BTS The Bitcoin Talk Show — broadcast on world_crypto_network, mad_bitcoins, identified by title, already inside those rows
• CAB Crypto Audioblog — broadcast on world_crypto_network, identified by title, already inside those rows
• CBC Chris Before Coffee — broadcast on world_crypto_network, identified by title, already inside those rows
• BTCIOT BTCIOT — Bitcoin & the Internet of Things — broadcast on world_crypto_network, identified by title, already inside those rows
• BCB Bitcoin Coffee Break — broadcast on world_crypto_network, identified by title, already inside those rows
• JOE The Joesmoe Show — broadcast on world_crypto_network, identified by title, already inside those rows
• CCV Crypto Convos — broadcast on world_crypto_network, identified by title, already inside those rows

WCN strands — named runs inside WCN, deliberately not promoted to shows, because promoting one changes every WCN figure ever quoted:

• Proof of Work — 103 rows (inside WCN)
• Crypto Audioblog — 39 rows (inside WCN)
• BTCIOT — 36 rows (inside WCN)
• Bitcoin Technical Analysis — 35 rows (inside WCN)
• Chris Before Coffee — 27 rows (inside WCN)
• Bitcoin Coffee Break — 26 rows (inside WCN)
• ProTip — 25 rows (inside WCN)
• The Bitcoin News Show — 24 rows (inside WCN)
• This Week In Cryptos — 24 rows (inside WCN)
• Dark News — 16 rows (inside WCN)
• Crypto Convos — 14 rows (inside WCN)
• MiniBits — 12 rows (inside WCN)
• Off Chain — 10 rows (inside WCN)
• World Crypto Book Club — 10 rows (inside WCN)
• FutureRant — 9 rows (inside WCN)
• The Joesmoe Show — 9 rows (inside WCN)
• Adventures in #NFTs — 8 rows (inside WCN)
• WCN Town Hall Meeting — 8 rows (inside WCN)
• On The Record — 5 rows (inside WCN)