Ross Ulbricht goes camping for a weekend with a girlfriend and another couple. The Dread Pirate Roberts side of his laptop posts: I'm going away for the weekend. Do X, Y, and Z, and I'll be back. No geolocation on the campground photo. The other photos: one of Ross and the girl by a car on a street in San Francisco. One past the Golden Gate Bridge, someone's hand visible. One at the campground, no GPS tag.
Bilton wants this chapter. So he opens Google Maps. Time between the street-photo and the Golden Gate Bridge photo: known from the EXIF timestamps. Time from Golden Gate Bridge to campground photo: 48 minutes. He maps every wooded area within a 48-minute drive of the Golden Gate Bridge. He walks them. He finds the campsite. He sits where they sat. He writes the chapter.1
This is the Tetris database method in its purest form — what makes Bilton's reporting craft singular and what most narrative non-fiction reporting doesn't even attempt. The method has four pieces and one composing principle.
Piece 1 — Substantive corpus. For Ulbricht: the chat logs from the Dread Pirate Roberts side of his laptop, 2.5 years of conversations with employees, every line timestamped. Most stories don't have something this rich. But every modern story has some corpus — emails leaked in discovery, court documents, social media archives, deleted Reddit posts on Wayback Machine. The first move is finding and ingesting the substantive corpus.
Piece 2 — Human-interview database. Bilton interviews everyone — Ulbricht's roommate, girlfriend, mom (who agreed only on the courtroom steps and then changed her mind), college and high school friends. Every interview goes into the database with timestamps and locations and who-was-where-when. The mom-on-the-courtroom-steps moment is a footnote in the book; the rest of the database is the substrate that lets him verify timeline.
Piece 3 — Social media exhaust. Tweets. Facebook posts. Venmo transactions. The geolocated photos. The check-ins. This is what the FBI is also using, and Bilton is essentially doing parallel investigation. Everything goes into the database.
Piece 4 — Physical fieldwork. Once the database tells him where Ulbricht was, Bilton goes there. He sits in the Glen Park library where Ulbricht ran Silk Road. He goes to the same sushi place and orders what Ulbricht ordered. He sits in the same seat. He notices the wooden chopsticks and the bathroom on the left. He makes notes of smells — pine needles, crickets, sky. (Smell is what most writers leave out.1)
Bilton calls the composing logic Tetris.1 Each interview, document, social-media post, and location-visit is a Tetris piece. He's not reading them sequentially; he's looking for fits. Where does this piece slot in? Where does this one? Where's a row he can clear by stacking the timeline-matched pieces in sequence? Where's a piece that doesn't fit yet — set it aside for now, come back to it.
This is fundamentally different from how most narrative non-fiction is reported. Most reporters interview, write up notes, organize chronologically, draft chapters. Bilton's method is non-linear: he assembles a database first, then queries it for chapter-shaped subsets, then physically validates the subsets through fieldwork. The chapter on Ulbricht's camping trip wasn't planned — it emerged from the database query "what was Ulbricht doing the weekend the Silk Road had its crazy moment?" and then resolved through the 48-minute-radius walk.
Scenes the source can't remember and won't tell you. The whole point of the Sorkin painting-the-room technique (mahogany table, chocolate croissants) is to extract scene-detail from sources who can't or won't give substantive content. The Tetris method removes the source entirely. The database tells you what was there.
Verifiable detail at the scene level. Bilton can say Ulbricht sat in this seat at this sushi place because he sat in that seat at that sushi place and verified the menu was the same. The detail isn't speculative; it's witnessed retrospectively.
Counterfactual reasoning at the time-coordinate level. If you know everyone's location every day, you can reason about who could have known what and when. The Tetris database makes Ulbricht's social network legible at a granularity that conventional reporting never reaches.
This works for stories with digital trails — tech, crime, business, contemporary politics. It works less well for stories without trails — immigrant labor, undocumented life, pre-2000s subjects, anyone who lives outside digital infrastructure. Bilton's method is a child of the platform era. A 1985 version of American Kingpin would not be possible to write with the Tetris technique because the pieces don't exist.
It also works less well for stories where the substantive corpus is sealed or paywalled — diplomatic cables not yet leaked, corporate communications that never become court exhibits, family conversations that leave no digital exhaust. The method is asymmetrically powerful where digital trails are dense.
Bilton and MacCulloch's sense-of-place methodology converge on the principle that physical visiting matters — the writer must sit where the subject sat, walk where the subject walked. MacCulloch goes to Suffolk parish churches and to St. Basil's in Moscow. Bilton goes to Glen Park library and Ulbricht's sushi place. The two practitioners use the same physical-presence discipline.
They diverge on what produces the warrant to visit. MacCulloch goes to places because the historian's instinct says place-experience changes interpretation (the St. Basil's plan-vs-experience principle). Bilton goes to places because the database told him the subject was there. MacCulloch's place-visits expand interpretive possibility; Bilton's place-visits validate database-derived locations. Two complementary applications of the same physical-fieldwork discipline.
Bilton diverges from Orlean's three-phase research-think-write method on whether you can think while reporting. Orlean's discipline is research → think → write as sequential phases. Bilton's Tetris method is research-and-think-together: as he ingests each piece into the database, he's already composing the chapter-fits in his head. The thinking doesn't wait for a separate phase; it happens piece-by-piece. The split reveals that different narrative-nonfiction methodologies have different relationships to the question of when synthesis begins.
The Tetris database method maps onto two adjacent domains where the same investigative-reconstruction logic appears under different names: behavioral-mechanics (the intelligence analyst building a target's pattern-of-life from heterogeneous data sources) and history (the historian working from fragmentary primary-source corpora to reconstruct events whose participants have left no continuous account).
Behavioral-Mechanics: Pattern-of-Life Construction — intelligence analysts (military, law enforcement, corporate security) routinely build target profiles by ingesting heterogeneous data streams (signal intercepts, financial transactions, location data, social media, human-source interviews) into a unified timeline-and-relationship database. The analyst then queries the database for behavioral patterns and operational vulnerabilities. Bilton's Tetris method is the journalist's parallel construction of the same workflow. The structural identity: both crafts assemble fragments into a navigable model of a subject's life and use the model to generate previously-inaccessible knowledge. The insight neither domain generates alone: investigative journalism in the platform era has converged on intelligence-analyst methodology because they face the same data-environment — sources who won't speak, partial corpora, dense digital exhaust, the need for scene-level granularity. Bilton's reconstruction of the Ulbricht camping trip is structurally identical to a counterterrorism analyst reconstructing a target's movement to a meeting. The journalist's product is a chapter; the analyst's product is a target package. The craft is the same.
History: Secondary Sources Before Primary — Cliche Mapping (MacCulloch) — the historian working from a fragmentary corpus (Anglo-Saxon Christianity from Bede plus archaeology plus letters; Tudor England from Cromwell papers plus state records plus Hilary Mantel novels) is doing structurally the same database-assembly that Bilton does with chat logs and Venmo records. The Tetris method is platform-era journalism's rediscovery of the historian's discipline. The structural identity: both crafts reconstruct lives whose owners have left only partial trails by composing the trails into a unified model and querying the model for scene-level reconstruction. The insight neither domain generates alone: the journalism-vs-history disciplinary boundary is fictional — they are the same craft applied to different time-horizons. Bilton is a historian of the 2010s using digital primary sources; MacCulloch is a journalist of the 1530s using paper primary sources. The 500-year gap is the only difference.
A third connection worth marking: the Tetris piece-by-piece composing logic has structural similarity to Macfarlane's mosaic-from-set-pieces writing — both build long-form work non-linearly from validated sub-units. The reporter and the literary nature-writer have converged on the same composing architecture from opposite ends.
The Sharpest Implication
If the Tetris method is what made American Kingpin possible at the level of scene-density Bilton achieves, and if the method requires dense digital exhaust to work, then the future of narrative non-fiction is structurally biased toward subjects who lived inside the surveillance economy. Stories about people who left no digital trail are getting harder to tell at Bilton's level of granularity. Stories about people who left maximum digital trail (anyone born after 2000) are getting easier. The uncomfortable implication: the people whose lives become reportable at scene-level are disproportionately those who participated in platforms — which biases the future canon of narrative non-fiction toward extractable subjects. The off-grid, the undocumented, the contemplatively private — they become harder to write about as the platform-trail subjects become easier. The narrative-non-fiction reader of 2050 will know everything about the platform-lives and almost nothing about the lives lived outside them.
Generative Questions