The ATEM
chronicles

Four kinds of data, each with its own way in

The study of role-playing games has tended to work ontologically — asking what role-play is, what defines a campaign, what makes a mega campaign. This page works the other way: instead of defining essences, it observes a practice and reconstructs it from what it leaves. A campaign, taken as a historical event of shared imaginative projection, leaves a heterogeneous material and memorial record; sorting that record is the precondition of any claim about what happened at the table.

The four kinds rest on two conditions: the context of the practice — who played, when, where, how often — and the rules put into play, which are the device through which the world is projected. That is why each table's system appears on its card, ratified by the game master chronicle by chronicle.

The Mega Campaign has two clocks. In the time of play its chronicles follow one another; in the time of the world they run out of order — a chronicle played in 2019 can take place in 1300. That bi-temporality is structural, and the instrument below makes it measurable: The world, The table and The leap are three orderings of the same 30 cards.

From the table to the page

Thirty-five years, one game master, partially persistent cores of players. The Archive has three beginnings, not one: the practice itself from about 1990–91; the workshop and Torregrises from 1996; the sound archive from 2007. The following pipeline describes the work; individual sources and editions remain at different stages.

1. Play context · rules

   The session itself. Its context (players, dates, places, frequency) and its system are recorded with it, because both condition everything the table produces.

   Yields: the event — everything below is a trace of it.

2. Write notes

   Before, during and after play the game master writes by hand: 123 notebooks, coded C1 onward, indexed by period. They are the oldest layer of the Archive and they prevail over every later one.

   Yields: the manuscript core.

3. Record recordings

   From 2007, and systematically from February 2013, sessions are recorded in audio and, for some tables, video.

   Yields: 437 recordings in the register.

4. Transcribe notebooks · audio

   Notebooks are scanned and transcribed page by page; a transcription may expand abbreviations but may not add names. Audio is transcribed automatically. Every draft entry keeps its coordinates — notebook and page, such as C38 p43 — so that it can always be checked against the original.

   Yields: traceable drafts and raw session transcripts.

5. Consolidate one account per session

   Raw transcripts become a clean, dated, readable account of each session: dice, rules talk and table chatter are removed, duplicated transcription is removed, and nothing is invented — confused passages stay marked as unclear, doubtful names as uncertain. Consolidation is reconstruction, faithful but not verbatim; what belongs to the real table rather than to the fiction is kept apart, in the [R] layer.

   Yields: the consolidated chronicle of each campaign.

6. Connect and curate canon

   Notebook pages and consolidated sessions feed the glossary: one entry per person, place, object or event, written in Obsidian and published with Quartz. Every link must resolve; no real player names appear. The Spanish entries keep the voices of the Archive's chroniclers; the English edition is a fact-first reference — one article per Spanish entry, the same address under /en/, disputes kept under Conflicting accounts and gaps under Open questions.

   Yields: the glossary in Spanish and English, with a concordance that joins them.

7. Retell decks · comics

   The cards above give each chronicle a short English introduction, linked to its glossary sources. The original illustrated decks remain available in Spanish. Comics offer a longer retelling where an edition exists. Both are derived works: they rank below the Archive.

   Yields: 29 decks · 17 English comics so far.

8. Return to the register the visualizer

   The table itself, recording by recording: full text, audio narration and, where the Archive keeps it, the session recording. It sits behind a password because the real names of the people who sat at each table appear there, and those do not go into the open.

   Yields: the way back from any retelling to its source.

When layers disagree, the older one prevails — the hierarchy of sources:

1. The notebooks — the handwritten originals.
2. The transcription — the first copy of the notebooks.
3. The Calendar — the chronological consolidation of the shared cosmos.
4. The glossary — the curated entries.
5. The marginal gloss — commentary, not a source.

Decks and comics come after all of them. A later layer may overwrite an earlier one only if it says so.

An archive of play is also subject to what Schiffer called post-depositional processes, and this one measures its own: selective mining (an archive designed to index characters makes the device of play invisible), copies multiplying with none marked authoritative, notebook datings drifting between versions of the index, changes of method left undeclared. The Archive keeps several roots, and every audit inherits its root: “in which copy?” comes before any claim, as in textual criticism.