← srd 5.2.1

SRD 5.2.1, parsed

The second system on this surface, and the first one with positions. Cairn has no grid and no facing, which is why its arena refuses to draw a map. This game is nothing without one — cover, reach, opportunity attacks and cone templates are all distances — so the map here will be the model rather than decoration. Everything starts with getting the rules out of a 364-page PDF and proving they came out intact.

What came out of the document

Recovered from the official PDF by one scraper, and checked by 67 assertions before anything here is believed. That ratio is deliberate: Cairn's SRD is HTML with real table markup, so a bad parse there tends to throw. This one is typeset text in two justified columns, and its failures come back looking like English.

One thing the document itself gets wrong

The Archmage's XP does not match its Challenge Rating. SRD 5.2.1 prints CR 12 (XP 8,000). Challenge Rating 12 is worth 8,400 XP, and the Erinyes — also CR 12, a few pages away — prints 8,400 correctly. So this is an erratum in the published document, not a bad parse.

The data keeps 8,000, because it says what the book says. Quietly correcting it would make this corpus disagree with the page someone is reading at the table, and that is a worse failure than being wrong in the same way the source is. The selftest pins it as the only known discrepancy, so if a later revision fixes it — or introduces another — the tests say so instead of the exception rotting.

Worth saying plainly: this was not found by reading. It was found because the parse is graded against the official Challenge-Rating-to-XP table, which is knowledge from outside the parse — the same trick that caught seven swarms being welded onto the creatures above them, and four dragons whose CR line is printed back-to-front as CR 3 (700 XP).

The bestiary

The official encounter budget

This table is the reason this system was chosen. Cairn publishes no challenge rating at all, so its oracle had to invent a metric and label it as ours. Here there is an official answer — and it can be graded. Build encounters to exactly these budgets, simulate them on a real grid, and see whether "moderate" means anything.

Not built yet