Current definition cat-v1. Charts stamped with it were computed by src/resolution/categories.py, and every value plotted for a past date could have been computed on that date.
Neither venue ships a usable category on the endpoints we pull. Kalshi's series ticker is a category of sorts but only for Kalshi; Polymarket's Gamma market payload carries no tag at all (scripts/polymarket_ingest.py pulls 31 fields and none of them is one). So the category is derived from the question text, here, by keyword.
That is a shallow rule and it is meant to be. Its whole job is to put 132k questions into buckets big enough to compare dispute rates across. Every classification records the term that produced it, so a suspicious bucket can be audited rather than believed — the same standard flags.py holds itself to.
Order matters and is deliberate. A question can match several vocabularies ("Will Trump create Bitcoin reserve in first 100 days?" is both politics and crypto); the first rule in RULES order wins, and classify returns which one it was. Sports goes first because it is 60% of the corpus and its markers ("vs.", "O/U", "Spread:") are unambiguous — leaving it late lets a team nickname get caught by another vocabulary and quietly inflate that bucket.
Newest first. Each entry says what changed and whether values that had already been published moved with it, followed by the sentence from the module that decided the verdict.
eight buckets, keyword and marker rules, first match wins.
The first version of this metric. There is no earlier definition to restate.
3 charts name this metric in their readout line. None of them names it first, so a count taken from the leading stamp alone reports 0 rather than 3. A chart built on several definitions carries all of them, and being second in that list is not the same as being unused.