DATA METHODOLOGY
How a count becomes a question.
Your browser never contacts the adult-content data source. Counts first enter internal snapshots, pass review, and are then frozen into questions.
Updated: 2026-07-15Collection
The prototype queries aggregate post totals for one or two approved tags through the Rule34 authenticated DAPI. It requests the smallest result set needed to read the total and does not store, display, or relay source posts.
AI-exclusion scope
The condition labelled “Exclude tagged AI-generated” subtracts posts carrying the source tag “ai_generated” by using the reviewed negative tag “-ai_generated”. It does not claim to detect untagged AI content, mislabelled posts, or every AI-related category.
Review and freezing
- Every character needs traceable evidence of adult status and manual approval.
- Every condition must come from the internal allowlist.
- Each count snapshot must pass validity review.
- Ties, counts below the condition threshold, and gaps under roughly 7% are not published.
- Difficulty is calculated from the log10 gap between the two counts.
Timing and possible differences
The page shows a snapshot time, not a live view of a third-party site. Deleted posts, tag merges, and indexing delays can create differences. Use the report action beside a round if something looks wrong.
Failure fallback
If collection fails, the game continues to serve the most recent approved frozen rounds. A failed response is never written to the database as a zero count.