The Elfcyclopedia
A methodological triptych, hosted on ElfSurgery.com.
The Elfcyclopedia is three connected projects on trans-relevant evidence and infrastructure, sharing methodological commitments and a single host. Each panel asks a different question and answers it on its own terms; together they form a working epistemic apparatus that a reader can consult and check.
Panel IWhat Actually Works
A comprehensive research report on gender-affirming care — what the evidence says, what the practice looks like, and where the gaps are. The report does not take a position on whether passing should be a goal. It takes a position on the unconscionability of withholding information from those who have decided it is.
Panel IIElfvaluation
Seventy-six gender-critical and anti-trans claims scored on a fifteen-variable, four-category algebra: Evidence, Method, Consensus, Logic. p(V), the variable-product score, is the geometric mean across the four categories — a zero in any category propagates to the whole. Every claim is traced to its source; every score is backed by citations a reader can check.
The matrix surfaces shape, not just truth or falsehood: two sources can be wrong in entirely different ways, and the composition of a source's scores across the four categories is itself a reading of that source's epistemic posture.
Formerly published as Miserable Bastards; renamed to keep the corpus inside the Elfcyclopedia naming family.
Panel IIIThe Rankings
Every trans-supporting organisation we could find was put to a single question: does a dollar given to this org achieve more good than a dollar given directly to a trans person in need? If yes, the org scores positive. If we cannot tell — because the org doesn't publish its budget, or its service numbers, or both — it scores negative. Not because it's bad. Because a girl with five dollars and a choice to make cannot afford guesswork.
Scores marked verified have both revenue and throughput confirmed from published sources — tax filings, annual reports, impact data. Everything else is a best estimate, with the working assumed and surfaced rather than hidden inside a confident-looking number.