Feat-first by default
Profiles should prioritize concrete, reconstructable feats. Statements and narrative implications can exist, but they should be labeled and weighted separately.
Methodology Sheet
How The Archive Thinks
This page explains the ranking philosophy so users understand why profiles, votes and staff actions are structured the way they are.
Category structure is fixed, entries are flexible
The archive uses a stable category and subcategory tree so future imports from external feat-sorting workflows stay consistent.
Classification is methodological
Realistic, semi-realistic and unrealistic are not insults. They indicate the style and plausibility level of the intelligence portrayal being evaluated.
Admin overrides are visible
Because public votes can be biased or manipulated, staff must be allowed to change rankings with explicit written reasoning.
Base score and live score are different
Each character starts from a dossier-based base score on a 10-point scale. Public voting then nudges that score instead of replacing it, so popularity can move a profile without erasing the original evidence package.
Rank is contextual, not decimal
A rank number is not converted into a score pattern. Rank is produced after judging dossier quality in the current archive context, which means a score and a rank can be strong without mirroring each other mathematically.
AI comparisons are dossier readers
Battle outputs should read approved feats across both dossiers, compare category by category, explain each verdict and then synthesize an overall winner with a stated difficulty level. They are tools for analysis, not eternal verdict machines.
Community movement is damped, not explosive
Vote effects should scale gradually. A small burst of votes can show momentum, but large movement only happens when a wider voting pattern keeps reinforcing the same direction over time.
Ranking Logic
How Scores Move
The archive uses a two-layer ranking model so evidence remains primary while community activity still matters.
1. Base score
The starting score is assigned from the dossier itself: feat quality, breadth, traceability, consistency and resistance to weak interpretation.
2. Community pressure
Votes create a damped adjustment that grows with volume and agreement. Small bursts move less; sustained consensus moves more.
3. Final ordering
Characters are ordered by the live score, then refined by vote distribution and deterministic tie-breaks so rank collisions do not remain unresolved.