Evidence Grade
Score a case the way the desk does. Witness quality, sensor data, documentary record, and remaining gaps are weighted separately so a vivid story cannot outrun a thin file.
Inputs
0 none · 3 trained multi-witness · 5 independent experts, contemporaneous.
Radar, video, telemetry, medical, physical trace. 0 = testimony only.
Official reports, timestamps, chain of custody, contemporaneous notes.
0 closed file · 5 decisive missing pieces. Gaps subtract.
Conventional hypothesis
Result
A
Publishable as documented, high confidence
- Score / 100
- 100
- Confidence
- High
- Sensor share
- 34%
- Gap penalty
- −20
12·3 + 22·2 + 16·3 − 10·2 + 0 = 100
Methodology
Four piles, kept separate. Testimony can raise a case; it cannot carry one. Sensor and documentary weight more than witnesses. Gaps and a live conventional explanation both pull the grade down. This is an editorial rubric, not a Bayesian posterior.
score = clamp(12·W + 22·S + 16·D − 10·G + mod, 0, 100) mod = −18 strong conventional | 0 open | +8 conventional fails A ≥ 85 · B ≥ 70 · C ≥ 55 · D ≥ 40 · else F
Limits A high grade means the file is strong enough to publish as documented. It does not mean “aliens,” “hoax,” or any other conclusion. Interpretation stays in the article. The tool only scores the record.