How to read a match percentage
Every pair of works on this site carries one number for how much they share. This is what that number is made of, and — the part no page said — what a given value is actually worth against the 51,040 pairs the index holds.
What a number is worth
The median pair of works in this index matches at 59%, which is the figure to hold everything else against. Above that the air thins fast: 7% of pairs reach 80, 0.5% reach 90, and exactly 21 of 51,040 pairs in the whole index get past 95. A number in the nineties on this site is rarer than it sounds.
| A match of | Beats this share of pairs | Pairs above it |
|---|---|---|
| 50% | 30.9% | 35,290 |
| 60% | 54.6% | 23,193 |
| 70% | 77.5% | 11,474 |
| 80% | 93.3% | 3,396 |
| 85% | 97.5% | 1,280 |
| 90% | 99.5% | 280 |
| 95% | 100.0% | 21 |
What it is made of
Distance between two works on the five axes that measure fear, weighted, then turned into a percentage. The weights: Pace of Fear ×1.15, The Threat ×1, Viscera ×1.1, Tone ×1, Reality Anchor ×0.95. Pace of Fear counts for most and Reality Anchor for least, because a disagreement about how a work frightens you matters more than one about how grounded it is.
The percentage is bounded at 99, not 100: two different works are never the same work, and a scale that can print 100% invites reading it as identity.
Why era bias is in your tape and not in a pair
Comparing two works leaves Era Bias out. It measures texture rather than fear, so two works sharing it share a look, not a feeling, and saying otherwise would be a claim this site does not stand behind — see what each score looks like.
Comparing your tape to a work keeps all six, because there era is not an observation about a work: it is a dial you set. If you asked for analog texture, a recommendation that ignores it is not answering you.
Compare any two works → · Both ends of an adaptation, measured → · What the whole index says → · How the scoring works →
Questions about match percentages
What does a VHSCREAM match percentage mean?
How close two works sit across the five axes that measure fear, weighted and turned into a percentage. The only way to read one is against the distribution, which is what this page is for: the median pair in the index matches at 59%, so that is the figure every other number should be held against.
Is a 90% match high?
Much higher than it sounds. Only 0.5% of the 51,040 pairs in this index reach 90, and 21 get past 95. The air thins fast above the median, which is exactly why a percentage with nothing to compare it to is not an answer.
Why does the match stop at 99 instead of 100?
Because two different works are never the same work, and a scale that can print 100% invites being read as identity rather than as closeness. The bound is deliberate: it is the difference between saying two films frighten you the same way and saying they are the same film.
Why is era bias left out of a match between two works?
Because it measures texture rather than fear, so two works sharing it share a look and not a feeling. Comparing your own tape to a work keeps all six axes, because there era is not an observation about a work — it is a dial you set, and a recommendation that ignores what you asked for is not answering you.
Which axes count for most in a match?
Pace of Fear counts for most and Reality Anchor for least. A disagreement about how a work frightens you matters more than one about how grounded it is, and the full set of weights is printed on this page rather than described, so the figure can be recomputed from the published dataset.