Mark Flekken: raw computed results
The raw ranking from the production similarity calculation (face-embedding centroid cosine → gender-cohort z-normalization). The web-opinion crown, curation, and fusion re-rank are removed.
- 1
82.2%z 4.19 - 2
81.7%z 4.16 - 3
81.2%z 4.12 - 4
69.2%z 3.90 - 5
68.2%z 3.89 - 6
59.1%z 3.79 - 7
47.7%z 3.65 - 8
46.2%z 3.55 - 9
46.1%z 3.54 - 10
45.9%z 3.53 - 11
45.4%z 3.50 - 12
45.2%z 3.49 - 13
44.7%z 3.45 - 14
43%z 3.34 - 15
42.9%z 3.33 - 16
42.6%z 3.32 - 17
42.4%z 3.30 - 18
42.1%z 3.29 - 19
41.8%z 3.26 - 20
41.7%z 3.26 - 21
41.5%z 3.24 - 22
41.2%z 3.23 - 23
41%z 3.22 - 24
40.7%z 3.19 - 25
40.5%z 3.18 - 26
40.4%z 3.18 - 27
40.4%z 3.17 - 28
40.3%z 3.17 - 29
40.2%z 3.16 - 30
40.2%z 3.16 - 31
40%z 3.15 - 32
40%z 3.15 - 33
39.9%z 3.14 - 34
39.9%z 3.14 - 35
39.4%z 3.11 - 36
39.4%z 3.11 - 37
39.2%z 3.10 - 38
39.2%z 3.10 - 39
39.2%z 3.10 - 40
39.1%z 3.09 - 41
39.1%z 3.09 - 42
39%z 3.08 - 43
38.9%z 3.08 - 44
38.9%z 3.08 - 45
38.6%z 3.06 - 46
38.6%z 3.06 - 47
38.6%z 3.06 - 48
38.6%z 3.06 - 49
38.4%z 3.04