Brad Gilbert: 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
83.1%z 4.27 - 2
82.5%z 4.22 - 3
73.5%z 3.95 - 4
72%z 3.93 - 5
66.4%z 3.87 - 6
64.9%z 3.85 - 7
62.2%z 3.82 - 8
62.1%z 3.82 - 9
49.2%z 3.68 - 10
47.9%z 3.66 - 11
47.6%z 3.64 - 12
44.3%z 3.43 - 13
44.3%z 3.42 - 14
43%z 3.34 - 15
42.8%z 3.33 - 16
42.7%z 3.33 - 17
42.6%z 3.32 - 18
42.5%z 3.31 - 19
42.3%z 3.30 - 20
42.3%z 3.30 - 21
41.8%z 3.27 - 22
41.5%z 3.25 - 23
41.4%z 3.24 - 24
41.3%z 3.23 - 25
40.7%z 3.19 - 26
40.6%z 3.19 - 27
40.6%z 3.18 - 28
39.7%z 3.13 - 29
39.6%z 3.12 - 30
39.4%z 3.11 - 31
39.4%z 3.11 - 32
39.2%z 3.10 - 33
39.1%z 3.09 - 34
39.1%z 3.09 - 35
39%z 3.08 - 36
38.9%z 3.07 - 37
38.7%z 3.07 - 38
38.6%z 3.06 - 39
38.5%z 3.05 - 40
38.5%z 3.05 - 41
38.2%z 3.03 - 42
38%z 3.02 - 43
38%z 3.02 - 44
37.9%z 3.01 - 45
37.8%z 3.01 - 46
37.6%z 2.99 - 47
37.6%z 2.99 - 48
37.4%z 2.98 - 49
37.3%z 2.97 - 50
36.9%z 2.95