Brad Wilk: 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
90.4%z 4.91 - 2
87%z 4.59 - 3
83.9%z 4.34 - 4
83.3%z 4.29 - 5
82.9%z 4.26 - 6
81.7%z 4.16 - 7
81.5%z 4.14 - 8
73.8%z 3.95 - 9
65.8%z 3.86 - 10
54.1%z 3.73 - 11
49.1%z 3.68 - 12
46.4%z 3.56 - 13
46.2%z 3.55 - 14
45.6%z 3.51 - 15
45.5%z 3.50 - 16
43.7%z 3.39 - 17
43.5%z 3.37 - 18
43.3%z 3.36 - 19
42.9%z 3.34 - 20
42.8%z 3.33 - 21
41.9%z 3.27 - 22
41.8%z 3.26 - 23
41.3%z 3.23 - 24
40.9%z 3.21 - 25
40.7%z 3.19 - 26
40.7%z 3.19 - 27
40.7%z 3.19 - 28
40.7%z 3.19 - 29
40.6%z 3.19 - 30
40.4%z 3.17 - 31
40.2%z 3.16 - 32
40.2%z 3.16 - 33
39.9%z 3.14 - 34
39.8%z 3.14 - 35
39.8%z 3.13 - 36
39.8%z 3.13 - 37
39.7%z 3.13 - 38
39.7%z 3.13 - 39
39.6%z 3.12 - 40
39.5%z 3.12 - 41
39.4%z 3.11 - 42
39.3%z 3.10 - 43
39.3%z 3.10 - 44
39.2%z 3.10 - 45
39.1%z 3.09 - 46
39%z 3.08 - 47
38.9%z 3.08 - 48
38.9%z 3.08 - 49
38.8%z 3.07 - 50
38.6%z 3.06