Mark Parrish: 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
83.6%z 4.31 - 2
82.1%z 4.19 - 3
81.6%z 4.14 - 4
81.3%z 4.12 - 5
69.9%z 3.91 - 6
65.7%z 3.86 - 7
60.2%z 3.80 - 8
58.3%z 3.78 - 9
57.9%z 3.78 - 10
53.9%z 3.73 - 11
47.5%z 3.64 - 12
47.3%z 3.62 - 13
46.6%z 3.58 - 14
46.4%z 3.56 - 15
46.2%z 3.55 - 16
46.1%z 3.54 - 17
44.2%z 3.42 - 18
43.8%z 3.39 - 19
43.5%z 3.38 - 20
43.3%z 3.36 - 21
42.9%z 3.33 - 22
42.5%z 3.31 - 23
42.2%z 3.29 - 24
41.8%z 3.26 - 25
41.6%z 3.25 - 26
41.1%z 3.22 - 27
41%z 3.21 - 28
40.3%z 3.17 - 29
39.8%z 3.14 - 30
39.7%z 3.13 - 31
38.4%z 3.05 - 32
38.3%z 3.04 - 33
38.1%z 3.03 - 34
38.1%z 3.02 - 35
37.7%z 3.00 - 36
37.7%z 3.00 - 37
37.5%z 2.99 - 38
37.5%z 2.99 - 39
37.5%z 2.99 - 40
37.5%z 2.99 - 41
37.4%z 2.98 - 42
37.4%z 2.98 - 43
37.3%z 2.97 - 44
37.1%z 2.96 - 45
37.1%z 2.96 - 46
37.1%z 2.96 - 47
37%z 2.95 - 48
36.9%z 2.95 - 49
36.6%z 2.93 - 50
36.5%z 2.92