Patrick Tam: 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
90.7%z 4.96 - 2
87.9%z 4.66 - 3
85.3%z 4.45 - 4
83.6%z 4.31 - 5
83.1%z 4.27 - 6
82.3%z 4.21 - 7
76.6%z 3.98 - 8
74.8%z 3.96 - 9
73%z 3.94 - 10
68.3%z 3.89 - 11
67.4%z 3.88 - 12
65.8%z 3.86 - 13
61.8%z 3.82 - 14
60%z 3.80 - 15
59%z 3.79 - 16
57.8%z 3.77 - 17
52.7%z 3.72 - 18
51.3%z 3.70 - 19
50.2%z 3.69 - 20
50.1%z 3.69 - 21
49.8%z 3.69 - 22
47.8%z 3.65 - 23
47.8%z 3.65 - 24
47.3%z 3.62 - 25
46.3%z 3.55 - 26
46.2%z 3.55 - 27
46.2%z 3.55 - 28
45.9%z 3.53 - 29
45%z 3.47 - 30
44.8%z 3.46 - 31
44.8%z 3.46 - 32
44.6%z 3.45 - 33
44.6%z 3.44 - 34
44.5%z 3.44 - 35
43.9%z 3.40 - 36
43.9%z 3.40 - 37
43.7%z 3.39 - 38
43.7%z 3.39 - 39
43.6%z 3.38 - 40
43.5%z 3.37 - 41
43.4%z 3.37 - 42
43%z 3.34 - 43
42.8%z 3.33 - 44
42.8%z 3.33 - 45
42.5%z 3.31 - 46
42.4%z 3.30 - 47
42.3%z 3.30 - 48
42%z 3.28 - 49
41.8%z 3.26 - 50
41.4%z 3.24