Matthew Sweet: 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
84.1%z 4.36 - 2
78.9%z 4.00 - 3
73.4%z 3.94 - 4
61.6%z 3.81 - 5
54.8%z 3.74 - 6
49.4%z 3.68 - 7
48.3%z 3.67 - 8
46.3%z 3.56 - 9
45.8%z 3.52 - 10
45.5%z 3.50 - 11
45%z 3.48 - 12
44.8%z 3.46 - 13
44.4%z 3.43 - 14
43.9%z 3.40 - 15
43.6%z 3.38 - 16
43.5%z 3.38 - 17
43.5%z 3.37 - 18
43.4%z 3.37 - 19
43.3%z 3.36 - 20
43.2%z 3.36 - 21
42.9%z 3.34 - 22
42.6%z 3.31 - 23
42.4%z 3.30 - 24
42.3%z 3.30 - 25
42.3%z 3.29 - 26
42.2%z 3.29 - 27
42.2%z 3.29 - 28
41.6%z 3.25 - 29
41.6%z 3.25 - 30
41.6%z 3.25 - 31
41.2%z 3.22 - 32
41.1%z 3.22 - 33
40.7%z 3.19 - 34
40.6%z 3.19 - 35
40.5%z 3.18 - 36
40.5%z 3.18 - 37
40.4%z 3.18 - 38
40.1%z 3.15 - 39
40.1%z 3.15 - 40
39.9%z 3.14 - 41
39.8%z 3.14 - 42
39.8%z 3.13 - 43
39.4%z 3.11 - 44
39.2%z 3.10 - 45
39.2%z 3.09 - 46
39%z 3.08 - 47
38.7%z 3.07 - 48
38.7%z 3.07 - 49
38.5%z 3.05 - 50
38.2%z 3.03