Tom Gregorio: 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
91.7%z 5.15 - 2
86.3%z 4.53 - 3
84%z 4.34 - 4
83.2%z 4.28 - 5
74.2%z 3.95 - 6
72.6%z 3.94 - 7
68.2%z 3.89 - 8
65%z 3.85 - 9
59.6%z 3.79 - 10
53.8%z 3.73 - 11
53%z 3.72 - 12
47.8%z 3.65 - 13
46.8%z 3.59 - 14
46.2%z 3.55 - 15
46%z 3.54 - 16
45.7%z 3.52 - 17
44.6%z 3.45 - 18
44.4%z 3.43 - 19
43.9%z 3.40 - 20
43.3%z 3.36 - 21
43.2%z 3.35 - 22
43.1%z 3.35 - 23
43%z 3.34 - 24
43%z 3.34 - 25
42.8%z 3.33 - 26
42.8%z 3.33 - 27
42.7%z 3.32 - 28
42.5%z 3.31 - 29
42.3%z 3.30 - 30
41.8%z 3.27 - 31
41.7%z 3.26 - 32
41.3%z 3.23 - 33
40.8%z 3.20 - 34
40.7%z 3.19 - 35
40.3%z 3.17 - 36
40.3%z 3.17 - 37
40.1%z 3.16 - 38
40.1%z 3.15 - 39
40%z 3.15 - 40
39.9%z 3.14 - 41
39.5%z 3.12 - 42
39.5%z 3.11 - 43
39.4%z 3.11 - 44
39.4%z 3.11 - 45
39.4%z 3.11 - 46
39.2%z 3.09 - 47
38.9%z 3.08 - 48
38.7%z 3.06 - 49
38.5%z 3.05 - 50
38.1%z 3.03