Marco Haber: 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
84.5%z 4.38 - 2
80.5%z 4.06 - 3
45%z 3.47 - 4
43.2%z 3.36 - 5
42.9%z 3.33 - 6
42.6%z 3.32 - 7
42.1%z 3.29 - 8
41.3%z 3.23 - 9
41.2%z 3.23 - 10
41%z 3.21 - 11
40.9%z 3.21 - 12
40.7%z 3.20 - 13
40%z 3.15 - 14
39.2%z 3.09 - 15
39.1%z 3.09 - 16
38.7%z 3.07 - 17
38.7%z 3.07 - 18
38.5%z 3.05 - 19
38.4%z 3.05 - 20
38.3%z 3.04 - 21
38.1%z 3.03 - 22
37.7%z 3.00 - 23
37.6%z 3.00 - 24
37.5%z 2.99 - 25
37.3%z 2.97 - 26
36.4%z 2.92 - 27
36.2%z 2.90 - 28
35.9%z 2.88 - 29
35.8%z 2.88 - 30
35.8%z 2.87 - 31
35.7%z 2.87 - 32
35.7%z 2.87 - 33
35.7%z 2.87 - 34
35.2%z 2.83 - 35
34.8%z 2.81 - 36
34.8%z 2.81 - 37
34.4%z 2.79 - 38
34.4%z 2.79 - 39
34.3%z 2.78 - 40
34%z 2.76 - 41
33.9%z 2.75 - 42
33.8%z 2.74 - 43
33.7%z 2.74 - 44
33.3%z 2.71 - 45
33.1%z 2.70 - 46
33.1%z 2.70 - 47
33%z 2.70 - 48
33%z 2.70 - 49
33%z 2.69 - 50
32.8%z 2.68