Post ARvqR6fHk7ZxtpsyPY by halcy@icosahedron.website
 (DIR) More posts by halcy@icosahedron.website
 (DIR) Post #ARvqR5dTZMXOhw20x6 by halcy@icosahedron.website
       2023-01-23T10:18:47Z
       
       0 likes, 0 repeats
       
       ASSUMPTION: If you have, for a number of models, a large number of files per model, say, 139, each rated by 5 raters, then it is vanishingly unlikely to get the exact same mean score twiceREALITY: I just had a batch with 14 models where there are two pairs with completely identical mean scores but different distributions, and one with a difference to the order of 10^-6
       
 (DIR) Post #ARvqR6ARamuQMBSLL6 by halcy@icosahedron.website
       2023-01-23T10:18:58Z
       
       0 likes, 0 repeats
       
       maybe I should play he lottery this week
       
 (DIR) Post #ARvqR6fHk7ZxtpsyPY by halcy@icosahedron.website
       2023-01-23T10:20:15Z
       
       0 likes, 0 repeats
       
       central limit theorem (bit-exact version)
       
 (DIR) Post #ARvqR7ApqoofTgeAaW by marsxyz@pleroma.soykaf.com
       2023-01-23T10:35:29.319650Z
       
       0 likes, 0 repeats
       
       @halcy central limit theorem any %