-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Matthew Paul Thomas wrote on 24/06/11 16:17: >... > So it seems to me that one test case for our final formula should be > that if something has no ratings at all, *or* if it has only 3-star > ratings, its ELR should be 3.00. This means it's not really an > "Expected Lowest Rating" (ELR), but more like a "Dampened Rating" > (DR), modelling the pattern that the average rating usually dampens > towards 3.00 as more ratings come in. > > So I tweaked your Python code to subtract 3 from each star rating before > multiplying it by the corresponding Wilson score, and then to add 3 to > the final result. >...
I've now updated the "Top Rated" specification to reflect this, and added a sketch. <https://wiki.ubuntu.com/SoftwareCenter#top-rated> - -- mpt -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.11 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAk4VqGUACgkQ6PUxNfU6ecoGGwCggrkCo4f2Q6n0yYSXyveinRp1 wVsAn3EQagYOShH3A4MAb3d1GqBecgOt =+Kyg -----END PGP SIGNATURE----- _______________________________________________ Mailing list: https://launchpad.net/~software-store-developers Post to : [email protected] Unsubscribe : https://launchpad.net/~software-store-developers More help : https://help.launchpad.net/ListHelp
