O problema e' que a formula do beeradvocate funciona bem pq eles tem muitas avaliacoes e muito, muito, mais usuarios que a gente. Eles normalizam as notas usando a media geral e o numero de avaliacoes, de tal maneira que mesmo que a cerveja tenha notas altissimas dadas por poucas pessoas ela ainda esta rankeada abaixo de uma com notas um pouco mais baixas dadas por muitas pessoas. Segundo eles, e' preciso que para ser devidamente rankeada nas top beers a cerveja tem que ter pelo menos 108 avaliacoes escritas, o que ainda e' uma utopia aqui. Mas a gente chega-la um dia!
Deem uma olhada:
This Best of BeerAdvocate (BA) list is generated using statisical formulas that pull data from millions of user reviews; they are not hand-picked. The general formula uses a Bayesian estimate:
weighted rank (WR) = (v ÷ (v+m)) × R + (m ÷ (v+m)) × C
where:
R = review average for the beer
v = number of reviews for the beer
m = minimum reviews required to be listed (currently 108)
C = the mean across the list (currently 3.76)
The formula normalizes scores, that is pulls (R) to the mean (C) if the number of reviews is not well above (m). So if a beer has only a few reviews above (m), its (WR) is decreased a little if it is above the mean (C), or increased a little if it is below the mean (C) in accordance with the normal distribution rule of statistics.
Currently, a beer must have 108 or > reviews to be included in any calculations. And (m) is calculated by averaging the number of reviews for beers that have 108 or > reviews within the list being viewed, while (C) is the mean (average) overall score for all beers that have or > reviews within the list.
Example 1: (a beer with a 4.35 review average and 105 reviews)
(105 ÷ (105+108)) × 4.35 + (108 ÷ (105+108)) × 3.76 = 4.05 = WR
Example 2: (a beer with a 3.1 review average and 11 reviews)
(11 ÷ (11+108)) × 3.1 + (108 ÷ (11+108)) × 3.76 = 3.7 = WR