[Defago]

I actually meant mathematically. The equation rewards quantity of ratings, mostly when working with items with low ratings. Every item, when calculating its rating, works as if it had 20 ratings worth whatever the site mean rating is (usually around 75-76). This means every item gets 20 "ghost" ratings of 75.

If an item gets five 100/100 ratings, then the average will actually be 80. An item with twenty 85/100 ratings gets the same rating. That's favoring quantity of ratings, and it makes all the well-known big bands jump up to the top of the ratings. I understand the need of having a bayesian average to avoid having items with one rating jump to the top, but a limit of 5 would do the trick and encourage lesser known items to get higher scores on the ratings chart. We already have the ranked chart to see how popular items are.

[EyeKanFly]

Bohemian Rhapsody
Like a Rolling Stone
Tomorrow Never Knows
Karma Police
Paranoid Android
Strawberry Fields Forever
Sympathy for the Devil
Sunday Morning
The Boxer
Bridge Over Troubled Water

Only one of these would be in my overall top 10 (Bohemian Rhapsody).

It was kind of hard picking. There's a lot of stuff I might've thought would be in the 100, but wasn't ("Thunder Road", "Do You Realize??", "I'm Waiting for the Man")
_________________
51 Washington, D.C. albums!