Henrik, looks to me like some publications realize lists are a great marketing tool and pile up a huge number of extremely similar lists in their own name, while others don't repeat themselves incessantly. Do you have some system to devalue the lists of publications that go trigger-happy making them?
Yes, I do. If a magazine publishes more than one list within a ten-year span, and the same album appears in both/all lists, these lists get a lower weight.
I have a question about how your statistical system works. What do you do about outliers and skew?
Like, if some publication came out tomorrow and said "Pet Sounds is the worst album ever! 0/10!" Would it get penalized, or do you do something to correct for outliers? I ask because I notice on metacritic an album might have a 90 or something, then just one review comes along and slams it and it's suddenly down five to ten points.
I'm also curious how you weigh reviews written thirty years ago against reviews written just recently.
Henrik can obviously answer these questions better, but I thought I'd contribute two cents:
The PET SOUNDS question is kind of misleading, BillAdama - "worst" or overrated albums lists are obviously not included here. Albums are "punished", I suppose, for not being included on lists that Henrik adds to the database, but in the case of a high-ranking album that's a negligible concern (e.g., I don't think TO BRING YOU MY LOVE suffered much for being omitted from the 1001 ALBUMS YOU MUST HEAR BEFORE YOU DIE book).
As for the yearly Metacritic scores, a skewed rating for a particular album probably gets balanced out if said album winds up appearing on a lot of end-of-year lists. WHITE CHALK is severely underrated on the chart right now solely because of that absurd 20, but I suspect once the best-of-07's start coming out it will end up closer to where it deserves to be.
