It was a wonderful job. I'll miss this

I never got many of the clues, though. But that's my fault and was part of the fun. Finding out who was who.

Good luck to Stefan ... and whoever takes over.

How about a collective effort?

We could create "users" on Rate Your Music for different magazines that rate 1-5. A meta-user "Acclaimed Music" is also created and it befriends the magazine "users".

Then the custom charts of the "user" Acclaimed Music can easily be used to see which albums receive the most acclaim around the world.

With as little as 10 non-american and non-british magazines I think this would supplement metacritic very well. If magazines from those countries were included, it would of course be even better.

Sounds like a brilliant idea, but I don't know if the RYM moderators would allow critics' accounts. Maybe someone here is a moderator?

They probably wouldn't like it for copyright reasons, but we wouldn't have to advertise it. The magazine
"users" could be called Acclaimed Music 1, 2, 3, etc.

Oh, I meant that they wouldn't like it because the ratings wouldn't be personal.

We could, however, do something like that on a website. I might be able to set up a PHP script where anybody can go to the website, enter a review source, an album, and a score, and have it saved to a central database, which one can then just run an algorithm on.

For the past 3.5 years this has been a very welcome source for finding good new music. Would be great if it could continue in some kind of fashion. The RYM construction sounds good to me, if that's allowed to do so.

As an early winner of one of the CDs (TVOTR's Dear Science) I give special thanks to you, Stefan. Wishing you all the best!


Edit: Just saw BillAdama's post crossing mine. Collecting the data on a separate AM webserver would also be great.

If someone knows how to set this up, I think it is the better solution. The best solution would be if it could be integrated with AM I suppose. But I think we should get i started sooner rather than later so that we don't fall behind from the start.

While we're waiting to see what happens concerning the above discussion, I've created a list that I think could be of interest to visitors of Acclaimed Music. It's a list of the most acclaimed albums according to Swedish media so far 2011. It is based on more than 30 media outlets.

http://rateyourmusic.com/list/dr_Robert/most_acclaimed_albums_in_swedish_media_2011/

Hopefully, a few others could do the same for their countries.

Cool! Have you done anything else than taken the results from kritiker.se and removed the albums with less than 10 ratings? Since you're taking into account the number of ratings anyway, you might want to use an adjusted score like Stefan did. However, it seems like the number of sources has increased rapidly during the year (or is there another explanation to why Anna Järvinen has got 11 more ratings than PJ Harvey?) and then the adjusted score will not work so well.

Anyway, the formula is
Adjusted Score = (N×R + D×M) ÷ (N+D) where:
R = Average rating for the album
N = Number of ratings for the album
M = mid-range value (this could also be the average across all albums, but presumably this is unknown). kritiker.se says that "Recensionsbetyg räknas om till en gemensam skala och ett medelvärde mellan ett och fem beräknas" so the mid-range value seems to be 3.
D = Degree of disbelief. A higher number means less belief in each rating, albums will then get an adjusted score closer to M (particularly albums with few ratings). You have to decide the value of D. In your case it should probably be somewhere between 5 and 10.

Looking a bit more at kritiker.se and the number of sources for other albums that has come out during 2010-2011, it seems that the number of sources is not increasing so fast and that the adjusted score should perform well. But, if you don't want/have time to use the formula that's fine. It's only a suggestion.

If I'm going to do something along these lines I'm going to need a server which can run PHP scripts and SQL.

I wanted to "publish" a result fast so I took the easiest route. I might use your formula in the future since most of the work with creating a list has now been done, but it depends a little bit on how time-consuming the calculations are. I don't think using the formula will affect the list MUCH and the interest in the list seems to be limited.