List of the most popular albums overall

An idea for this Web site came to me during a discussion in one of the other forums, and I wondered if it would be possible. Well, actually I think it would be possible, but it might not be of practical use. But here is my idea anyway:

The site already has a list of the top ten most visited bands on the site. I'm assuming that the list is generated dynamically (i.e. 'on the fly') or at least derived by the Webmaster on a reasonably regular basis. How about a list of the top ten (or twenty) most popular albums overall? From the reviews lodged on the site it might be possible to derive a top ten (twenty would be better) list and display it. As the number of reviews increases, the list could become more and more indicative of the most popular prog albums. OK, the number of people who visit this site is a probably a relatively small percentage of prog rock enthusiasts, but that still amounts to a lot of people. I would be interested to know what are the collective fan-base's top ten (or twenty) albums.

The problem is how to compute the ranking of each album and - just as importantly - how to get differentiation between the hundreds of albums that could theoretically end up with equal ranking based on the site data. As the site is not selling the albums, popularity cannot be gauged by number of sales. The two fundamental variables available per album are: number of reviews and number of stars per review. It would be interesting if the Webmaster could arrive at a solution.

One possibility (not that I'm advocating it) would be the following:

Ranking Value (RV) of an album := sum of the stars for all reviews of that album

The higher the value of RV, the higher up the list the album would be placed. If there are no reviews, then RV is zero, ergo that album is deemed not popular. If an album has 20 reviews and the sum of the stars is 100 stars (average of 5 stars), then RV = 100. If an album has 50 reviews and the sum of the stars is 200 stars (average of 4 stars) then RV = 200 and it is therefore deemed more popular than the aforementioned album.

But how can the album with an average review of 5 stars be ranked lower than an album with an average review of 4 stars? It depends how one defines 'popular', doesn't it? One could say that an album with 50 reviews and an average of 4 stars *is* more popular than an album with 20 reviews and an average of 5 stars. A problem appears to occur with the above-mentioned formula when we consider extreme cases: an album with 1000 reviews and an average of 2 stars per review versus an album with 10 reviews and an average of 5 stars per review. Common sense dictates that the first album is unpopular and the second one is popular.

Any ideas? Or is it just not practical?

MAX@, Thanks for the kind words.

I do have another idea (I'll try not to bombard you with these!) and have posted it as a new topic.