Rank score - Suggestions - Forums : Best Ever Albums

albummaster
Janitor

Gender: Male
Location: Spain
Site Admin

Previously, the 'average ranking' calculation on this site was a bit mathematically flawed as it was based on the 'mean' rank of the charts that an album appeared in, but because of the number of different chart-sizes that there are on this site, the average ranking could fluctuate wildy without really telling you anything about how good or important an album is; so, in their place, rank scores have been added.

The 'rank score' is determined using a formula whereby (in a chart of 100 entries) #1 gets 100 points, #2 gets 99, #3 gets 98, etc (in a chart-size of 10, #1 gets 100 and #2 gets 90 etc.)

The total rank score of an album indicates how well an album has ranked in the greatest album charts and the higher its rank score, in theory the more prominent the album is. The total rank score is used on this site to determine the overall prominence of an album in the greatest album charts and is listed as a reference statistic across the site.

Last edited by albummaster on 10/08/2012 15:44; edited 1 time in total

californication

Are user charts calculated into the overall chart?

YellowBook

Location: Andalucia, Spain

Hi californication, yes, user charts are included in the totals for the overall chart.

Richie Hunt

Gender: Male
Age: 110

The change to the calc sounds good.
I question though, old charts ie that were done in 1999 etc exclude albums created post chart publish date. Is this timing of charts which favor older albums taken into consideration of the ranking? like the older the chart the less value given to an albums score - because some charts are out of date.

Xavygravy

albummaster wrote:

The 'rank score' is determined using a formula whereby (in a chart of 100 entries) #1 gets 100 points, #2 gets 99, #3 gets 98, etc (in a chart-size of 10, #1 gets 100 and #2 gets 90 etc.)

I don't see why charts with fewer entries should get less points for their #2 (and so on) albums. It seems completely illogical, as a #2 album is still a #2 album, and should receive the same number of points as if it were a chart of 100 entries.

Also, were the old charts updated points-wise?

albummaster
Janitor

Gender: Male
Location: Spain
Site Admin

The ranking system has been discussed at length elsewhere, and it's not going to change. No system is perfect, but if you find it completely illogical, you could use the top-rated chart instead (but please bear in mind that some people try to manipulate the ratings), or alternatvely try the custom chart page where you can customise the parameters.

Charts older than ten years no longer count towards the rankings.

Xavygravy

albummaster wrote:

The ranking system has been discussed at length elsewhere, and it's not going to change. No system is perfect, but if you find it completely illogical, you could use the top-rated chart instead (but please bear in mind that some people try to manipulate the ratings), or alternatvely try the custom chart page where you can customise the parameters.

Charts older than ten years no longer count towards the rankings.

I'm not asking for an alternative, I'm asking for the reasoning behind it, because it doesn't seem to have any logic behind it.
The current system says that if someone makes a chart of the '10 greatest albums of all time', their #10 is only contributing as much as the #90 of someone else's 100 chart. There is clearly something wrong here, as in charts of 100, there are generally significant differences between the greatness of the bottom 10 and the greatness of the top 10. Unless there's something I'm missing, I don't see why you'd want to keep this obvious weakness in the system. Think

albummaster
Janitor

Gender: Male
Location: Spain
Site Admin

Xavygravy wrote:

I'm not asking for an alternative, I'm asking for the reasoning behind it, because it doesn't seem to have any logic behind it.
The current system says that if someone makes a chart of the '10 greatest albums of all time', their #10 is only contributing as much as the #90 of someone else's 100 chart. There is clearly something wrong here, as in charts of 100, there are generally significant differences between the greatness of the bottom 10 and the greatness of the top 10. Unless there's something I'm missing, I don't see why you'd want to keep this obvious weakness in the system. Think

You didn't ask for the reasoning, you said that you found the current system completely illogical, so I gave you the alternatives.

In brief, people wanted larger charts to count more than smaller charts (which has been discussed elsewhere). There are strengths and weaknesses in every system, and that's why the other options exist, so you can use the one that you prefer.

RFNAPLES
Level 8

Gender: Male
Age: 75
Location: Durham, NC, USA

Xavygravy wrote:

albummaster wrote:

The 'rank score' is determined using a formula whereby (in a chart of 100 entries) #1 gets 100 points, #2 gets 99, #3 gets 98, etc (in a chart-size of 10, #1 gets 100 and #2 gets 90 etc.)

I don't see why charts with fewer entries should get less points for their #2 (and so on) albums. It seems completely illogical, as a #2 album is still a #2 album, and should receive the same number of points as if it were a chart of 100 entries.

Also, were the old charts updated points-wise?

Is the above true? I thought #1 received the same number of rank points regardless of chart size.
_________________
Top 100 Greatest Music Albums by RFNAPLES
Bubbling Under The Top 100 Greatest Mus...y RFNAPLES

albummaster
Janitor

Gender: Male
Location: Spain
Site Admin

yes, the top-ranked album does count the same, and subsequent albums are scored according to chart-size e.g. 90, 80, 70 (in a chart of ten) or 99, 98, 97 (in a chart of 100).