[RFNAPLES]

An easy algorithm was already stated by seraphlaim: Critical representation and user representation should be averaged as separate entities and then averaged together afterward, thus the site becomes an aggregator where the two voices are congruous.

Albums are ranked in most “recognized charts” either by numerous critics or by numerous voters. Most Best Ever Album charts are individual member recommendations. While I applaud your effort to giving members representation in the Overall Chart, I fear that with an increase in member charts and selections you will dilute the considered opinion of the recognized charts.
_________________
Top 100 Greatest Music Albums by RFNAPLES
Bubbling Under The Top 100 Greatest Mus...y RFNAPLES

[albummaster]

The average rating for the recognised lists is less than the average rating for the member lists, which seems at odds with the proposition that these are somehow 'better' charts. The average rating of recognised charts is 59/100 (the site average for member charts is 66.) Why should these lists be given more weighting? Accepting your point that these lists might be the aggregate opinion of 1000s of voters (but not in all cases), what do we gain by splitting them? (I don't personally think it serves any purpose.)

The current formula is not fool-proof and I welcome specific ideas to help improve it, but I personally think splitting out the recognised lists does not improve the algorithm.

A more likely route to tuning the algorithm would be to somehow take member ratings into account when calculating the 'worthiness' (or otherwise) of a chart. For example, the points available might depend on how highly a chart has been rated, but there are thousands of ways that the score could be determined. For the moment, imho, the current method seems as good as any.

albummaster

[RFNAPLES]

Let’s look a scenario:

One recognized chart with the opinions of thousands of voters and critics gives an album a score of 100. Then one individual member chart gives that same album a score of 2. Now the average score is 51. Thousands of people think it is a great album but one member doesn’t so the average rating is quite low because of it. Fair?


_________________
Top 100 Greatest Music Albums by RFNAPLES
Bubbling Under The Top 100 Greatest Mus...y RFNAPLES

[albummaster]

No, this example is not valid. We do not use the average ranking, we use the total rank score. If a chart appears top in a chart it gets 100 points and if the album subsequently appears in another chart, the score from the second charts gets added to the 100 points it has already earned. The total is cumulative (it is not an average.)

The only time averages are used on this site is for the ratings of albums, bands and charts (and all of these ratings come from our members, not from the 'professionals'.) For the table showing the 100 top member-*rated* albums, an album has to have at least 10 member votes to appear in that list.

albummaster

[RFNAPLES]

OK let’s change the scenario:

Album A: Recognized chart (with thousands of critics and other voters) gives 100 rank points and member chart gives 1 rank point=101 Total Rank Score.

Album B: Recognized chart gives 60 rank points and member chart gives 60 rank points=120 Total Rank Score.

Album B has a better rank, since Album B has a total rank score higher than Album A. Nevertheless I would feel that Album A is better than B regardless of the total rank score because A was ranked #1 by a poll of thousands of critics and other voters while B was ranked only #60 (assuming 100 album chart size) by both charts.


_________________
Top 100 Greatest Music Albums by RFNAPLES
Bubbling Under The Top 100 Greatest Mus...y RFNAPLES

[joannajewsom]

RFNAPLES, I see what you're saying, but your argument is ultimately a flawed appeal to popularity. Isn't this site about not appealing to the critics, and letting people like you and me have our say?

Not every recognized chart has been voted by thousands of critics and other voters. Some have been decided only by voters and some have been decided only by critics, some have been decided by 100 people and some by 1000 people, and each has their own set of stipulations (such as an album bank from which to vote). All in all, every recognized chart is very different. If you're going to say that recognized charts should be weighed differently because of the difference between recognized charts and user charts (such as number of voters), then you have to consider the huge differences between the recognized charts, and weigh their vote accordingly. I'm sure you see the problem with this system.

Overall, I think user and recognized charts shouldn't be separated. Quite frankly, looking at some of these critic's lists, I trust a user's chart more than a recognized chart. Too many cooks spoil a broth, maybe?

[seraphlaim]

I agree with that but why not have the best of both worlds? If you separate it allows users to see a breakdown. You could have a link to view the users chart, and link for the critics chart and then the overall that blends the two. I think that would be intriguing to monitor, frankly. It would also be interesting if the site saved the 3 (ill call them) master charts say monthly or something. This would show the evolution of opinion and feeling from the two camps. A person could study the ascent or descent of particular albums just as public and critical opinon ebbs and flows.

As for accounting for the differing criteria for recognized charts, this is merely muddying the waters to me. Its largely irrelevant how many critics participated in a chart. Surely individual writers and musicologists are accounted for more than once as they participate in numerous lists than have then been imported into the site.

Personally speaking, the interesting part of this site is the juxapostion of two voices, the press & the people. And while there are far more of us than there are critics I find keeping those two voice seperate but equal as an intriguing evolution. I don't expect my user chart to count as much as a a panel's at Rolling Stone representing several people. I woul like my chart to count as much as my 'user' peers charts and for user charts, collectively, to count as much as recognized charts, collectively.
_________________
The greatest wonder is the realization of a question one has not yet asked .

[RFNAPLES]

We may already be able to see the rankings separately to a degree.

You can get the rankings of the 100 top member-rated albums at http://www.besteveralbums.com/topratedstats.php

You can get the rankings of the 100 top recognized chart albums by going to http://www.besteveralbums.com/makechart.php
and deleting the check in the “Include member charts?" box.
_________________
Top 100 Greatest Music Albums by RFNAPLES
Bubbling Under The Top 100 Greatest Mus...y RFNAPLES

[albummaster]

seraphlaim, you have some very good ideas. I like the idea of showing the rankings over time (it's something already on the 'to do' list.) But there are usually only one or two 'best ever' critics lists per year so the critics overall list is unlikely to change much month-to-month if it stood on its own. We could certainly implement this for the overall list.

At the moment, you can generate an overall critics list from the 'customise chart' page. You have to bear in the mind that the newest recognised chart on this site is from 2007 so by virtue of that alone the critics automatically exclude any album released in the past two years (so no Rainbows etc.) Admitedly, this might be a good thing as sometimes recent albums can carry more favour purely because they are more likely to be on your current playlist.

As you say (and maybe it is a good compromise since there are two divided camps here), we could indicate separately on this site the recognised rank and the member rank (and have separate lists for each) but still use the overall list as the 'master' rankings if that would keep everybody happy.

To develop this will obviously take time but we could certainly put this on the 'to do' list for a future release (but not too much in the future!)

[seraphlaim]

Sounds cool to me and by the way, please don't be concerned about keeping me 'happy'. I already love this site. I'm going to be an active participant anyway. At the same time its exciting me me that the site is relatively new and that the admin(s) are open to suggestions and want the users who have ideas to share them. Maybe some of us have a ground floor chance to contributing to making it even better.

You never have to worry about me taking personally an idea of mine not being adopted. I don't know that you are but i wanted there to be no question. Thanks for the compliment on my ideas and as long as they are welcome I will keep them coming. Actually, they would probably come anyway (LOL) but thanks none the less. It seems admin is smart enough to know that the best resevoir for ideas to make the site better is the users that have invested their interest in the site in the first place, and that observation is reassuring.

(forgive me that this post is so disorganized with thoughts jumping with my stream of consciousness.)
_________________
The greatest wonder is the realization of a question one has not yet asked .