Current year chart (& algorithm)

[albummaster]

[hippo]

So, for the current year, only decade and year charts count for points? Seems fair as favouritism can affect rankings easily (putting an album released by your favourite artist as a top 10 of all time). This is a good change.

I don't understand how The Suburbs by Arcade Fire dropped 2 places. Is it because some people didn't update their overall chart since 2010?

[paladisiac]

I'm late to the discussion, but sometimes i think this is how it should work:

Overall chart influenced by overall charts, decade charts and year charts.

Decade charts influenced by decade charts and year charts.

Year charts influenced by year charts only.
_________________
fav artists NOW | ALL-TIME favs | i listen 2 more music than u so u don't have 2!

[albummaster]

This is a good idea & was considered when year/decade charts were introduced but the implementation means having three separate scores across the site for each album (year/decade/overall), plus potentially for other types of chart if other types are introduced. The 'all in one' score just seemed (at the time it was introduced) a bit less confusing for visitors and easier to change if changes were needed.

[blueandthemoon]

Ah! So now I know why the 2014 chart changed so drastically! I like the change, as there was definitely too much weigh given to 2014 albums that were put in overall charts. TBH, you can't really know if an album is deserving of an overall spot unless you have had time to sit with it (I'm talking months if not years here). I know that no album from 2014 would be on my overall chart (yet), and applying this to everybody seems fair.

BTW, this is my *first* forum post and I just want to thank AM for making a fantastic site. I consider myself a music buff, but it always surprises me how much amazing music I haven't even heard yet. Thanks for everything and keep up the good work!

[zexen47]

I'm not sure this is the solution. To me, it just means kicking the problem forward till Jan 1st. If outliers are the problem, and an album that appears in 10 year charts is ranked lower than an album that only is featured in a best ever chart, maybe the solution is to discard outliers, like not adding points for the best position in a best-ever chart, or only counting best ever chart points if it appears in at least X charts like that (say, 3 or 4).

Right now, current year albums are being brought down in the rankings artificially, and they'd be brought up as soon as the year ends, without ever actually solving the problem, in my opinion

[Rockdrigo]

IMO; This is just taking out the way of having an influence in the site for lots of people (as far as I know most people don't have decade or year charts), and the way the algoritm was before was already serving the purpose of showing the best albums of this year.

One quick first: One problem I have is sending the album hundreds of steps behind (but this is apparently going to change as soon as the next year begin). Hopefully the whole next year won't stay with the new system.

Now, the question is what we should value more: Lots of people giving little amounts of points or a few ones giving lots of points?

If we see the site as a whole, having an album with certain amount of points over another with fewer points means the first one deserves to go higher, because a percentage of the site's points has been given to it that could have been given to another but wasn't, so it doesn't matter how many users think (by giving their points) a certain album should occupy a position, rather how many of their points are willing to give to prove it.

If album A is higher than album B is because more value was put on the former, and having everybody the access to giving that value (via their personal greatest chart) we don´t need any more or less, the thing was working before the change.

(Excuse me for my grammar, still learning English)

[albummaster]

Just to respond, the main reason for implementing the change was to remove the effect of people coming to the site, leaving a chart, and then never being seen again. These are the charts that were causing most of the issues.

We'll have to see what happens in the new year, as it's too early to make any predictions. Obviously, we'll all be keeping a close eye on things & further changes can be made if they are needed. It'll be fairly obvious if there's a sudden rush of 2014 albums appearing at the top of people's overall charts (& I really don't think most people are out to do this).

[Sean]

I don't like this. I think if someone listens to an album ten times from 1979 as oppose to listening to an album from 2015 it will be the same experience. I'd like to see To Pimp A Butterfly compete with good kid right away
_________________
One evening 1945 -

http://www.besteveralbums.com/thechart.php?c=3385