Discogs is on a mission to build the biggest and most comprehensive music database and marketplace the world has ever seen. A site with discographies of all labels, all artists, all cross-referenced, and an international marketplace built off of that database. It’s for the love of music, and we’re getting closer every day. The Rank Point system is there to encourage people to add more submissions, create new Master Releases, edit existing Releases/Masters/Artists/Labels, add images to Releases/Artists/Labels – all of which help us towards our mission.
Top Contributors By Rank Points, October 2018
Coming out of SPIN there were (as expected) a few comments floating around the Database Forum implying that the Discogs Rank Calculator is not fully reflective of the varying effort that goes into different kinds of work in the DB, nor perfectly efficacious towards its end. So about a month ago I asked a the Discogs Database Community for suggestions on a different rank point calculation for the October Leaderboards: Alternative Rank Point Calculator Ideas.
Diognes_The_Fox on where the original idea stems from, and what it means to him today: “While rank doesn’t really translate into much, at least for me it’s a semi-tangible measure of effort, and points assigned should reflect that. At one point, the points assigned did make sense. In the early days, submissions were pretty limited. There weren’t any companies and there were just a handful of credits. Subsequent edits really couldn’t add much more data than the minimum amount.”
Given how much the Database has changed over the years perhaps it comes as no surprise to some that the system feels outdated. On the other hand I tend to agree with the sentiment that it’s at least a semi-tangible measure of effort. However, given the myriad different kinds of contributions people make, perhaps a system that shows people how much effort they’e put in within their specific domain of interest and/or expertise, and a system that reflects the complexity/impact of those contributions, would be nicer.
Within the replies there were some pretty consistent recurring themes across all of them:
- Obscure Releases are more important to archive than mainstream, since they are more at risk of disappearing without a trace:
- Many say physical is therefore more important than digital – but sometimes digital-only Releases are also lost in the ether
- Shellac/Pathé, Cassettes and 8-Tracks are the most important formats to add, since they’re most at risk of being forgotten
- Older is more important that newer
- Complete submission are more valuable incomplete ones
- Votes are also incredibly important
- Release images help to Release verify information, Artist/Label images are important because they help with disambiguation (and high resolution images are better than low)
- YouTube clips on Release pages are one of the most used features of the whole site, indicating they are valuable and therefore their addition should be rewarded
- ‘Bigger’ Releases (i.e. 100xCD box sets) should be worth more
- The BaOI / LCCN fields are very important for determining uniqueness, so they should carry more rank weight
- Successful merges are important but at the moment are unrewarded
Taking these factors into account I began trying to identify and list some of the most active Database Contributors from October, without relying on just rank points. The more I thought about it the more important it seemed to me to recognise all the different things that people do, rather than trying to roll them all up into a single index. Taking heed of Diognes_The_Fox‘s advice, who indicated his preference for a shift towards a more “a-la-carte system”, one that reflects the complexity/impact of the data being added, I decided to calculate different points for different data, so it’s clear where someone’s strengths are.
In the end I ended up making XXX leaderboards. You can see the top YYY Contributors for each category below.
Click the images to see the Discogs Top Contributors, October 2018 doc, with a list of the top 100 Contributors per category (where you can use ctrl+F to find your username).
NOTE: ‘Item Submissions’ refer to the number of individual vinyl records, CDs, tapes etc. submitted during October. For example submitting 1 Release that is a double LP would be worth 2 ‘items’.
^NOTE: This calculation looks at the number of years between Release Date and Submission Date (wherever the Release Date exists), and calculates the total number of years of age submitted, and average Release age of all Submissions.