I want to express this as briefly and neatly as possible. I’ve written before about what differentiates a [fixed] point-based rating system from relative ratings (my rantish is nothing new by now), but I revisited my AniDB account the other day, glanced at my votes, and realized how completely fixed they really were.
Yes, I’m sitting at home, drinking some Newcastle on my Spring vacation and feeling rather down that I didn’t quite enjoy any anime or manga this week 
My own fault really, as I wanted to be less productive, but somehow managed to fire-up over an xul application. Anyhow…
This is a little dev-shot of a small XSL transform on some of melative’s API interface (the Action-Stream thingy). Basically, there is this event wrapper I built within the system, and it’s broken down into 3 sections (Action-Announce-Auto… haven’t thought of a name for the 3rd). More importantly is the fact that all 3 are handled exactly the same way, and have one great, underlying requirement… context.
Earlier this year I wrote about using melative to leech data for season previews, but the ultimate issue has been whether the information is actually in the system or not. Unfortunately, this ridiculous one man show has to spit some hacks to make adding and editing title info “more efficient”, though it still takes quite some time to do the process.*
Nearing the beginning of each season, previews are a big ticket item, especially in the strong seasons of Spring/Autumn. Some of the best I’ve seen come from hashihime and I rather enjoy the group takes such as on THAT and Scrumptious, but what makes a good preview, and more importantly, why the hell does everyone do the work of gathering the basic information on series separately?
Just a short bump, but while pondering ratings with hikago in #ab, I decided to try and test another system’s values on the melative system. I saved the technicals for the melative blog, but I hope anyone that caught the previous post checks it out and maybe can express its shortcomings.
Also featured in that post is a descriptor of what the values mean, and though they do not show on the current implementation, they are calculated and stored in the db; so little time to fix things.
I would like to say that any of these cataloging sites like ANN could just up and use such a system, but I’m afraid it would break (series ratings would swap many places), because the premise was not there in the beginning. A typical issue would be that some users’ ratings would be zeroed due to the redundancy of their votes…. I’m sure if a bunch of ratings went for zero contribution it would cause a ruckus. While it is possible, I don’t recommend restricting list size.
Perhaps it isn’t a secret, I’ve a thing for wondering why individuals can sincerely trust finite point-scale ratings (10 balloons, 3 hearts, 5 clouds).
Melative Current Anime