Version 5.4 ~ Haruhi gave rock and roll to you.
karmaburn.com karmaburn.com

6 May 2013: Objectively ranking anime

Emi
I wonder how long Emi is going to keep that bandage?

For the past few years now I've kept a running update of anime rankings, mostly for use in spamming IRC. At the season's end, I also use those rankings to determine each show's position in my season summaries. This season, I've started keeping a spreadsheet on Google Drive, illustrating the changes in episode-by-episode rankings on charts. At the same time, I still maintain my IRC list of show rankings, based entirely on how I subjectively feel shows compare against other titles within the same cour.

Eren
Making the impossible possible since 850.

The newfangled spreadsheet is ostensibly objective, but the data that goes into it is still derived subjectively since all the figures come from how I personally score episodes on a linear scale of one to five. What I've noticed though is that the rankings determined via the spreadsheet method don't necessarily agree with my overall subjective rankings. For example, at this point I currently consider Hataraku Maou-sama! to be the best show of the season, mostly because Shingeki no Kyojin has had some pretty annoying Ellen moments, and because Suisei no Gargantia—although consistently good—has not had as many bits that wowed me.

Amy and Pinion
"Not entirely stable? I'm glad you're here to tell us these things."

But looking at the spreadsheet experiment, Maou-sama! should be tied for second based on median scores or arithmetic means. If comparing geometric means, then it's no longer tied with Attack on Titan, but it's still in second place behind Gargantia, the undisputed leader. Does this mean anything other than that opinions are opinions? I suppose not, but I wonder if some people would be tempted to game the scores to correct for what is arguably a flaw in methodology (or I guess philosophy).


«« Spring 2013 initial impressions
Giants eating people is a metaphor for class warfare »»

Related Posts

.