Topic: "Google trends" for e621 tags

Posted under e621 Tools and Applications

Hello everyone!

I built a "Google trends"-like website for e621 tags:
https://trends.anthro.fr/

Actually, it works for all kind of queries, not only isolated tags. So you can compare things like "dragon rating:s,dragon rating:q,dragon rating:e".

I am not sure that's a useful tool, but it was fun to build it :P

Updated

You didn't explain where the data comes from and what it represents.

Updated by anonymous

AoBird said:
You didn't explain where the data comes from and what it represents.

Indeed! It's based on the number of posts on e621, I added labels above the charts :)

Updated by anonymous

SUGGESTION: Please add a feature to see the average score or something! that would be cool!

MLP has taken a big hit while pokeporn has gone stratospheric.

sonic_(series) got aliased to sonic_the_hedgehog_(series) and that seems to have screwed up that data.

I'm not as much of a statistics nerd as I maybe wish I was, but this is pretty damn cool! I like being able to compare the performance of tags over time.

magnuseffect said:
Was 2018 when bowsette_meme happened?

(edit: Yes.)

Exactly what I was thinking. Further evidence if any were required that Bowsette should have been allowed her own character tag. (Also, Bowsette posts in just the last three months of 2018 managing to outnumber regular Bowser posts in the whole year is one hell of an accomplishment.)

Anyway, here's a Pokémon battle. Species were selected by the unscientific method of entering in every species with more than 4000 posts (plus Lycanroc and Incineroar who are too new to have hit that milestone yet), discovering that it made an absolutely unreadable graph, then deleting tags more or less at random until there weren't too many lines tracing similar trajectories. Interesting findings include that the two newer Pokémon are still outnumbered by the old classics in posts per year, so it's not just a matter of the likes of Charizard having a 20-year head start. On the other hand, Lycanroc is for some reason the clear winner in favs per post, which may indicate a generational disconnect between the site's userbase and the majority of furry artists, or may just be a reflection of only having posts during years when the site was super active.

I was also rather fascinated to learn that Lucario mania didn't really take off until after Braixen was introduced. Can any furry historians shed light on this tale? Are the 2013 post stats an inaccurate reflection of where the broader fandom was at that time, or were furries waiting for a waifu to go with their husbando? Also, I have no idea what caused the great Pikachu spike of 2018.

Finally, here's an Eeveelutions race. First place is a surprisingly intense battle between Eevee and Umbreon.

closetpossum said:
RIP, but maybe that should be built into the website anyway...for improvement.
Just a thought...
would have liked to look at what's trending and what gets some kind of surge boost
As I tried to, somewhat, convey in my OWN Forum post earlier

Agreed, if one of the admins/janitors (or someone else who developed these kinds or applications) cares enough they could get in correspondence with prettybird; they may still have the source code or application saved somewhere. looks like their last activity was a post tag change 3 months ago, but a lot of their other activity dropped off about 7 to 12 months ago, so they may be weaning off of active contribution to e621, which would explain that site not being maintained and then going down. Looks like some other people have made external statistics analyzers of e621, but prettybird's might have had some features those don't (currently) have.

  • 1