#squimbus

1) the effect of the popularity of a fandom on polls involving its characters

2) how much a less popular character fitting the theme offsets this (perhaps if squimbus hadnt been bipolar either they would have gotten even FEWER votes in the example) (...it was bipolar in the original squimbus post, right?)

so i have two ideas.

1) two polls.

a) would test the first part of the effect. I ask people to submit their NEUTRALS (characters from popular fandoms that are widely known but people dont have much of an opinion about, they wouldnt win any popularity contests in their fandom but people wont vote against them to see them lose) and their SQUIMBUSES (characters that are beloved in their fandoms, but those fandoms are small.) I randomly seed a bracket and run a standard popularity contest, with each poll also linking to a poll that asks which fandom you KNOW MORE ABOUT. I would expect the n v n and s v s fandom statistics to roughly match the results, but the difference between the n v s fandom and character polls should give an idea of how much the belovedness of squimbus can offset the popularity of the neutral.

b) i then run a poll much closer to the original squimbus post, where I pick a theme, then ask people to submit their PIKACHUS (popular characters that do NOT fit the theme) and their SQUIMBUSES [squimbi?] (niche characters who fit the theme VERY WELL. I again seed them randomly and connect fandom knowledge polls, and see how the offsetting effect of MATCHING THE POLL THEME compares to the offset from the INCREASED LIKEABILITY of the squimbi in 1a

2) I do only one poll, essentially the same as poll 1a but i ask which character is more likeable. the idea behind this one is that if people think the neutral isnt very likeable they might look into the (hopefully likeable) squimbus and vote for them. this is much less work but its not as similar as id like to the original squimbus poll since "likeability" is a lot more nebulous than "has bipolar" since likeability is very subjective and bipolar has signs that people, esp people who are bipolar, will recognize

of course the problem with both of these is that i couldn't run a properly blind experiment, because i'd have to take submissions to get enough squimbi if nothing else. but tumblr does seem to love doing science with polls so i hope they would try to vote as they would in any other poll rather than vote for squimbus bc they know theyre supposed to be the "better" option