Updated data on Twitter before/after @ChuckWendig being fired from @Marvel/@StarWars. The new part is the number of bots (fully automated accounts) tweeting about Wendig. I'm amazed by how many bots there were. 
The # of tweets from normal accounts is in this graph. Of note: Lots more real people paid attention when he was fired vs after his initial tweetstorm. Also, @EthanVanSciver is right that legit accounts found Wendig through him – there is a bump after he tweeted his YouTube link.
Also after EVS tweeted, sketchy twitter exploded. Graph shows likely bots (automated) & 2 kinds of likely semi-automated accnts -- anon (no user description) & puppets (tweet >70/day). Bars are stacked counts-e.g., bar after Wendig’s 1st tweet is 44 bots, 8 puppets, & 4 anons.
Now all the types of Twitter stacked. I’ve said before that fully automated accounts were not that active on SW twitter (goo.gl/Zo3kHT). But 20% of the tweets in this period of Wendig twitter were from likely bots! (19% of accnts) cc: @Stephen_Kent89 @mortenbay
FAQs: I used @Botometer to code bots. Anon. accounts are suspiciously generic. It isn't based on if someone uses a real name/picture. (Not sure how I’d know). Anon is unverified + no user description. Puppets are unverified and tweet a lot (>70/day, inc. retweets).
NB: An automated/semi-automated account can take cues from a pundit account (like EVS) without the pundit knowing or doing anything special. It is the same principle as the algorithms that are continually monitoring your feed and picking ads for you to see.
