SHILLARY AND HER SHILLBOTS

HILLARY & THE RETWEETS

There are already two threads about her DELETE THIS tweet but what I've been doing has got divided between them and that's useless.

The tweet is this one - archive.is/ZLGTT

The question was asked why did this tweet get so many retweets so quickly which turned to how many of her followers are bots and why TwitterAudit has changed her score dramatically in the past couple of months.
Her Real to Fake ratio improved by 20% in a month

I have been able to create two lists of users / urls and put them into a spreadsheet for the purpose of judging if they are real or fake and to count them as they are voted.


I've started to identify cells of shillbots that seem to work in groups of ~5 to give a burst of retweets. I doubt that's scratching the surface. (See the retweeter spreadsheet)


Need some friendly anons to comb through the links and place a few votes for human vs bot. I think the retweeters list yield more interesting results.


TWO LISTS:
~690 users who retweeted the tweet.
~630 of Shillary's most recent followers - Chronological - Newest first

RETWEETERS LIST

Found by using search function for “RT @HillaryClinton: Delete your account.” – Collected from under the "Live" tab.

690 Users retweeted in this fashion. Do these go onto the official retweet counter?

docs.google.com/spreadsheets/d/1FDVUSIschmL338yvPJtlENcGHgc-MRgVFSpcNrs2ETA/edit?usp=sharing


FOLLOWERS LIST

600 Most recent followers - Data collected yesterday. Random population sample.

docs.google.com/spreadsheets/d/1i4kXzkd7632ud284D0eZi6t8_9a3wjNRMfdL1N9-KpM/edit?usp=sharing


USING THE SPREADSHEETS


0 for HUMAN
1 for BOT

Other urls found in this thread:

truthy.indiana.edu/botornot/
docs.google.com/spreadsheets/d/11qP2M4nPNnqimUtmwHsma6SWBFDWA6-vk9b55tRJkK0/edit?usp=sharing
archive.is/NOiJF
theguardian.com/technology/2011/mar/17/us-spy-operation-social-networks
twitter.com/NSFWRedditVideo

GO INCOGNITO/PRIVATE IF YOU'RE LOGGED IN TO GOOGLE

If everyone votes on 5 profiles it might not take forever to populate the table.

When the bots are counted the data should be good enough to do some interesting statistics with.

It might even be possible to prove that TwitterAudit has manipulated their ranking system in favor of Hillary.

I would think that at the very least Hillary is spending big money on the bots that can't be detected as well by the twitter audit.

I imagine the twitter audit has a very simple algorithm that does something like only check if there's a profile picture, last activity, etc.

What are the possibilities then?

-The twitter audit algorithms stayed the same and Hillary just upgraded her bots
-Twitter audit directly manipulated their scoring system

...

Go look at the groups of bots on twitter. You can watch them work in unison live.

...

He seems to be responsible for the first batch of bots.

...

Who is this guy?

Nothing?

I think I've found a way to spot and identify shillbots and their groups.

I'll continue to work down through the list but I'm going to bump this once in a while to stop it dying.

I figured I'd do the same style of search for Trump's recent popular tweet


There's literally only 3 results

I've just checked Trump's follower list.

Is his follower count exploding right now? Scrolling down the list I'm seeing easily 30%+ egg accounts i.e. most obvious bot and doesn't include more sophisticated bots.

If his follower count has exploded, with bots, then I bet the Hillary campaign has something to do with it. They're going to try taking Hillary's gleaming most recent Twitter audit and compare it to Trump's.

Her last audit was June 9th.

Trump's was May 26th - He's due an audit and 1 million new followers can seriously skew his results, especially against a crooked system.

...

Are those bots saying it's okay to come out against the holohoax and other assorted jewish parlor illusions?

TwitterAudit also magically changed Anita Sarkeesian's scores a few weeks after Milo mocked her about it. TwitterAudit is probably for sale. Although probably only via PR agencies.

~15-20k increase each daily
Shit site won't give a nice line graph of total followers against time

It's practically PC friendly to be anti-Jew because it gives (((them))) something to cry anti-semetism about.

Did the Google manipulation scandal turn into anything interesting?

It's hard to prove satisfactorily manipulation in big data by Google and others. Everyone just /knows/ there's corruption but finding the proof is difficult. Compound to the problem of dealing with people who don't possess a mind that will actually react to proof or reason.

I think they're actually saying if we wanted to do some holocausting today, we could do it with 1/13th the manpower required in the immediate post-WWII era

I went about it the same way and have a list of over 5000 of Trump's most recent followers.

Intuition from watching the list scroll as I was saving it says that >30% are the most obvious of shitty shillbots.

Trump is being followed by the 10s of thousands and they're mostly shillbots. Who is responsible and why?

Those twitter feed screencaps… My wife has a job for a website called 99dollarsocial or something like that. The job is as follows:


basically.. my wife is a literal waifubot. She has only gotten clients that were like plumbers and dentists so far but the second she gets something like this shit im gonna leak it all here.

In those particular screencaps they appear to me to have too identical timing to be a paid human shill.

Any understanding you have about human based shilling is definitely welcome.

I bet the bot shills apply machine learning techniques to human shills

yes and no. What my wife does for this company is you write up the posts for the week or two week period and then you schedule them to post. There are plenty of websites and programs that allow you to schedule when posts will be made. So basically it is very possible for shillbot posts to be human designed and scheduled for when to be posted.

yes and no. What my wife does for this company is you write up the posts for the week or two week period and then you schedule them to post. There are plenty of websites and programs that allow you to schedule when posts will be made. So basically it is very possible for shillbot posts to be human designed and scheduled for when to be posted.

But while they're not fully automated and posted by a bot, they're orchestrated.

You need orchestrating to make astro-turf look like grassroots.

I salute your autism, noble digger-sempai, this is some good shit.

What sort of money could it pay to shill full time?

I've looked through a bunch more from that retweet list and unfortunately haven't found any more batches of bots like that.

There are a lot of niggers though and the more I see them the more I think a lot of them must be automated. Constant random inane crap interrupted by bursts of sports retweets.

Remember, Microsoft wanted Tay to be a nigger. I don't know AI but niggerspeak must be entry level. Therefore I think blacktwitter is made of bots.

I'm messing about with basic code and want to add a few little bits. I'll quietly continue bumping for a little while until I get bored with it and move on. Don't mind me.

I might be able to move one column into another one at the click of a button. Hold onto your dick…

truthy.indiana.edu/botornot/

This site looks promising. Getting some lists of usernames isn't too hard, should be possible to run it through their script.

That list was created from the latest retweets of her most popular tweet. At over 400k it was likely genuinely viral with humanfags. The high 80% People to Bot ratio is likely due to this.

Notice it's still wildly off Twitter Audit's verdict of 90%

They've got a python script that allows you to use their botchecking algorithms through twitter's API. Twitter wants a mobile number for api access so so I need to wait a day or two for some sim cards to arrive.


This list

docs.google.com/spreadsheets/d/11qP2M4nPNnqimUtmwHsma6SWBFDWA6-vk9b55tRJkK0/edit?usp=sharing

This latest list (see screenshot and link above) is a sample of 7 of her most recent tweets, with low numbers of retweets. The retweeter were gathered from the limited list offered by twitter behind the "Retweets" button. The choice to aim for early retweets was decided on the intuition that if you want something to spread you want to give it an early boost with your bots rather than drag it out over a long period of time. (Remember the sheet and list is open to everyone if you want to peruse and leave some opinions)

This list is looking much more populated with bots. ~30% human so far. I'm being a little harsher with how I've been deciding what is a bot because frankly the bots are designed to be difficult to spot and I don't want to believe humans are consistently this shit.


archive.is/NOiJF
theguardian.com/technology/2011/mar/17/us-spy-operation-social-networks

Identifiers of a bot
High tweet count. >100k for an account that's 5 years old is still very high.

Large batches of tweets about sport. Seemingly normal profiles from "people" with no obvious interest in sport will randomly start spewing a lot of NFL/NBA garbage. There's a lot of chatter about "LeBron".

Replicated timelines - Some profiles seem to share high percentages of the content they retweet and the retweets happen at the same time.

Unusual following/followers - Often too high but not always. Bot accounts will often follow and be followed by other obvious bot accounts.

Things to note.

The top replies to any of Hillary's tweets are always fucking golden.

It just doesn't stop.

It sure as hell isn't Hillary running the account but it's going to be people who monitor it. They can't escape all of this information.

considering she already has jewgle manipulating their search bar in her favor I'm leaning toward this one.