Experts: Elon Musk is wrong about counting fakes and spam on Twitter

Tesla Directed by CEO Elon Musk Twitter the stock collapsed on Friday when he said he was going to put his own Purchase of social network for $ 44 billion “pending” while he investigates the proportion of fake and spam accounts on the platform.

Although Musk later clarified that he remained committed to the deal, he continued to kill the issue of fake accounts. He wrote on Twitter that his team would do its own analysis, and questioned the accuracy of the figures Twitter reported in its latest financial documents.

In his earnings report for the first quarter of this year, Twitter has acknowledged that it has a number of “fake or spam accounts” on its platform along with legitimate daily active use or users (mDAU) that can be monetized. The company said: “We conducted an internal audit of the sample of accounts and estimated that the average number of fake or spam accounts in the first quarter of 2022 was less than 5% of our mDAU for the quarter.”

Twitter also acknowledged that in the last 3 years the number of users has exceeded 1.4 to 1.9 million users. The company wrote: “In March 2019, we launched a feature that allows people to link multiple individual accounts together to easily switch between accounts,” – reveals Twitter. “An error was made at this time, so actions taken through the primary account resulted in all related accounts being credited as mDAU.”

While Musk may be justifiably curious, experts on social media, misinformation and statistical analysis say his proposed approach to further analysis is very scarce.

That’s what SpaceX is and Tesla The CEO said he would do to determine how much spam, fake and duplicate accounts exist on Twitter:

“To find out, my team will make a random sample of 100 @twitter subscribers. I invite others to repeat the same process and see what they discover. ” He clarified his methodology in subsequent tweets, adding: “Choose any account with a large number of subscribers” and “Ignore 1000 subscribers first, then choose every 10th.” I’m open to better ideas. “

Musk also said, without providing evidence, that he chose 100 as the sample number for his study because it is a number that Twitter uses to count numbers in its earnings reports.

“Any reasonable random sampling process will do. If many people independently get similar results for% of fake / spam / duplicate accounts, it will be indicative. I chose 100 as the sample number because that’s what Twitter uses to calculate <5% counterfeit / spam / duplicate. "

Twitter declined to comment when asked if its description of its methodology was accurate.

Facebook Co-founder Dustin Muscovite spoke on the issue through his own Twitter account, noting that Mask’s approach is not really random, uses too small a sample and leaves room for massive errors.

He wrote: “I also feel that ‘don’t trust the Twitter team that will help make the sample’ is a kind of red flag.”

BotSentinel Founder and CEO Christopher Boozy said in an interview with CNBC that an analysis of his company shows that 10% to 15% of Twitter accounts are likely to be “invalid”, including fakes, spammers, scammers, infamous bots, duplicates and “Single-purpose” hate accounts, which are usually targeted at individuals and harass individuals, as well as others who deliberately spread misinformation.

BotSentinel, which is supported mainly through crowdfunding, independently analyzes and identifies invalid activities on Twitter using machine learning software and peer review groups. The company now tracks more than 2.5 million Twitter accounts, primarily English-speaking users.

“I don’t think Twitter realistically classifies accounts as fake and spam,” Buzi said.

He also warns that the number of invalid accounts may be higher or lower in different corners of Twitter depending on the topics being discussed. For example, BotSentinel found that more invalid accounts write in tweets about politics, cryptocurrency, climate change and covid than those that discuss controversial topics such as kittens and origami.

Carl T. Bergstrom, Professor at the University of Washington, co-author a book that will help people understand data and to avoid false claims online, CNBC said that a sample of one hundred subscribers to any single Twitter account should not serve as “due diligence” to acquire $ 44 billion.

He said the sample size of 100 people is an order of magnitude smaller than the norm for social media researchers studying such things. The biggest challenge Musk has faced with this approach is known as bias of choice.

Bergstrom wrote in a CNBC statement: “There is no reason to believe that followers of the official Twitter account are a representative sample of accounts on the platform. Perhaps bots are less likely to monitor this account to avoid detection. They may be more likely to follow to give up who knows? But I just can’t figure out that Musk is doing anything but trolling us with this stupid sampling scheme. “