Elon Musk takes a wrong approach to counting fake messages and spam on Twitter: the experts

Tesla CEO Elon Musk sent Twitter shares plummeting Friday when he said he would put his $44 billion acquisition of the social network “on hold” as he investigates the proportion of fake accounts and spam on the platform.

Although Musk later clarified that he was still committed to the deal, he continued to talk about the issue of the fake accounts. He wrote on Twitter that his team would conduct its own analysis and expressed doubts about the accuracy of the numbers Twitter reported in its latest financial filings.

Twitter acknowledged in its Q1 earnings report that there were a number of “wrong or spam accounts” on its platform, along with legitimate daily active use or users (mDAU). The company stated, “We conducted an internal audit of a sample of accounts and estimate that the average fake or spam accounts during the first quarter of 2022 represented less than 5% of our MDAU during that quarter.”

Twitter has also admitted to overstating the number of users by 1.4 million to 1.9 million over the past three years. The company wrote, “In March of 2019, we launched a feature that allows people to link multiple separate accounts together in order to easily switch between accounts,” Twitter revealed. “An error occurred at the time, actions were taken across the base account and in all linked accounts that count as mDAU.”

While Musk may be justifiably curious, experts on social media, disinformation, and statistical analysis say his proposed approach to further analysis is woefully inadequate.

Here’s what the CEO of SpaceX and Tesla said he would do to determine how many fake, fake, and duplicate accounts there are on Twitter:

“To find out the answer, my team will randomly sample 100 Twitter followers. I invite others to repeat the same process and see what they discover.” He explained his methodology in subsequent tweets, adding, “Choose any account with a lot of followers,” and “Discard the first 1,000 followers, then pick every ten. I’m open to better ideas.”

Musk also said, without providing evidence, that he chose 100 as the sample size number for his study because that’s the number Twitter uses to calculate the numbers in their earnings reports.

“Any reasonable random sampling is OK. If several people independently get similar results in percentage of fake/spam/duplicate accounts, that would be obvious. I chose 100 as the number for the sample size, because that’s what Twitter uses to calculate < 5% is fake/spam/duplicate.”

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

Dustin Moskovitz, co-founder of Facebook, studied the issue via his own Twitter account, noting that Musk’s approach isn’t actually random, uses a very small sample, and leaves room for serious mistakes.

He wrote: “I also feel that ‘lack of trust in the Twitter team to help draw the sample’ is kind of a red flag.”

Christopher Buzzi, founder and CEO of BotSentinel, said in an interview with CNBC that analysis conducted by his company indicates that 10% to 15% of accounts on Twitter are likely to be “inauthentic,” including fake, spammers, and scammers, nefarious bots, duplicates, and intentional hate accounts” that typically target and harass individuals, along with others who intentionally spread misinformation.

BotSentinel, powered primarily by crowdfunding, independently analyzes and identifies real activity on Twitter using a combination of machine learning software and teams of human reviewers. Today, the company monitors more than 2.5 million Twitter accounts, most of which are English-speaking.

“I think Twitter doesn’t realistically classify ‘false and spam’ accounts,” Bozzi said.

He also warns that the number of non-original accounts can appear higher or lower in different corners of Twitter depending on the topics being discussed. For example, more non-original accounts tweet about politics, cryptocurrency, climate change, and covid than non-discussed ones like cats and origami, BotSentinel finds.

“I just can’t understand that Musk is doing anything other than trolling us with a ridiculous sampling scheme.”

Carl T Bergstrom

Author, “Calling Bulls—“

Carl T. said, “For the acquisition of $44 billion.

He said a sample size of 100 is an order of magnitude smaller than usual for social media researchers studying this kind of thing. The biggest problem Musk will have with this approach is known as selection bias.

In a letter to CNBC, Bergstrom wrote, “There is no reason to believe that the followers of the official Twitter account are a representative sample of accounts on the platform. Perhaps bots are less likely to follow this account to avoid detection. Should I follow it to appear legitimate. Who knows? But I can’t understand that Musk He does anything other than trolling us with this ridiculous sampling scheme.”

Leave a Comment