Guardian Indicts Twitter With Their Own Bad Math Skills

  • Reporter Shanti Das says 99% of hate complaints on Twitter are unenforced because one left-wing pressure group made 100 complaints and only 1 was successful
  • Guardian then erroneously claims that proves Twitter ignores 99% of complaints

OUR RATING: Major Negligence. MSNBC-level basic journalistic negligence

Indicted Outlet: Shanti Das | The Guardian | Link | Archive | 11/20/22

Since Elon Musk bought Twitter in October 2022 for $44 billion, [1] the media has sought to demonize and prove that Musk’s statements about permitting free speech on Twitter will result in an expansion of hate and other chaos on the messaging platform.

The desired media narrative is that free speech equals hate and chaos everywhere.

Many celebrities have pre-emptively left Twitter concerned that a free speech Twitter is one that permit verbal violence against preferred left-wing minorities. [2] The mainstream media isn’t even trying to hide the fact they’re hoping Twitter will fail. [3]

Previously Twitter’s speech police were so extreme that they banned President Trump for unclear reasons, [4] they banned Gateway Pundit Publisher Jim Hoft for posting the video of the late-night Biden ballot dump on election night 2020 in Detroit, [5] among many other questionable bans. [6] Twitter also had multiple methods to police speech short of outright bans, including ‘shadow banning’ and other methods to suppress certain speech. [7]

The party line from Twitter was that shadow banning was not real. [8] The reality was that there were multiple tools online to determine whether you were being suppressed, [9] and to what extent. [10]

So in order to ‘prove’ that so-called ‘hate speech’ is rising on Twitter, the Guardian took a report of 100 complaints about specific comments by a motivated left-wing speech police group, and noticed that 99 of the 100 complaints were unresolved to their satisfaction.

Thus to the Guardian, 99% of Twitter ‘hate’ speech is being allowed now and not deleted. This is a classic logical fallacy and thusly probably not true.

Major Violations:

  • Statistics Abuse

This is what we call statistics abuse in reporting. It’s purposefully misusing statistics in order to get a desired political result.

It’s also a well-known logical fallacy, sometimes known as ‘hasty generalization’ [11] and also ‘fallacy of composition’ [12] where one takes a small set and assume that it is reflective of the whole.

Here’s what that fallacy is in a nutshell:

the fallacy of examining just one or very few examples or studying a single case and generalizing that to be representative of the whole class of objects or phenomena.

Twitter has more than 238 million users. [13] There are roughly 500 million tweets per day. [14] From that universe, the Guardian’s proxy, the Center for Countering Digital Hate, took 100 comments and singled them out for complaints.

It’s hard for the mind to think of numbers larger than a few thousand, but as a comparison point there are 537 million people who live in North America.

Here’s the relevant excerpt from the Guardian article:

Tweets hurling racist abuse at footballers, including the N-word, monkey emojis and calls for them to be deported, are not being removed by Twitter.

New research shows the platform failed to act on 99 out of 100 racist tweets reported to it in the week before the World Cup.

Only one was removed after being flagged on Wednesday, a tweet that repeated a racial slur 16 times. All the others remained live this weekend.

The abuse was aimed at 43 players including England stars Raheem Sterling and Bukayo Saka, who were among several players targeted after the Euro 2020 final.

The analysis, conducted by researchers at the Center for Countering Digital Hate (CCDH) and seen by the Observer, included 100 tweets reported to Twitter. Of those, 11 used the N-word to describe footballers, 25 used monkey or banana emojis directed at players, 13 called for players to be deported, and 25 attacked players by telling them to “go back to” other countries. Thirteen tweets targeted footballers over their English skills.

So they complained about 100 tweets on Wednesday and were upset that, by Saturday, only 1 was removed.

The Guardian and the CCDH identify two British footballers by name, Raheem Sterling and Bukayo Saka, and complain that tweets hurling racial abuse at those two were not promptly deleted, including comments using the ‘n-word.’

So your friendly neighborhood Fact Check agency went on Twitter and searched both of their names using the n-word.

The results were interesting.

Elon Musk took over Twitter on October 27, 2022. [15]

39 instances of referring to Sterling as an ‘n-word’ exist, all of which pre-date Musk’s takeover of Twitter. [16] Most of these mentions are by fellow blacks who are using the word to describe how they anticipate the media will unfairly cover Sterling.

The results for Bukayo Saka were similar. [17] There were no combinations of name and epithet that appeared after Musk’s takeover of Twitter. The people uttering the offending language appeared not to be using the language as an epithet, complicating any removal of a post that merely used the offending words.

I was able to find one tweet that called for Bukayo Saka to be deported, but in closer review it was actually a left-wing individual making a bad faith argument in the voice of his political opponents. It wasn’t anyone actually calling for his deportation, it was a left-wing person saying that’s what he would expect right-wing people to do if they were consistent about wanting Britain to be British-only.

Here’s the left-wing tweet trying to mimic what he thinks is right-wing thought by calling for Saka’s deportation:

Such is the challenge of policing speech: few people speak in plain ways. If a perceived right-wing person says someone should be deported and a left-wing person mocks a right-wing person saying they want to deport someone, the only visible difference may be in their intent.

None of that nuance is in the article and its missing context to the story. Many of these claims are hard to verify with particularity.

The supposed CCDH study is not reproduced. The supposed tweets are not identified and sourced. We have to take them at their word and they are a motivated left-wing pressure group. The story would have more credibility if the original offending tweets were included for anyone who wanted to verify the accuracy of the story.

But even if true, 99 out of 100 tweets surviving 48 hours in a pool of a billion potential tweets over that period of time isn’t a representative sample of the whole. Not to mention the methodological problems in properly determining what is an offending tweet and what isn’t, suggests this story is flawed on multiple levels.

Racial epithets and related kind of comments may be increasing on Twitter, but this Guardian article does nothing to actually advance public understanding of whether or not that is true.

OUR RATING: Major Negligence. MSNBC-level basic journalistic negligence

1 ]
2 ]
3 ]
4 ]
5 ]
6 ]
7 ]
8 ]
9 ]
10 ]
11 ],class%20of%20objects%20or%20phenomena.
12 ]
13 ]
14 ]
15 ]
16 ]
17 ]

Join the conversation

We have no tolerance for comments containing violence, racism, profanity, vulgarity, doxing, or discourteous behavior. Thank you for partnering with us to maintain fruitful conversation.


TGP FactCheck