I didn't even know what to say when Newt Gingrich claimed that the press is ignoring the awesome size of his . . . Twitter following. I mean, I'm second to none in my love for Twitter (you can follow me at @asymmetricinfo), but huh? This is supposed to make him a viable candidate? If so, Alyssa Milano for Speaker of the House!
But it has spawned at least one interesting sub-story: Gawker thinks that most of his followers are fake.
A former staffer tells them:
Newt employs a variety of agencies whose sole purpose is to procure Twitter followers for people who are shallow/insecure/unpopular enough to pay for them. As you might guess, Newt is most decidedly one of the people to which these agencies cater.
About 80 percent of those accounts are inactive or are dummy accounts created by various "follow agencies," another 10 percent are real people who are part of a network of folks who follow others back and are paying for followers themselves (Newt's profile just happens to be a part of these networks because he uses them, although he doesn't follow back), and the remaining 10 percent may, in fact, be real, sentient people who happen to like Newt Gingrich. If you simply scroll through his list of followers you'll see that most of them have odd usernames and no profile photos, which has to do with the fact that they were mass generated. Pathetic, isn't it?
To which the Gawker writer adds:
While it would be impossible to survey all of Gingrich's followers, a cursory glance immediately turned up a few accounts that featured odd names, no personal information, no followers, no posts, and a small follow list.
The friend who sent me this list is, to put it mildly, no fan of Newt Gingrich. But as he points out, the methodology of checking does not back up the claims. First of all, I looked at Gingrich's most recent 100 followers, and counted up all users with a low number of tweets (<10) and followers (<25). I got about 35, which is a third, not 90%. But even that is almost certainly an overestimate; some people do join twitter, follow a few people, and then never do anything with their account again.
But of course, this is not reliable statistical methodology. The latest followers may not have the same makeup as earlier followers. To really do this, you'd have to take a representative sample of all 1.3 million followers, and develop some solid metrics based on what percentage of twitter users simply let their accounts lapse. You'd also have to account for spam--I have some followers who are clearly spam accounts, just waiting for me to say something about pR0n or other lucrative keywords.
But even my primitive method is a lot better than eyeballing the account and saying that a few of the followers sure look like fake accounts. I have so say, having looked at Newt's feed, it sure does look like someone is creating fake accounts to add to it--it got to the point where I could predict pretty reliably which users would have a low-to-zero tweet count. For illustration, all but 2 or 3 of these accounts look fake to me:
So I wouldn't be surprised to learn that Newt has hired some firm that is inflating his twitter count (except that I still don't understand why anyone would bother spending good money on such a silly exercise.) But either way, the number of accounts which are clearly real far outweighs the number which are clearly fake.
But let's say Newt is juking his stats, and further that the latest 100 followers are a representative sample of the whole list. That would still give him 845,000 followers, well over Sarah Palin's 620,000.
Of course, the non-representativeness of the sample could go either way--maybe more of his early followers are fake, and now that he has this huge twitter base, he's picking up more real people who see his numbers and figure his feed must be awesome. So 90% is possible. But it's probably worth doing more than an "eyeball" before you publish this sort of claim.
We want to hear what you think about this article. Submit a letter to the editor or write to firstname.lastname@example.org.