So, apparently, I don’t know a good tweet when I see one. According to the New York Times blog The Upshot, a group of researchers from Cornell University developed an algorithm designed to predict which of two tweets would be retweeted more often.
The Times created a nifty quiz showing 25 pairs of similar tweets, and giving us ordinary folk the chance to predict which one got passed around more often on Twitter. I scored correctly on 14 out of the 25. The algorithm scored correctly on 18 of them.
I’m in good company, by the way. Marc Andreessen apparently only got one more right than I did, and I tied with Twitter co-founder Ev Williams and CBS news political director John Dickerson. (The best score recorded when I looked was 21/25 by Katie Notopoulos of Buzzfeed. Of course.)
There is hope for us mortals, according to the Times’ blog post on the topic:
That an algorithm can make these kinds of predictions shows the power of “big data.” It also illustrates a fundamental limitation of big data: Specifically, guessing which tweet gets retweeted is significantly easier than creating one that gets retweeted.
Why don’t you try the quiz and let me know how you did?