If you work in social media, or any online media site really, for very long, you learn that it’s hard to predict which post or piece of content will go viral. That doesn’t stop people from trying.
The latest attempt? The New York Times has the details on a collaboration by three computer scientists who developed an algorithm that, with relative accuracy, can tell you which of two tweets to the same content by the same user will more likely be reshared. This is how those developers explain their project:
… [W]e take advantage of the surprising fact that there are many pairs of tweets containing the same url and written by the same user but employing different wording. Given such pairs, we ask: which version attracts more retweets? This turns out to be a more difficult task than predicting popular topics. Still, humans can answer this question better than chance (but far from perfectly), and the computational methods we develop can do better than an average human …
How is that possible? A huge body of data to pull from. In A/B tests, it predicts which tweet will be more popular correctly 67 percent of the time, compared to the 61 percent of tweets more likely to be retweeted that humans guess correctly, according to the NYT. Before you get too depressed, read the full article to see why your computer won’t be replacing you or your social community manager anytime soon.
Then just for fun: The NYT’s The Upshot takes this idea one step farther and put together this fascinating 25 question gut check to see if YOU can beat their algorithm and predict with more success whether one tweet will go viral or one tweet will go silent.
It’s harder than it sounds! I got 15 vs. the computer’s 19. So what do you get?