Adam Ostrow of Mashable posted an article “The Algorithm That Predicted Swine Flu” earlier today. The general synopsis is that a company—Replikins, Ltd—predicted the swine flu epidemic a year ago. Many scientists have come out of the woodwork over the last few weeks claiming to have made similar predictions (evoking the hindsight bias), but Replikins has a paper trail to back up their claim—a press release from a year ago indexed by Google.
In any event, the lesson here is clear: if it’s not on the Web, it didn’t happen. So if you have a bold prediction to make, be sure to document it somewhere – whether it’s your blog, YouTube, Twitter, a press release, or any of the hundreds of other options available for sharing content online.
There is nothing wrong with Ostrow’s advice, but it’s only part of the picture and raises some interesting questions. It’s a wise move to put your good ideas into writing if you want anyone they were yours in retrospect, but his anecdote is an example of the survivorship bias. Incorrect and inconsequential predictions don’t warrant articles in Mashable.
We have no idea how many inaccurate predictions were made about the swine flu or how many predictions were made—whether by Replikins or anyone else for that matter—about other ailments that haven’t become pandemics. You could make any number of random predictions and most would fade into obscurity.
Given the number of people publishing content on the internet, there is a certain tendency for a monkeys on typewriters effect to occur. If enough people make enough predictions, someone will eventually have to be correct. It’s quite possible that Replikins’ findings was the result of acumen than random chance, but Ostrow’s advice should be reviewed with a bit of skepticism.
Given a long enough timeframe, any prediction may prove to be correct.
(Photo by Olivander)
