Short answer: no.
This interesting infographic from London-based agency 1000heads, which analyses the buzz surrounding Saturday’s boxing title fight between Briton David Haye and Ukrainian Wladimir Klitschko, compares the opinions of folks on Twitter with what is perhaps the more established views of the boxing pundits.
Bottom line: sometimes it pays to listen to the experts, and Twitter isn’t a reliable indicator of an outcome for anything in which its users cannot influence.
As we know, Klitschko won. So how did Twitter fare as a prediction tool for the fight? Pretty poorly – just 32% of the tweets survey expected Klitschko to emerge victorious, compared with 80% of boxing pundits.
Twitter is often touted as a way to crowd-source data, but as I’ve argued before this only really works if those surveyed can actually have an impact on the end result – for example, during voting for a show like American Idol. In examples like that the results on Twitter, if wide enough, should be pretty indicative of the result on the show, as people who care enough to voice their opinion on Twitter about the likely winner of American Idol are probably voting for that person, too. Same goes for Presidential runs.
But something like a boxing match (or the Academy Awards), upon which the users of Twitter have absolutely no influence, becomes nothing more than a guess, no matter how many people are polled. The results can have no impact on the actual fight, and it doesn’t make any different what Twitter thinks. If accurate, it’s simply a coincidence. Correlation does not imply causation.
Still, that won’t stop us asking. And why not?