Â鶹´«Ã½

SPORTS fans who tweet while watching the game could help create more exciting highlights than those compiled by professional broadcasters.

and of the University of Calgary in Alberta, Canada, used Twitter to crowdsource highlights from American football games. Their system, called , gauges fans’ excitement by sudden spikes in Twitter activity, then selects video from the previous minute.

The pair found few similarities between their highlight reels and the professional versions. That’s because sportscasters assemble their reels retrospectively, constructing a narrative around the game’s outcome, while the crowdsourced highlights reflect fans’ emotions moment by moment. Tweeters typically select impressive displays of skill, rather than just scoring touchdowns. The crowd also spiked on “lowlights” when teams performed poorly or referees made controversial decisions.

Tang and Boring categorised tweets by the teams’ hashtags, making it possible to produce customised highlight reels for fans of either side. They will present their work in May at the Conference on Human Factors in Computing Systems in Austin, Texas.