How do you predict which models are going to have a breakout season? Ask a top casting director like Noah Shelley? Or just wait to see who bagged a Prada exclusive? Well, a group of researchers from the Indiana University School of Informatics and Computing have come up with an alternative method: maths.

The team has come up with a complicated sum which brings social media engagement rates and the number of shows booked the previous season into consideration. As Emilio Ferrara, the university’s research assistant professor explained to The Cut: “We counted how many posts the models did, and how many likes and comments each post would get, on average... We collected the data on their portfolios: how many runway walks the models had already done, and a few other figures such as height (and) shoe size.”

This information is fed into a computer, which spits out the results, with an 80 per cent accuracy rate. Impressive. However this approach is not without its limitations. For one, though social media does have an increasing bearing on models careers, it’s certainly not the be-all and end-all. Take Dazed cover star and Prada campaign girl Lineisy Montero for example. With 9,540 Instagram followers, her success far surpasses her engagement rates – down, no doubt, to her unique natural beauty.

Likewise some models or “show girls” have great success on the runway but not so much when it comes to campaigns or editorials. And that’s simply because shows are what they do best. Take Waleska Gorczevski, who was last season’s most popular girl, walked an astonishing 67 shows throughout fashion month. However when it came to major campaigns, she didn’t land a single one.