Sochi Results Help Infostrada Sports Hone Approach

(ATR) As is true for the athletes, improvement is always one competition away for those projecting the Olympics.

Guardar
Speed Skating: 2014 Winter Olympics:
Speed Skating: 2014 Winter Olympics: Netherlands Marrit Leenstra, Jorien Ter Mors, and Ireen Wust in action during Women's Team Pursuit Finals at Adler Arena Skating Center. Netherlands won gold. Sochi, Russia 2/22/2014 CREDIT: Robert Beck (Photo by Robert Beck /Sports Illustrated/Getty Images) (Set Number: X157692 TK2 R1 F932 )

As is true for the athletes, improvement is always one competition away for those projecting the Olympics.

Infostrada Sports says though it fell just shy of its goal of predicting 50 percent of the medal winners, the process of tweaking and improving its formulae from race to race was a positive experience.

"It’s during the Games where it becomes the most interesting, I think, because then you can see what sort of sports are predictable on the basis of data because obviously what we’re trying to do here is get as close as we can," Simon Gleave of Infostrada Sports tells Around the Rings.

The company makes adjustments even as the data is still coming in, as was the case with the speedskating team from the Netherlands.

"We’ve looked at each individual performance with the Dutch - we were doing that from the beginning - and it was fairly obvious, from about three or four days into the Olympic Games, that everybody that we had in the top eight … was actually winning a medal."

Infostrada Sports correctly predicted 141 medalists of its 294 predictions for Sochi, just six shy of half.

"Comparing that to the other models that are around, that was the best, so we’re quite happy with it."

Among the highlights was their work in the sliding sports, like skeleton, luge, and bobsleigh. Gleave says they produce enough data prior to the Games that they are able to get a handle on how the talent stacks up globally.

Figure skating and ice dancing both fell in line with projections as well. Infostrada Sports correctly predicted that Meryl Davis and Charlie White of the U.S. would take gold in ice dancing.

They also were ahead of the curve on a few of the Games’ perceived upsets. Their numbers showed that Shaun White would fall short of gold in the snowboard halfpipe, which he did. Another set of data showed that Sven Kramer of the Netherlands would not win despite being a heavy favorite in the 10,000 meter speedskating event.

On the flipside, Gleave says there was "no way of predicting" Jorien ter Mors would spring a surprise in the women’s 1,500 meters as she lacked long-course experience.

And in the men’s downhill, champion Matthias Mayer of Austria had not placed highly in prior events and was a bit of an afterthought with American Bode Miller attracting much of the pre-race attention.

Gleave concedes the downhill lends itself to more random results.

"It’s a very difficult one."

Infostrada Sports was not alone in underestimating Russia’s dominance. Though their metrics already account for the boost a team receives by competing in a home Olympics, Russia even outperformed that.

Gleave pointed to the role a host team has in designing courses and the amount of access they have prior to the Games as reasons for success.

However, he pointed out that Infostrada Sports would be careful not to deviate too far from the established advantage for host countries in moving toward the next Games.

"Our feelings at the moment have not changed the way that we did it. If we’re going into Rio, we will not change the home advantage parameters from our London model, though we are sure to sharpen our predictions on other aspects."

Written by Nick Devlin

20 Years at #1: Your best source of news about the Olympics is AroundTheRings.com, for subscribers only.

Guardar