Germany's World Cup Loss Proves AI Isn't Perfect

Sports analysis might be greatly improving thanks to artificial intelligence data processing software, but that doesn't mean it's perfect.

The German researchers who predicted either Spain or Germany would win the world cup watched its secondary choice get eliminated from the World Cup during an engaging match against South Korea. 

A paper lead by Andreas Groll at the Technical University of Dortmund in Germany took machine learning and statistical analysis in order to pinpoint who would be victorious in the World Cup 2018 held in Russia.

[see-also]

As previously reported by Interesting Engineering, the researchers picked Spain as the most likely winner with a 17.8% probability of success and a 73% chance of reaching the quarter-finals. However, the researchers added that if Germany were to clear the group phase of the competition, its chances of reaching the quarter-finals would increase to 58%.  

Oddly enough, in the process of getting to the finals and claiming victory, the researchers put Germany beating Spain should they go head to head. That would've given Germany its second consecutive World Cup victory. The researchers explained this was because Germany would be more likely to beat Spain in the semifinal round, but the team has a harder road to get there and more chances of getting knocked out of contention.

Looks like at least that part of the researchers' analysis came true. Spain is still hanging strong in the quest for the cup (at least, as of this writing). 

Using Artificial Intelligence for Sports Predictions

Data analytics have always been an integral part of sports. Seasonal player stats can make or break a professional player's chances of higher salaries or switching teams. Horse race fans check the performance history of horses before waging bets. However, the rapid evolution of artificial intelligence has led to a fast-paced new world of seeing what group can produce the most accurate outcomes of events given the sports analytics. 

Data science company STATS LLC is one such group. The sports analytics company had been around for nearly four decades and leader Patrick Lucey said he's seen it evolve over time. 

"I feel very lucky to do what I do," said Lucey "I tell people what I do, and, you know, I feel very lucky to work in this domain and work with very talented people."

The German researchers might've had more success had they attempted to analyze and predict the specific movements of the players themselves, which is exactly what Lucey does using AI-based sports analytics.

"We have the ability now to -- using all of the tracking data that we have -- to kind of simulate what players will do in a certain situation, so that's really exciting," Lucey said.

These allow the STATS team to ask "how often does a team do this? How often does a player do this? What's the likelihood of a player making a shot in that situation? What if I switch that player with another player? What if I want to simulate how that team would react in that situation?"

While the German loss was devastating to both the researchers and to millions around the world, the use of machine learning to determine a winner is far from over. Plus, it can have a second crack at picking the correct winner --  in another four years.