Roger Federer and Novak Djokovic played out an epic five-set final at Wimbledon 2019, which will go down in folklore as one of the greatest matches ever. The Serb prevailed after winning all three of his sets via tiebreaks and despite winning fewer points than his opponent on the day. Most significantly, he did so after saving two match points on Federer's serve late in the fifth set.
The match was also the first in the history of Wimbledon to feature a deciding set tiebreak, after the fifth set reached a 12-12 stalemate.
Recently, researchers at Stanford University recreated the incredible final using Artificial Intelligence software, and came up with different simulations and possibilities based on the two players' styles and strengths. One of the simulations even showed what might have happened if Federer had hit his forehand a little differently on the first of his match points.
AI system recreates the match-point unforced error made by Roger Federer off Novak Djokovic's return
The system developed by the researchers creates stunningly realistic controllable graphics of tennis matches. And the example they used to illustrate the software was the Wimbledon final between Roger Federer and Novak Djokovic.
Roger Federer was in top form for most of the match, as he took the fight to the defending champion. The Swiss maestro was surprisingly the better returner during the final too, winning 15 more points against Djokovic's serve than the Serb could against Federer's.
But two poor tiebreak performances took the match into a deciding set, where Roger Federer has historically never had a lot of success against Novak Djokovic. Still, the Swiss was able to muster two breaks of serve in the fifth, and at 8-7 he even served for the match.
The 39-year-old had two match points at 40-15, but gave away the first through an unforced error on his forehand. The second match point was all Novak Djokovic though, as he ripped a brilliant passing shot to leave Federer flat-footed at the net.
The software developed at Stanford used the first match point to show how alternate possibilities for real-life exchanges can be visualized. In the video, Roger Federer can be seen hitting his inside-out forehand a foot or so to the left of where he actually did, which would have put the ball out of Novak Djokovic's reach but still inside the line.
Despite a few kinks sought out by keen eyes (like fans and officials never moving, no shadows under the ball and players and players appearing to float across the surface), the visuals are supremely realistic.
Stanford University researchers were able to develop the system by building a statistical model that calculated the responses of players in match-play situations.
The AI also gives the user full functionality of the match function. So you can completely control the movement of players, and decide the shot-placement on any area of the court.