28 September 2014

Nick Kygrios caught the world’s attention in July at Wimbledon 2014 when he beat world number 1 Rafael Nadal. After the match McEnroe commented:

“We’ve been waiting for this for a while. We keep saying, `Who’s the next guy?’, and I think we found that guy right now.” http://dailym.ai/1jlEfrc.

but Nadal seemed to beg to differ:

“He has things, positive things, to be a good player. But everything is a little bit easier when you are arriving.”

I scraped the data from the Wimbledon web site, for the first two rounds of the tournament to compare Nick’s play with the other competitors. Making plots of these statistics we can learn that Nick really was pushing his game, or maybe he always pushes his game, making an extremely high number of unforced errors, but balancing this with a large number of winners. His statistics on first serve, receiving points and net points were comparable to the players who made it to the semi-finals. His second serve percentage was good but on the low side compared to the semi-finalists.

The relationship between the statistics and how far the competitors make it through the tournament is not simple. For example you might think that the more winners a player hits the better the chance that they win a match. This isn’t true, and a wonderful example is Bethanie Mattek-Sands, who hit 81 winners in her 2012 Australian Open first round match, 30 more than any other player, and lost the match. It wasn’t her that unforced errors were disproportionately high either, they were on the right side of the winners at 65. Players typically need to keep their statistics in a range, not too high and not too low, with a few exceptions. Hitting winners is important, but only up to a point, and it helps if there are more winners than unforced errors. Simply modeling the statistics against progression through the tournament doesn’t work well. Making plots we can see the relationships, but we can also imagine relationships when they don’t actually exist.

Making plots of the data in the context of plots where there is no relationship, using permutations, can help assess the importance of the relationship. See Chance column for more information on how this was done for Nick Kygrios’ statistics.

Since that article was written we have had the luxury of seeing Nick in his next grand slam tournament, the 2014 US Open. He had a tough draw, Youzhny in the first round, Seppi in the second, and Robredo in the third. His statistics were up there again, comparable with the best. But his main problem right now is the wildness: against Robredo, as it happened against Raonic in Wimbledon, playing against a steady opponent willing to wait for their chances to strike back, toppled his game. He has the physical game, but his mental game is not at its necessary level to be a champion yet.