At the October 18th Masters 1000 Shanghai final, Jo-Wilfried Tsonga stood across the net from Novak Djokovic and was easily dismissed 6-2 6-4. This lopsided score was routine over the Asian swing for Djokovic. In the Beijing and Shanghai tournaments, he average surrendering only 4.3 games/ match and easily controlled both tournaments. This domination isn’t just contained to the Asian swing of the ATP tour, but has been continuing over the past 12 months where Djokovic record is 83-5. He has won 3 of the 4 grand slam events, 6 of the 9 Masters 1000 (didn’t play Madrid and made it to the finals of the other 2), as well as won the 2014 ATP World Tour finals. If you just include those top tier events, he currently holds 15,900 of the 18,500 points he could possibly have. True domination, but can we draw any conclusions on how a few have beat Djokovic?
Tennis is a simple game in that it matters how a player serves and returns. Hold serve more often than your opponent will result in a better chance to win. Along those lines, if we look at Djokovic’s service hold percentage in every match against his return of serve break percentage, it is easy to infer a line of demarcation between a win and a loss. This invisible line is trivial, but the distribution of data demonstrates some important features. First, Djokovic is an extremely gifted on the return of serve breaking the opponents serve more than 25% of the time in 74% of his matches. For example, against Nadal, Djokovic broke Nadal’s serve 44%, 50%, and 56% of the time in their matches over the past 12 months. Another clear point is that Djokovic is susceptible to being broken himself, being broken at least once in 65% of his matches.
But to draw more conclusions, it is useful to understand these service hold and break % against the skill of the opponents. For the skill for the opponents, I use the Riles Clubhouse serve and return efficiencies and plot them against the corresponding serve or return percentage in the match against Djokovic.
As expected, Djokovic’s serve hold percentage becomes lower as the opponents return rating increases. The losses, although generally on the right hand side of this chart, do include Ivo Karlovic on the far left hand side. So it is inconclusive that the opponent return rating is a major factor in being able to beat Djokovic.
The flip-side does provide a little more insight. Djokovic’s return break % has a stronger correlation to the serve efficiency of his opponent. Also, all of Djokovic’s losses are clustered in the bottom right of this chart meaning he did not break many service games while playing a strong service player.
Therefore, Djokovic will most likely have difficulty with the top servers like Federer, Wawrinka, and Karlovic when they can consistently hold serve against Djokovic. This is poorer news for players like Nadal, David Ferrer and Richard Gasquet who are better returners. But it is clear that the way to beat Djokovic is to hold your own serve (>80% of the time results in a 28% chance of winning and >90% of the time results in a 67% chance of winning).
Even for a top server like Federer, it is not always easy to hold serve. Federer’s 5 matches against Djokovic in the last 12 months are accentuated on the chart below. Federer only 2 wins were when he didn’t allow Djokovic to break him once.
So as top players line up across from Djokovic at the Masters 1000 in Paris and the ATP finals in London, they should be extra focused on holding their own serve. But it is easier said than done, just ask Djokovic’s latest victim, Jo-Wilfried Tsonga.