Novak Djokovic advanced to the final of the Miami Masters today when Kei Nishikori withdrew from the event due to injury. Oddly, it was the second match at the Sony Open that Djokovic didn’t have to play, as Florian Mayer pulled out before their scheduled third round match.
It’s a rare occurrence in professional tennis–so rare that it had only happened once since 1968, when several players benefited from multiple walkovers at the French Open. In Miami two years ago, Andy Murray also skipped his third round and semifinal matches, as both Milos Raonic and Rafael Nadal dropped out due to injury.
The fact that it was Djokovic who got the free pass immediately gave rise to all sorts of speculation. Will the lack of match play hurt the Serbian? Does Novak get more walkovers than most? Are opponents more likely to withdraw if they’re facing a top player?
Let’s take these questions in order. I addressed a similar issue a couple of years ago in this post. Walkovers are rare, but the available evidence suggests that there’s no positive or negative effect from winning via withdrawal. A player’s chances of winning his next match are roughly what they would’ve been anyway.
Djokovic does gain from walkovers more often than the average player, but he’s far from the top of the list. Opponents have withdrawn five times in his 695 matches, good for 0.7%, roughly the same rate as opponents of Murray, Nadal, Roger Federer … and Donald Young and Dmitry Tursunov. Jo Wilfried Tsonga has benefited from six walkovers in 432 matches, a 1.3% rate, highest among tour veterans.
Top players win by walkover more often than others–but as we’ll see in a moment, it isn’t because they are top players. It’s intuitive to figure that mildly injured players are more likely to take the court if they think they have a better chance of winning, but the evidence suggests there’s little, if any, effect.
Men ranked in the top five win by walkover 0.6% of the time, while those in the next five get free passes 0.3% of the time, and most of the rest of the pack benefits at the tour average rate of 0.2%–once every 500 matches. (All of these aggregate rates are based on tour-level matches from 1991 through 2014 Indian Wells.)
For the most part, top players get walkovers because they hang around until the late rounds of tournaments. Walkovers occur at the highest rate in the quarterfinals of events, when 1.1% of matches end before they begin. Round-of-16 contests are almost as bad, at 1.0%, and semifinals are also considerably more walkover-prone than average, at 0.6%.
When we take these dangerous middle rounds out of the equation, the number of walkovers shrinks, as does the difference between top players and the rest of the pack. Less than 0.15% of pre-R16 matches end in walkover, and the rate at which top-five players benefit from them falls to 0.4%. That’s still more frequent than the rate for the rest of the field, but keep in mind the tiny numbers we’re dealing with here. It’s 13 walkovers in over 3000 matches. Take away five of those withdrawals–roughly two per decade–and the top five would benefit at the same rate as players ranked 16-20.
It’s not as interesting a narrative, but it appears that players usually withdraw when they are too injured to compete, and that’s most likely to happen midway through a tournament. The highest-ranked players benefit–because of their previous success on the court, not their intimidating influence off of it.
As always, a sober analysis. I will say that as a fan, and a non-mathematician, I habitually assume that a walkover says something about the physical state of the player who retired and not much else. I do think it’s worth noting, as you do, that top players benefit just a tad more simply because they tend to go deep enough for other players to start retiring. In this case what’s most interesting is that Nishokori, if he is to become a “top player,” will have to transition from the fellow who retires because his body got banged up, to the fellow who doesn’t retire and benefits from others doing so. Ironic, I guess?
a similar analysis for retirements (RET) would be interesting. and also if RETs are “used” more often, if players play against a better ranked player, etc.
thx for the w.o. analysis though.