Retirements Aren't Only About Standing the Heat

Heat was the main topic of discussion on Day 4 of the US Open. The brutal effects of sweltering temperatures and stifling humidity took centerstage during the afternoon play when American Jack Sock collapsed in the third set of his match against Ruben Bemelmans. Sock–who looked unconscious as the medical team on site rushed to his aid–was forced to retire and was subsequently diagnosed with heat illness.

Sock’s prognosis is good but that must be little consolation to a top competitor who was up two sets and had to quit because he was unlucky enough to get a 12:30 match time. Athletic competition isn’t a real contest if luck has too much of a role. This is what makes penalty shoot-outs in association football so frustrating.

Tournament directors could throw up their hands and say it’s the weather’s fault, but weather isn’t the only issue. Retirements and walkovers–the most readily available proxies of player fatigue or injury that we have for most matches–have been on the rise. From 1991 (the first years of the ATP World Tour) to 2014, the expected number of retirements or walkovers has increased from 4.4 to 6.9, which corresponds to a 5% probability of a retirement or walkover at today’s Grand Slams (Figure 1).

These are average, of course. Figure 1 shows that the data on retirement is quite noisy and, like the overall average, shows some evidence of getting noisier over time. So, although, there has been a systematic upward shift over time, the expected retirements and walkovers in any given year could be much higher or lower from the mean shown by the trend line in Figure 1. Since 2000, 3 years at the US Open (2002, 2011, and 2014) have had 10 or more retirements or walkovers. The increasing volatility combined with the rise in expected retirements and walkovers make this pattern even more of a concern.

What could be the cause of the upward trend in retirements and walkovers? Clearly weather conditions and player fitness have an important role, but–unless climate change has been particularly cruel to Flushing Meadows in September–improvements in player fitness would predict retirements and walkovers to have dropped over the course of the modern era. The fact that we observe the opposite suggests that there is more to the picture.

One striking parallel with the pattern in retirement and walkovers is the dramatic increase in match lengths on the ATP Tour. Figure 2 compares the average match time for completed matches at the US Open to the average in 1999, when public data on match duration was first regularly collected (Source: Jeff Sackmann). Over 15 years, the average match length has increased by 15 minutes. For the players that make it to the finals, that increase corresponds to nearly 2 extra hours of total match time in the tournament–no wonder god-like fitness like Novak Djokovic’s is needed to be at the top of today’s game.

The similarity in the rising trends in match duration and the frequency of retirement/walkovers suggests that retirements and walkovers might not be totally random. The modern baseline game is simply more physically demanding than the quicker-paced play of 15 or 20 years ago, and the additional wear and tear this implies has consequences. Yet tournaments have done very little that would suggest they recognize these trends. The US Open has arguably done more to curb these trends than the other Slams by introducing the 5th set tiebreak and moving the men’s final to Monday. But, Sock’s dramatic exit yesterday and 12 retirements overall before the Round of 16, are frustrating reminders that these efforts haven’t gone far enough.