At the conclusion of the Australian Open I wrote a post about the tournament. In a comment one of my friends raised questions about withdrawals due to injury in tennis and whether they are increasing and Nadal’s potential role in that. This made me very curious about trends in tennis withdrawals. Are retirements in men’s tennis increasing? Are the best players retiring more frequently?
I looked for stats on this and could not find them. I decided to compile my own. After some thought I decided to start my study with a 20 year sample size (1990-2009). I went through the results of every match played in every major (Australian Open, Wimbledon, French Open, U.S. Open) during that period.
Disclaimer: There is a chance that at some point during the hours of scouring old match results I made a transcription error, and my statistical analysis was done very quickly and is most definitely not publishable beyond the blog level.
Fasten your seatbelts, we’re about to embark on a wild ride through tennis stats.
We’ll start by getting on the same page with terminology. There are numerous ways to have a match end without a resolution on court. A walkover occurs when a player withdrawals before a match begins. A retirement occurs when a player decides to stop playing during a match due to injury or some other reason. A default occurs when a player is ejected from a match. A true withdrawal occurs when an invited player chooses not to play in a tournament before it begins. I was not able to find a way to effectively collect data on withdrawals. I tabulated retirements, walkovers, and defaults for the past 20 years. Walkovers and defaults are rare and almost perfectly evenly distributed over the last two decades, so I decided to use retirements as my main data set.
At times I will refer to seeded and unseeded players. For this discussion seeded players refers to players seeded #1 through #16. These are the elite players in the tournament, the upper 12.5% .
Distribution of Retirements
By my tabulation there were 322 retirements at tennis majors between 1990 and 2009. Seeded players accounted for 31 retirements (9.6%). At first glance this might seem to say that seeded players are less likely to retire, however a goodness of fit test reveals this statement cannot be made (p=0.11).
Statistically speaking retirements are equally distributed across the majors (Figure 1). The Australian Open has 83, Wimbledon has 62, the French Open has 87, and the U.S. Open has 90. This surprised me. I expected the hard court tournaments at the beginning and end of the season to have higher numbers.
Figure 1. Distribution of retirements in men's tennis majors (1990-2009).
The Leading Retirees
In the past 20 years the most frequent major retiree has been Wayne Ferreira. He has retired 7 times, completing a retirement career grand slam (retiring from all four majors). The only other player in the past 20 years to compile a retirement career grand slam is Andrei Pavel, who has only retired 5 times. The second highest number of major retirements in the past 20 years was accrued by Jerome Golmard with 6 (though he failed to retire in the French Open, shattering his grand slam hopes).
The data were very interesting. When I charted simple line graphs for each tournament it seemed like retirements were increasing, but it was difficult to tell exactly what was going on. Wimbledon and the French Open both had more notable increases, yet the numbers tended to fluctuate greatly year-to-year, resulting in a messy chart with high peaks and low valleys. The Australian Open and U.S. Open were very difficult to read.
Running regressions on the numbers over time yielded a surprise. I played around with various trend lines and finally settled on a linear set-up. Wimbledon was the only major with a significant increase in retirements. The French Open narrowly missed being significant, and both the Australian Open (Figure 2) and U.S. Open (Figure 3) had were far from significant.
Figure 2. Australian Open retirements over time (R2=0.09; p=0.20).
Figure 3. U.S. Open retirements over time (R2=0.003; p=0.81).
My next step was to combine all the data from the four majors to run a cumulative retirement regression (Figure 4).
Figure 4. Cumulative tennis major retirements over time (R2=0.47; p=0.001).
So it appears across all majors that an increase is happening in retirements. The two majors that are contributing the most to this are Wimbledon and the French Open.
This leads to another question: are seeded players more likely to retire now than they were in the past? I isolated the retirements of seeded players and ran a linear regression. The results indicate that there has not been an increase in seeded players retiring (R2=0.004; p=0.79).
My next move was to compare decades. I split the data into a 1990s group and a 2000s group. I then looked at retirements by decade, continuing to isolate seeded players. It turns out that retirements significantly increased among unseeded players in the 2000s (p=0.005), though there was not an increase in seeded players (Figure 5).
Figure 5. Tennis retirements by decade.
To answer the questions I started this study with:
Are retirements in men’s tennis increasing? Retirements in men’s tennis majors have increased over the past 20 years. The greatest increases have come at Wimbledon and the French Open.
Are the best players retiring more frequently? The increase in retirements is something that is occurring among unseeded players.
What is Nadal’s role? Nadal has only retired from one major (he also has withdrawn once). The current elite player most prone to retire is Novak Djokovic, who has retired from majors 4 times. Djokovic has greatly improved his fitness over the past two years, however, so hopefully major retirements are a thing of the past.
I suspect the long season and rise of the hard court are partially to blame for the increase in retirements. The best way for a player to improve his ranking (and paycheck) is to play a lot of tennis. Hard court tournaments have become more common than clay or grass tournaments here in the U.S. and numerous tournaments abroad that had been clay or grass based have switched to hard courts.
It is undeniable that a change in attitude toward injury has happened in the past decade. The players of the past must marvel at the frequent injury timeouts and mid-match sessions with trainers, both of which did not exist all that long ago.
One Last Parting Note of Interest
In the past 20 years there has been only one major in which every match of the tournament was played to completion on the court. That was the 1993 French Open, in which Sergi Bruguera won the title in a five set final over Jim Courier. The 1990 Wimbledon tournament was the only other retirement-free major in the past 20 years, but it featured a walkover in the second round.