Does Interrupting a Match Impact the Outcome?
The intermittent rain and late nights at the 2018 French Open have resulted in multiple interrupted matches, including the bottom half of the men’s quarterfinals. We often have the impression that reversals and big swings are more likely with an interrupted match. Does the data back up this impression of the impact of suspended play?
Rain has been a foe at Roland Garros for several years. 2016 was one of the most challenging years for the tournament in recent memory, with entire days being completely washed out. Although play has been able to proceed on all days in 2018, the smattering of sprinkles throughout have resulted in a high number of suspended matches, 8 men’s singles matches and 2 women’s singles matches so far.
Historical data on suspensions is hard to come by. But the number of suspensions in 2018 at the French Open is more than the interrupted matches there between 2014 and 2017 combined.
With the rise in suspended matches, especially on the men’s side, this year, it is natural to ask what impact the interruptions might have had on how the event has unfolded? and who has been able to advance?
Suspended Matches Are Long
Looking at the stats on the 12 men’s matches that finished after suspension (excluding any ending in retirement) at the French Open between 2016 and 2018, the first thing that stands out is there duration. Compared to men’s matches in the same period that were not interrupted, the majority of suspended matches fall in the upper 50% of match durations.
The typical continuously played men’s match at the French Open in the past 3 years has lasted 200 points. For suspended matches, the typical points played is 260.
Whether suspension in itself belabours play is less clear. The more likely explanation is that matches that would have gone long no matter what are the very matches with the highest change of getting interrupted.
The storyline you often here when a match is interrupted is that we can expect a shift in momentum to come. Tennis commentators will often discourage the player in the lead from stopping play if there is any way for it to continue, the implication being that they are more likely to lose their advantage if play is completed the following day.
But is the change after a suspension more likely to bigger than the change after a similar amount play in a regular match? Rafael Nadal’s reversal over Diego Schwartzman yesterday would suggest so, but maybe that would have happened whether there was a suspension or not.
As most suspensions have come around 100 points into a men’s match, we can compare the stats pre and before the 100 point mark in regular matches to get an idea about unusual swings in suspended. Below, we look at reversals in the leading player, that is, the player who was ahead in points won compared to the player that actual won.
In matches not suspended, this kind of reversal has happened in an average of 23% of men’s singles matches between 2016 and 2018 at Roland Garros. For the suspended matches, that average is 31%, a 30% increase. Now, there is a lot of uncertainty around that average given the small number of interrupted matches, but it is suggestive of an uptick in lead reversals.
The eventual winners of suspended matches had an average increase of 5 percentage points in points won after the interruption compared to before. This is a greater change than 75% of winners of continuously played matches before and after a similar amount of points played.
Nadal and Benneteau were two players who showed the biggest gains in points won after a suspension this year. Chardy, on the other hand, was one of the few who saw the opposite effect, though he had put enough work in before the suspension to still get the win. Both Nadal and Benneteau were 1-0 down in sets when suspended, while Chardy was 2-0 up in sets.
It does seem that there is evidence to back up our sense that suspensions make an impact on match outcomes and how players perform. If true, it is one more reason we can look forward to a new Chatrier with a roof in the years to come.