By Alessandro Mastroluca and Gabriele Ferrara “When we are on site to watch a tennis match, we think to understand the game very well. The more I looked at the stats, the more I started to understand that our eyes are not the best way to evaluate a tennis match”.
Craig O’Shannessy, the newest entry in Novak Djokovic’s team, used his experience to change his approach as a tennis coach and create a peculiar match reporting style, working for the New York Timas and the ATP. Graduated in journalism, he worked for a year and a half in Australia, for the Border Mail in Albury, then he switched to focus for 20 years on his role as a tennis coach, working among others with Kevin Anderson, bringing him back to the top 50, Rajeev Ram, helping him to move back from No.270 into the top 100, Marsel Ilhan, Melinda Czink and Milos Raonic.
Now, he has the great chance to prove that also in tennis strength is in numbers. He started his personal and professional paradigm shift more than ten years ago when he began to use Dartfish, a an integrated solution across multiple platforms (software, online, mobile) that allows users to capture videos, tag events real-time, and upload them, that helped the Us Davis Cup team in 2011 and supported 400 athletes who won medals at the 2012 Olympic Games.
“In 2005 or 2006 I first announced the tagging module in Dartfish. Over the years, I’ve used it for the junior tennis players and for the professional players. I’ve used to record data from best players in the world, Roger, Novak, Andy, Rafa”.
From 2015, he began to extract data and stats produced by IBM during the Grand Slams. “Once I started doing more with data, I discovered there’s a different look”. The first discovery came as a sort of an upset.
“When I’ve started to consider tournaments as a whole, 70% of all rallies were in the 0-4 shots range. And I had no idea about, and I was involved in tennis for so many years”. As he wrote last year on The New York Times, during the 2015 Us Open “71 percent of the points in men’s singles matches and 66 percent of the points on the women’s side came on rallies of four shots or fewer.
Only 9 percent of men’s points and 11 percent of women’s points came on rallies of nine shots or more”. Even more relevantly, “the players who won more zero-to-four-shot rallies won the match 90 percent of the time in men’s singles and 83 percent of the time in women’s”.
Those numbers induce to rethink even to the 14 clashes between Novak Djokovic and Roger Federer from 2014 to 2016. During that period, Federer beat the Serb only once on clay, in Montecarlo, collecting his other wins on hardcourts at Dubai, Shanghai, Cincinnati and the ATP Finals where the Court Pace Index is generally higher.
So, it would seem that the main issue for the ageing Swiss was his performance in the longer rallies. Numbers, without concepts and visions, are nothing. This principle guided O’Shannesy’s activity as a coach, particularly on the practice court.
“Players were expecting to come to the practice court and do their usual work: side by side, thinking the more is better, the more I sweat the better. I saw it was very ineffective because you weren’t taking moments from the match and copy it into the practice court”.
So, he started to change his approach. “My practice court became very specific, segments of the match. It took away a lot of the grinding. It’s much better in practice what happens in a match, what it’s repeatable in a match.
Touches on the ball will be less but they will be practising something that could help them win matches. I showed them 10% of all points pass the double-digit and in junior tennis even less. So why would you spend 90% of your practice time developing something that happens 10% of the time in a match?”.
More pieces of information make decisions simpler. Numbers clearly show that the beginning of a rally becomes more relevant than its conclusion, and to use time more productively on the practice court, focusing on serve and return.
Something that could help a player like Lucas Pouille to bring his career and his evolution to the next step, given his percentage of flat serves in reaches 37,5 percent from the deuce side and 39 percent from the ad side.
And clearly, the Frenchman has reached a point in his career where only specific attention to tiny details could help him make the difference. Also for a player like him, serving down the T could seem an attractive option when the pressure mounts.
But numbers reveals it’s the worst choice. “Players that won more, serve out wide more,” he said. “The T is the smallest target to hit. It gives you more aces but it’s harder to hit. If you don’t look to data you could think that it’s better but it is not”.
Nadal used this approach to switch his second serve in 2017 and become the oldest No.1 in the year-end-ranking since 1973. “At ATP World Tour Masters 1000 events and the Nitto ATP Finals this season, Nadal hit 86 second serve out wide in the deuce court.
Last season, he attempted only four to the same location” he wrote on the ATP website. “Nadal’s commitment to hitting more wide second serves in the deuce court was well rewarded, as he had the highest win percentage (64 percent) there than in any other location”.
Besides, during the Us Open final, Nadal served 35 of 72 first deliveries out wide, to open up the court and give himself more chances to clinch the point with his first shot after the serve. It’s not a coincidence that Anderson won just 6 of the 43 rallies ended in less than 4 shots on Nadal’s serve.
The “serve + 1” pattern represents another telling element in a player’s strategy. A “serve + forehand” choice, as Roger Federer normally does, remains more effective than the “serve + backhand” approach.
And the same schemes can be applied to junior or lower ranked players. “This new methodology removes opinions and personal bias. Let’s look at the actual reality, the irrefutable evidence. You can’t argue. Use it as a guide to understanding what matters in winning matches and what can happen on a practice court”.
Players would win more matches and follow. So, why is tennis struggling to follow the same path of team sports? Soccer fans, reporters, pundits and players are familiar with stats and analysis by Opta. NBA.com, even more tellingly, has a partnership with SAP, the same provider of the WTA.
But the level of detail is not comparable. NBA offers to the average reader metrics like the True Shooting Percentage, that factors in a player’s performance at the free-throw line and considers their efficiency on all types of shots using the formula Pts/(2*(FGA + (.44*FTA)) where field goal attempts are taken integrally and free throws attempts are related to the .44 multiplier because research has determined that about 44% of all free throws take up possessions.
But the possibilities are almost neverending. You can check the Usage Rate, the number of possessions every 40 minutes, the number of touches for every single player in a match, the number of points realized and conceded every 100 possessions.
“When you look at team sports, they have a lot of manpower a lot of people involved and a lot of money to invest in technology,” O’Shannessy said. “In tennis, it’s individual, people are on their own.
It’s naturally easier for team sport to get involved. Other team sports have just one organizing body, too. In tennis, we have ATP, WTA and Grand Slams whose data are made by three different providers, Infosys, SAP and IBM that don’t always cooperate and don’t share things together.
Tennis is far more fragmented. The access to that data is difficult”. Tennis could and should do much more, also trying to define when a player switches from a primary to a secondary pattern. Often, remaining confident in his own gameplan looks like an underrated quality.
During the legendary Wimbledon final in 2008, as he wrote in his autobiography, despite the psychological ups and down Nadal continued to repeat himself to attack Federer’s backhand. In different situations, when they’re in control, the player can choose to change parts of their main strategy to upset their opponents.
Define how and when they do it, would help to understand why they have won or lost. Tennis, O’Shannessy concludes, “should give better and clearer data to understand what really happens in a match”.
A first step could be counting the amounts won and lost in the different rally ranges (0-4 shots, 5-9 shots, ten or more shots) and the forced errors, “the No.1 way a rally ends” in modern tennis. Then, we could add, changing the way the breakpoint opportunities are presented and relate them not to the total chances but to the number of games when a player has had those chances during the match.
In this way, fans, coaches and players could improve their decisions and their comprehension of the game. Because eyes are not enough. .