For more than a decade, the conversation around US Tennis has been that of a failure by the organization to properly develop and train promising players in the pipeline. The 2017 US Open, where the final four women standing were all American (Venus Williams, Coco Vandeweghe, eventual champion Sloane Stephens, and Madison Keys) alleviated some of the criticisms of the USTA and their inability to procure another capital-S Star from the deep bench of players throughout the US, but it also puts pressure on the association to continue that momentum and continue to prove American tennis can consistently find and put forth talented, entertaining players who can win majors.
There will always be players like Donald Young who never became the Next Big American Player despite training with the USTA and its coaches, or Sloane Stephens, who doesn’t completely align her training with the organization’s coaches and training centers yet still wins a US Open title. The importance for the US is to find the players in between those two poles — those who want the help the USTA is offering them who also have the raw talent and skill to reach the same heights as a player like Stephens.
The question looming over American tennis isn’t a new one: what happens Serena Williams retires?
In the face of that concern, the USTA is making a big a move in the hopes of securing a future at the top of the game for many American players to come. They’re using an exclusive partnership with IBM to start treating player development like a business venture, not just a sporting one.
With the help longtime partner IBM, US Tennis is finally stepping up its game when it comes to the implementation of advanced stats and data tracking of all of its players. It’s 2018, so it’s not like players and coaches from all countries have been completely ignoring statistical help before now. But by increasing the efficiency of training and freeing up human capital throughout the organization, US Tennis hopes it will give the US a competitive advantage on and off the court that will launch even more players into the Top 10.
While using 1980’s Wall Street-esque terms when it comes to player development sounds callous on some level, this is also something the USTA should have done years ago. Other sports (the NBA tops amongst them) are leaps and bounds ahead when it comes to having an overall system to track and organize advanced stats.
It’s 2018, so it’s not like tennis players, coaches, and organizations haven’t been using serve analysis or court position tracking before now. For example, certain WTA tournaments (not Slams) allow on-court coaching and players and their coaches can use tour-approved devices to look at statistics and scouting reports during a match. Some players may simply pull out their own tablet from their bag and re-visit a report prepared for them in advance of the match, others may talk through adjustments with their coach with the help of additional statistical rundowns.
The stats have been there, and coaches have incorporated them into training regimens and match strategy before, but US Tennis is now hoping that exclusive access to certain metrics can give their players even more of a leg up, especially now that more coaches are embracing the use of analytics than ever before.
The AI technology that IBM has used to make the US Open more accessible to fans for years now — via the SlamTracker scoring application, match insights used during the broadcast, and AI highlight sorting that makes it easier for the tournament to identify the most exciting moments to clip and push to its site — will now also be used to more efficiently catalogue player matches, easily collect stats like serve and return habits, and identify new statistics that aren’t currently in the rotation at all.
Stats tracked in this system include simple things like forehand winners, backhand winners, break points, and points that lead to break points called pressure points but can also track more advanced data like serving and returning habits or serve-plus and return-plus numbers — which can help players have a plan of attack for the first four shots of a rally, where 65-70% of points are decided. IBM’s technology can also tag videos from matches with lower quality match video than human taggers would be able to do, allowing video scouting reports to be built from tournaments outside the Slams and Masters 1000 events.
All of these advanced stats could be collected by hand before now and had been, but between the manpower it took and the time logged doing it, the benefits for players (especially during a tournament rather than just during training) weren’t nearly at the level they could reach now that IBM’s technology is involved.
Martin Blackman, GM of USTA Player Development, described the main upside of the partnership as freeing the USTA from the financial and manpower restrictions they hit when trying to support so many players at once.
For example, before the partnership, tagging a 90-minute match would take approximately two hours. IBM’s technology does the same thing in a few minutes, and can then export that data in a clearly organized set that can be analyzed by coaches and staff. While it’s objectively more convenient, it also has the added benefit of allowing those staffers to spend their time improving other areas of the development chain or interacting with players more than before.
USTA employees who would be otherwise tied up tagging matches for hours on end can commit those hours to working with junior players who need extra attention at a time when techni.
A player on the road can receive the same feedback on their match or scouting report on an opponent in one hour rather than a dozen — and when you consider the tight turnarounds during tournaments where players have matches every day, knowing an opponents’ tendencies on break points or at the net a day before the match rather than the morning of could be all the difference between a win and a loss.
Madison Keys, 14th in the world right now, is embracing this partnership at the US Open after being a finalist last year. Keys could come right off the court, do her press availability, and immediately sit down with her coach to break down all the things she did right and wrong before the match leaves her mind. And thus have a full day to wrap her mind around what she has to fix and what her next opponent will try to take advantage of in her game.
Coaches can pinpoint what data they want to focus on and receive “playlists” of those aspects of a certain players game to show them, all done automatically by the system after a match is processed. In turn, when a player fixes an issue and executes something well on the court, they can then look at a report that zeroes in on all the times they succeeded in taking advantage of an opponent’s cross court shots or drastically improved their serve toss location.
There’s also an opportunity for the system to grow and evolve as needed, eventually incorporating player wearables, nutrition, or sleep and training habits into the existing statistical models. Crucially, it points the USTA towards the areas they should be paying attention to before actually providing the answers.
Blackman admits that the organization wasn’t at the level they could have been before IBM got involved.
It’s made us raise our game in player development … We really have to make sure that we’re at the top of our game with knowing what we’re asking, why we’re asking it, being very fact driving and objective in our approach to this.
It’s not just top players who will receive these benefits, a key feature of this new system. Players as young as 11 will be analyzed using the technology, helping local communities identify players who could be true talents despite those players not having the resources to put themselves in front of USTA scouts at junior tournaments or go to camps that can elevate their skills.
If we accept that local courts and community organizations are some of the least mined development opportunities and that USTA is missing out on talented players because they don’t have the resources to scout everywhere all the time, local coaches being able to upload video of their players to this new program to be analyzed and elevated accordingly could be a real game changer. Not every family can afford to put kids in private classes or travel to tournaments, and limited financial resources shouldn’t automatically mean that player never gets a shot at a bigger stage.
If a local coach uploads a video of a young player to the USTA’s servers and IBM’s technology identifies them as a promising talent, the USTA can then invited them to an upcoming camp to be further evaluated. At that camp, they will then be recorded and USTA coaches can provide feedback for them to work on before the next camp based on video breakdowns of their serve, forehand, and backhand. A simplified version of what is happening at the top levels of the sport.
Once a player graduates to the junior tour, they will then have some of their matches tagged and analyzed in a similar way to the pros and be able to receive more in-depth feedback on things like court positioning or return competency.
With any luck this partnership will recalibrate the USTA development system back towards a more democratized system of funneling players to the top, with players who might not have even gotten a look or been invited to a camp in the first place now given a chance to make it to the junior tour and beyond.
The human element of training isn’t completely gone though, which may actually be the most important part of this new strategy. The USTA has finally realized that choosing between analytics and human competency in the development pipeline isn’t a choice at all, it’s a handicap to the system.
Since there is a lot of information to sort through after any given training session or match, players don’t receive any of it directly. It all goes to their coaches, who then sort and prioritize what they want to focus on with their player on that specific day. Madison Keys, a key player taking advantage of the new technology, may have nine or ten parts of her game that IBM identifies as possible improvement areas, but her coach can look at the report and only show her problems with her court movement because it will be the most important thing to fix before a match against a player who sprays the court with shots.
A personalized touch to scouting and coaching remains just as important as the information available to each player. Because otherwise why have coaches around at all? Blackman says the USTA has already seen a noticeable increase in players and coaches sitting with their computers post-match to absorb the advanced analysis of a match and immediately use it to tweak a player’s game.
After years of flailing for solutions to their ongoing development struggles, the USTA has finally found something that will enable them to consistently provide their players with more information, and faster, than other countries’ players will have, while still enabling each individual to use it in their own way as often or infrequently as they’d like.
There’s one thing this system won’t be able to do though, and that’s provide American players a competitive advantage against their compatriots. Should an American play an American at any tournament, no reports will be provided for either player on their opponent. So right now, with Madison Keys’ next opponent the American up-and-comer Bernarda Pera, neither player will get an oppositional report. Pera and Keys play each other well, so in this case the lack of a personalized report could make all the difference in who has the upper hand.
Maybe it’s good to leave some blind spots in the process, even as scouting and training improves. Tennis’ imperfections are a feature not a bug, but USTA’s new approach to data could help smooth out some rough edges and lead its players to the trophy ceremony more often than before.
Be the first to comment