Most tennis players spend hours on the practice courts. And for good reason - tennis is darn tough. The question is, are these hours on court productive hours or redundant? How can we know? To assess whether our training is effective (and that it'll transfer to matchplay) we must first understand the demands of elite tennis.
In this post, we’ll review a study by Pereira et al (2016) that dives into the movement details of professional tennis. Other studies have previously analyzed movement characteristics; but, those studies replaced tournament matches with simulated matchplay. The present study observed movement characteristics via official ITF sanctioned matches.
Why Should we Know the Mechanical Demands of Tennis?
Most research to date has focused on physiological responses to tennis play (heart rate, blood lactate, oxygen uptake) and activity profiles (rally lengths, work to rest periods etc). What’s been lacking however, is specific mechanical characteristics of tennis movement - like running velocities, the direction of movements and distances covered by each player. More specifically, past research hasn’t done a very good job of analyzing these variables during official matches. None of us would argue that competition is more demanding than practice matches from both a physical and cognitive perspective. With advancements in technology, researchers are better able to use video software to analyze specific player variables - position, trajectory of movements, kinematics and so on. This information is extremely valuable and can influence the organization of training for elite players.
Study Purpose & Hypothesis
To analyze the physical and technical demands of official tennis matches and to investigate whether there are decrements in physical and technical performance during competitive matches. Also, to analyze the physical demands between players who are serving and those that are returning.
The authors hypothesized that tennis players would present decreases in physical and technical performance over the course of an official match. Additionally, they expected players to cover greater distances when returning the ball, compared to when serving.
Eight professional tennis players, over 4 matches on clay courts, were observed in this descriptive analysis. Before the start of the study, two expert coaches concluded that all players were baseliners. Two 30 Hz digital cameras were positioned around the court and recorded all 4 matches - the recordings were transferred to a computer and synchronized between cameras.
Player movement trajectories were gathered via a tracking system. Players were segmented from other aspects of the court into what the researchers called ‘blobs’, using a video software system. The players tracking was then found via foot positions (this was determined by finding the maximum y-coordinates of the blobs and their associated middle point in the x-coordinate (figure 1). Prior to matches, a coordinate system of the court was determined so that player coordinates could be identified and then measured.
A single operator used the a video software system to watch all 4 matches. When a technical action was performed, the operator identified it with a computer mouse by clicking on one of the following - serve, forehand, backhand, volley, slice, smash. The operator also identified which player had performed the action and whether it was a ‘correct performance’ (ball lands in the court) or an “incorrect performance” (an error).
Variables Analyzed in the Study
Distance Covered by Each Player:
when serving vs when returning
Player Movement Velocities & Percentages in Each Zone:
The athletes covered approximately 5.2m per rally when serving and about 7.7m when returning - this confirmed the authors hypothesis that servers cover less court than returners. Players performed a total average of 92.5 and 59.5 lateral displacements & 25 and 15 forward displacements during the first and second sets, respectively. No differences were found between rally durations between the first (4.6 seconds) and second (4.3 seconds) sets. Table 1 also presents the technical performances of the athletes during the first and the second sets of the matches. A significant decrease was found only for the forehand ground strokes - this contradicts the researchers hypothesis as they believed all technical strokes would decrease from set 1 to set 2. Lastly, table 2 highlights the various velocity ranges and the percentage of time spent in each range.
The authors made 3 conclusions with respect to their findings: (1) professional tennis players did not present decreased physical performance during the match; (2) when the player was serving, the distances covered during the rally were shorter compared to the returning player distances and (3) players maintained technical efficacy during the match, with the exception of forehand ground stroke performances, which decreased from the first to the second set. Let's dive a little deeper into the discussion.
Average rally durations (4.6s in 1st set & 4.3s in 2nd set) were shorter than those previously reported - 6.3 ± 4.1s for advanced and 7.6 ± 5.5s in recreational players (Fernandez-Fernandez et al 2009, Hoppe et al 2014). The authors believe this is due to the differences in levels of play within these studies. This is likely true because of differing playing styles between pros and amateurs. Generally, pros have more aggressive game styles. The authors suggest that training should reflect this with greater emphasis on creatine phosphate utilization (less than 10 seconds of work) along with a focus on muscular power training INSTEAD OF lactate-producing and endurance drills.
Lateral vs Forward Movement
The authors observed that lateral displacements occurred more frequently (79% of the time) vs forward displacements (21% of the time). That said, when moving forwards, players ran longer distances compared to when moving laterally. Even when moving forwards, the distances in tennis aren’t very long which means that shorter, more explosive plyometrics should be employed when training movement and change-of-direction qualities specific to tennis. Unfortunately, no backward displacements were reported. Either the authors included backward movements as part of forward/horizontal displacements OR they simply didn’t measure them. Either way, this was not mentioned. I hypothesize that more forward movements occur in tennis (and with greater distances) BUT we should not discredit the importance of backward movements. They do occur very frequently in tennis rallies, especially when playing opponents who hit higher/with more spin (think Rafa on clay). Furthermore, from a mechanical perspective, moving backwards presents greater challenges for players and should likewise be targeted in off-court training sessions.
Serving vs Returning
This is the first study to analyze the different movement characteristics between serving and returning. As hypothesized, servers moved less than returners. The authors attribute this to the commanding nature of the serve, which allows players to have a greater chance of controlling the point from the onset. From a training perspective, it’s advisable that players spend more time practicing first serves, as well as the first and second shots after the serve. Traditionally, serves are left to the end of practices, however, dedicated serving sessions may be more appropriate. This will allow for the diversity of serves to be trained (flat, spin, slice, wide, body etc) along with specific serve patterns (ex: wide serve followed by aggressive forehand into the open court).
Other Movement Variables
The majority of time (79%), players running velocities were between 0 and 7 km/h (table 2). The authors reiterate that this does NOT mean tennis is a low intensity sport. Rather, tennis is a sport characterized by a number of decelerations, accelerations and an overall high amount of changes of direction. Recall from a previous post, changing direction has 3 phases - braking, planting and propulsion. These 3 phases have specific muscle contraction properties that are best trained with maximum strength training exercises for the lower body.
No differences were found between the physical performance variables (rally durations, distances covered, running velocities etc) between sets. The authors propose that because of shorter rally durations and recovery times between points, games and sets, players have an adequate amount of time for restoration of both neuromuscular and physiological parameters. However, keep in mind that only the 2 first sets of each match were recorded. The results for 3-set matches may affect fatigue and consequently player movement characteristics to a greater degree. Future analyses should address this.
There were some differences between correct and incorrect performances between sets. More specifically, players made significantly more errors on the forehand side in the second set versus the first set. It is difficult to attribute this decline to fatigue as both mental and tactical errors could influence results. Furthermore, players may perceive greater fatigue (because of time spent on court, the scoreline etc) when in fact the physical variables indicate otherwise - second vs first set rally lengths were actually slightly shorter. Also, all other technical shots (backhand, serve, volley, slice and smash) did not show any significant differences between sets. The authors suggest that this may be due to players decreasing the speed at which they hit these shots but this cannot be confirmed as ball speeds were not measured.
Shorter, more explosive type drills should be implemented more often in the training programs of elite players - both on and off the tennis court.
In terms of conditioning for tennis, work capacity that reflects rally lengths and distances covered (along with appropriate rest intervals) should be the main focus of training for this quality. This is obviously developed in matchplay settings (which provides further importance to competition) but can also be emphasized off the court with power endurance circuits.
More serving practice should be performed - along with associated serving patterns - as players do less physical work when serving vs returning. This is primarily due to the aggressive game styles of pro tennis.
Although most stats point to shorter rally lengths in tennis, these are still averages. Some players - even at the pro level - may play a higher percentage of points in longer rallies (5 or more). It’s important to know your player and how they fare in each specific rally length. This info could have significant implications on the organization of practices and off-court training regimes.
More studies from official matches still need to be conducted and perhaps comparisons between 2 set matches, 3 set matches and even best of 5 set matches. Furthermore, different court surfaces, balls, climates and so on, may also impact the movement characteristics of elite tennis. These parameters should be considered in future studies.