Performance demands of professional male tennis players

Abstract

To quantify the performance demands in professional male tennis.
Games from three grand slam tournaments were analysed by an elite tennis player from video recordings. Game related data were collected on 22 players (French Open, 8 (186 games); Wimbledon, 11 (206 games); US Open, 9 (224 games)). Total number of strokes per game was quantified separately for service and return games. Strokes were categorised by type and designated as forehand or backhand. Differences in the types of strokes in a game were analysed using one factor (type of stroke) repeated measures analysis of variance. Differences in total strokes and stroke distributions between playing surfaces were analysed by analysis of variance (surface type) with Tukey's post hoc pairwise comparisons.
For service games there were more serves per game than any other type of stroke (p<0.001), with topspin forehand and topspin backhand the only other strokes averaging more than one per service game. For return games there were more forehand and backhand returns and topspin forehands and backhands than other types of stroke (p<0.01). Total number of strokes per game was greater in the French Open than Wimbledon (p<0.01), with more topspin forehands (p<0.01) and more topspin backhands (p<0.01). Total strokes per game in the US Open were not different from the other two tournaments.
The serve was the predominant stroke accounting for 45% (French Open) to 60% (Wimbledon) of strokes during service games. The greater number of strokes per game on clay v grass may contribute to earlier fatigue.

Full text

ORIGINAL ARTICLE
Performance demands of professional male tennis players
C D Johnson, M P McHugh
...............................................................................................................................
See end of article for
authors’ affiliations
.......................
Correspondence to:
Christopher Darren
Johnson, Nicholas Institute
of Sports Medicine and
Athletic Trauma, Lennox
Hill Hospital, 100 East
77th Street, New York,
NY10021, USA; critter@
nismat.org
Accepted 11 October 2005
.......................
Br J Sports Med 2006;40:696–699. doi: 10.1136/bjsm.2005.021253
Objective: To quantify the performance demands in professional male tennis.
Methods: Games from three grand slam tournaments were analysed by an elite tennis player from video
recordings. Game related data were collected on 22 players (French Open, 8 (186 games); Wimbledon,
11 (206 games); US Open, 9 (224 games)). Total number of strokes per game was quantified separately
for service and return games. Strokes were categorised by type and designated as forehand or backhand.
Differences in the types of strokes in a game were analysed using one factor (type of stroke) repeated
measures analysis of variance. Differences in total strokes and stroke distributions between playing
surfaces were analysed by analysis of variance (surface type) with Tukey’s post hoc pairwise comparisons.
Results: For service games there were more serves per game than any other type of stroke (p,0.001), with
topspin forehand and topspin backhand the only other strokes averaging more than one per service game.
For return games there were more forehand and backhand returns and topspin forehands and backhands
than other types of stroke (p,0.01). Total number of strokes per game was greater in the French Open
than Wimbledon (p,0.01), with more topspin forehands (p,0.01) and more topspin backhands
(p,0.01). Total strokes per game in the US Open were not different from the other two tournaments.
Conclusions: The serve was the predominant stroke accounting for 45% (French Open) to 60%
(Wimbledon) of strokes during service games. The greater number of strokes per game on clay vgrass
may contribute to earlier fatigue.
Professional tennis is a year round sport with a different
tournament or competition every week. Most injuries in
this population of athletes involve the shoulder and are
secondary to overuse.
1–3
It has been reported that over 50% of
world class players experience shoulder symptoms during
their career and 80% of theses cases stem from overuse.
34
The
areas of the shoulder most commonly affected include one or
more of the following: the rotator cuff, biceps tendon,
scapular region, glenohumeral ligaments, and the glenoid
labrum.
5
As the overhand racquet motion subjects the
shoulder girdle complex to similar stresses as those seen in
throwing, injury patterns and glenohumeral internal rotation
deficits among elite tennis players are similar to those of
professional baseball pitchers.
67
In contrast to baseball,
where various pitch statistics are maintained for all pitchers,
no such statistics are maintained for tennis players.
Additionally the effect of different tennis playing surfaces
on the number of strokes and stroke selection is not known.
Our aim in this study was therefore to determine the
performance demands of professional male tennis by
documenting the number and type of strokes during
professional tennis matches on different surfaces. While
several papers have been written about tennis related
injuries,
1–13
this is the first one to our knowledge that has
attempted to quantify the performance demands of the sport
among world class players.
METHODS
Games from three grand slam tournaments during the 2003
season were analysed (French Open, Wimbledon, and US
Open) by an elite tennis player from video recordings. These
tournaments were selected on the basis of the differences in
playing surface (French Open, clay; Wimbledon, grass; US
Open, hard). Game related data were collected on a 22
different male players (eight in the French Open, 11 at
Wimbledon, and nine in the US Open) with three players
analysed in all three tournaments and three players analysed
in two of the three tournaments. Games were analysed
separately for service and return games. The total number of
games analysed for each tournament was 186 for the French
Open, 206 for Wimbledon, and 224 for the US Open. An equal
number of service and return games was analysed for each
tournament. The total number of strokes per game was
quantified separately for service and return games. Strokes
were categorised as serves (first and second), topspin, slice,
half volley, volley, return (return games only), and overhead,
and designated as forehand or backhand as appropriate. As
ball velocity is markedly higher for serves than for other
strokes, service returns were not grouped with other ground
strokes and were categorised as a forehand or backhand
return regardless of the spin placed on the ball (for example,
topspin or slice). Backhand overheads were categorised as
backhand volleys because the ball velocity is significantly less
in backhand overheads than in forehand overheads.
Differences in the type of strokes executed within a game
were analysed using one factor (type of stroke) repeated
measures analysis of variance with Bonferroni corrections for
post hoc pairwise comparisons. Differences in total strokes
and stroke distributions between playing surfaces were
analysed using analysis of variance (surface type) with
Tukey’s post hoc pairwise comparisons. Results are reported
as mean (SD). Total strokes and strokes per game are
reported separately for servers and returners.
RESULTS
For service games (table 1) there were more serves per game
(mean (SD), 8.9 (4.7)) than any other type of stroke
(p,0.01), with topspin forehand (4.4 (4.2)) and topspin
backhand (3.0 (3.6)) being the only other strokes that
averaged more than one per service game.
For return games (table 2) there were more forehand and
backhand returns (2.3 (1.7) and 3.0 (1.9)) and topspin
forehands and backhands (3.0 (3.4) and 2.6 (3.1)) than other
types of stroke (p,0.01).
Combined data from all three tournaments on the number
of strokes and stroke distribution are given in table 3. The
696
www.bjsportmed.com
group.bmj.com on May 20, 2010 - Published by bjsm.bmj.comDownloaded from

total number of strokes per game was higher in the French
Open than in Wimbledon (service game: 21.0 (10.2) v16.0
(8.9), p,0.01; return game: 14.8 (9.2) v10.4 (6.0), p,0.01).
The difference in total strokes was primarily accounted for by
more topspin forehands (service games: 6.0 (4.2) v2.9 (3.4),
p,0.01; return games: 3.2 (3.8) v2.0 (2.2), p,0.01) and
more topspin backhands (service game: 4.2 (4.0) v1.3 (1.9),
p,0.01; return game: 3.7 (3.7) v1.8 (1.8), p,0.01). Total
strokes per game in the US Open (service game: 17.9 (12.1),
return game 12.2 (10.0)) did not differ significantly from the
other two tournaments.
There were more forehand and backhand volleys (p,0.01)
for service games in Wimbledon (forehand: 0.6 (0.9);
backhand: 0.9 (1.5)) than in French Open (forehand: 0.2
(0.4); backhand: 0.1 (0.4)) or the US Open (forehand: 0.2
(0.4); backhand: 0.3 (0.7)).
Serves (first and second) accounted for 45 (12)% of total
strokes during service games in the French Open, which was
less than for both Wimbledon (60 (17)%, p,0.01) and the US
Open (56 (18)%, p,0.01). Topspin forehands accounted for
28 (11)% of service points in the French Open, which was
more than for the US Open (21 (12)%, p,0.01) or
Wimbledon (16 (14)%). Additionally, topspin backhands
accounted for 18 (12)% of service points in the French Open
which was more than for Wimbledon (7 (8)%, p,0.01).
Similarly, for return games there was a higher proportion of
topspin forehands and topspin backhands in the French
Open (24 (12)% and 21 (15)%, respectively) than at
Wimbledon (16 (13)% and 16 (14)%, respectively; all
p,0.05)
DISCUSSION
Stroke production in tennis involves generating repetitive
forces and motions that are of high intensity and short
duration. These forces consistently subject the shoulder
region to high stress over the course of games, practice
sessions, and match play.
14
This is particularly evident in the
case of the serve, which has been documented to be the most
strenuous stroke on the upper extremity.
15
Over half of the
total force developed during the serve is generated from the
lower extremity and trunk musculature.
14
The shoulder plays
a crucial role in the kinetic chain to transfer these forces to
the hand and racquet. This leads to high levels of muscle
activity not only to enhance the bony and ligamentous
systems of the shoulder region but also to produce motion,
which is accomplished by an explosive contraction of the
internal rotators with the shoulder in an abducted position.
Fleisig et al
16
documented internal rotation velocities of the
humerus among elite players to reach 2420˚
/s during the
acceleration phase of the serve. Similar to professional
baseball pitchers, range of motion demands on the dominant
shoulder are also extremely high. Dillman reported maximal
shoulder external rotation values of 154˚during the serve
(Dillman CJ, unpublished data presented at the United States
Tennis Association National Meeting, Tucson, Arizona, 1991).
Competitive baseball pitchers and tennis players also show
shoulder internal rotation range of motion deficits on the
dominant shoulder. This is most probably the result of
repetitive microtrauma during the deceleration phase of the
pitching and service motion which leads to scar formation
and subsequent posterior capsule contracture.
6
Considering
Table 1 Data on the number of strokes and stroke distribution for service games in the
three tournaments: service games
Stroke type US Open French Open Wimbledon
Total strokes 17.9 (12.1) 21.0 (10.2) 16.0 (8.9)
Serves First 6.4 (3.2) 6.5 (2.3) 6.4 (2.9)
Second 2.5 (2.1) 2.4 (1.7) 2.6 (2.0)
Top spin Fore 4.3 (4.3) 6.0 (4.2) 2.9 (3.4)
Back 3.4 (3.8) 4.2 (4.0) 1.3 (1.9)
Slice Fore 0.1 (0.3) 0.4 (1.3) 0.1 (0.3)
Back 0.5 (1.0) 0.7 (1.1) 0.3 (0.7)
Half volley Fore 0.1 (0.2) 0.1 (0.5) 0.3 (0.6)
Back 0.1 (0.3) 0.03 (0.2) 0.2 (0.5)
Volley Fore 0.2 (0.4) 0.2 (0.4) 0.6 (0.9)
Back 0.3 (0.7) 0.1 (0.4) 0.9 (1.5)
Overhead 0.1 (0.4) 0.2 (0.6) 0.2 (0.6)
See results section for statistical analysis. Values are mean (SD).
Table 2 Data on the number of strokes and stroke distribution for return games in the
three tournaments: return games
Stroke type US Open French Open Wimbledon
Total strokes 12.2 (10.0) 14.8 (9.2) 10.4 (6.0)
Returns Fore 2.0 (1.5) 2.8 (1.9) 2.3 (1.5)
Back 3.2 (2.3) 3.0 (1.7) 2.9 (1.6)
Topspin Fore 3.2 (3.8) 3.2 (3.8) 2.0 (2.2)
Back 2.5 (3.5) 3.7 (3.7) 1.8 (1.8)
Slice Fore 0.2 (0.5) 0.4 (1.1) 0.1 (0.4)
Back 0.9 (1.4) 0.7 (0.9) 0.8 (1.2)
Half volley Fore 0.03 (0.2) 0.06 (0.2) 0.1 (0.3)
Back 0.05 (0.3) 0.02 (0.2) 0.08 (0.3)
Volley Fore 0.04 (0.2) 0.09 (0.3) 0.09 (0.3)
Back 0.09 (0.3) 0.07 (0.3) 0.1 (0.5)
Overhead 0.0 (0.0) 0.03 (0.2) 0.04 (0.2)
See results section for statistical analysis. Values are mean (SD).
Performance demands of professional male tennis players 697
www.bjsportmed.com
group.bmj.com on May 20, 2010 - Published by bjsm.bmj.comDownloaded from

the high joint velocities, extreme external range of motion
during the serve, and internal rotation deficits coupled with
the fact that serves account for approximately 45% (French
Open) to 60% (Wimbledon) of the total strokes during service
games, it is not surprising that shoulder injuries are so
prevalent in elite tennis players.
The winner of the 2003 US Open averaged 7.8 (3.2) serves
per game for 31 service games analysed. Over the two week
period of the tournament he had seven matches including
approximately 120 service games. Therefore it is estimated
that he hit over 1000 serves in singles match play alone when
factoring in serves in tiebreakers (he played seven tiebreakers
in the tournament). By contrast, an elite professional baseball
pitcher typically pitches every four days with an average of
approximately 100 pitches per game. For example, during the
2004 Major League Baseball (MLB) playoffs, a prominent
pitcher threw in four games over a 16 day period. During this
time, he averaged 102.8 (16.5) pitches and 6.3 (1.0) innings
pitched per game. The total number of pitches was 411,
which is markedly less than the total number of serves a
professional tennis player hits in a similar time period. Given
the combination of high demand and limited rest for tennis
players, it is understandable that impingement of the rotator
cuff and biceps tendon, anterior capsule attenuation, or
intrinsic tendon overload of the posterior shoulder muscu-
lature, or combinations of these, occur frequently.
In addition to the serve, ground strokes place additional
stress on the shoulder, though to a lesser degree. Our results
showed that for service games topspin ground strokes were
the second most frequently hit strokes, while for return
games there were more topspin ground strokes and service
returns than all other strokes. While muscle activity during
the preparation phase of ground strokes is minimal, the
acceleration and follow-through phases yield much higher
activity.
17
Electromyography during the forehand yields high
activity in the subscapularis, biceps brachii, pectoralis major,
and serratus anterior. The serratus anterior, subscapularis,
infraspinatus, and biceps are also moderately active during
the follow through. With regard to the backhand, the middle
deltoid, supraspinatus, and infraspinatus show a high degree
of activity during acceleration. These muscles are also active
during the follow through, along with the biceps, though to a
lesser degree. While service returns are also frequently hit
strokes, the forces placed on the shoulder are not known; it is
likely that they would be similar to those of ground strokes
during the acceleration phase. While the other stroke types
documented do occur during match play, they are less
prevalent and most probably do not play a major role in
contributing to injuries sustained by elite tennis players.
The impact of tennis court surface was evident when
comparing the total number of strokes across tournaments.
The fact that total number of strokes per game was greater in
the French Open than at Wimbledon is consistent with clay
being a slower court surface than grass. The difference in
total strokes was primarily accounted for by a larger number
of topspin ground strokes and is consistent with longer
rallies. The greater number of strokes on clay may contribute
to earlier fatigue and possibly to a higher prevalence of injury,
especially if players are forced to compete on consecutive
days.
This study was based on data from grand slam events,
which require players to win three of five sets. It should be
mentioned that this format does not exist for every
tournament. The other format used on the ATP Tour requires
the player to win two of three sets, and tournaments are
typically structured so that players compete on a daily basis
for approximately seven to 10 days. With this schedule,
players essentially compete in a different tournament every
week with minimal or no rest. Thus it seems that it may be as
demanding as competing in a grand slam event.
The present study is the first attempt to our knowledge to
quantify the performance demands of tennis among elite
players. This information is valuable for several reasons. First,
Table 3 Combined data from all three tournaments on the number of strokes and stroke
distribution (see results section for statistical analysis). Service and return games
Service games Return games
Stroke type Stroke type
Serves First 6.4 (2.9) Returns Fore 2.3 (1.7)
Second 2.5 (1.9) Back 3.0 (1.9)
Topspin Fore 4.4 (4.2) Topspin Fore 3.0 (3.4)
Back 3.0 (3.6) Back 2.6 (3.1)
Slice Fore 0.2 (0.8) Slice Fore 0.2 (0.7)
Back 0.5 (1.0) Back 0.8 (1.2)
Half volley Fore 0.2 (0.5) Half volley Fore 0.1 (0.3)
Back 0.1 (0.4) Back 0.1 (0.2)
Volley Fore 0.3 (0.7) Volley Fore 0.1 (0.3)
Back 0.4 (1.0) Back 0.1 (0.4)
Overhead 0.2 (0.5) Overhead 0.02 (0.2)
See results section for statistical analysis. Values are mean (SD).
Whatisknownonthistopic
NProfessional tennis is a year round sport with a
different tournament or competition every week. Most
injuries involve the shoulder and are secondary to
overuse
NSeveral papers have been written about tennis related
injuries, stroke biomechanics, racquet characteristics,
injury prevention, and rehabilitation and conditioning
programmes, but none has attempted to quantify the
performance demands of the sport among elite tennis
players
What this study adds
NThis study provides objective information that may
improving training techniques, coaching tactics, and
clinical decision making
NIt can serve as a template to be applied to other
populations of competitive tennis players, especially at
the junior level, in an effort to safeguard against injury
698 Johnson, McHugh
www.bjsportmed.com
group.bmj.com on May 20, 2010 - Published by bjsm.bmj.comDownloaded from

it may provide the necessary information to develop data
based rehabilitation programmes that can safely return elite
male tennis players to competition. While tennis rehabilita-
tion programmes do exist
118
they are not based on objective
data but rather on an expert’s knowledge of the sport and are
modified according to a player’s skill level. Second, as serves,
service returns, and topspin ground strokes are the pre-
dominant strokes, coaches should emphasise proper
mechanics and training of these stroke types. Finally, this
study may serve as a template that can be applied to
competitive junior players. By determining the performance
demands of the sport in this population, we will gain valuable
data that may justify the need to modify tournament
structure and training routines to safeguard against injury.
Authors’ affiliations
.....................
C D Johnson, Nicholas Institute of Sports Medicine and Athletic Trauma,
Lennox Hill Hospital, 130 East 77th St, Black Hall, New York, NY, USA
M McHugh, National Institute of Sports Medicine and Athletic Trauma
REFERENCES
1Ellenbecker TS. Shoulder injuries in tennis. In: Andrews JR, Wilk KE, eds. The
athlete’s shoulder. New York: Churchhill Livingstone, 1994:399–409.
2Lehman RC. Shoulder pain in the competitive tennis player. Clin Sports Med
1988;7:309–27.
3Winge S, Jorgensen U, Nielsen AL. Epidemiology of injuries in Danish
championship tennis. Int J Sports Med 1989;10:368–71.
4Priest JD, Nagel DA. Tennis shoulder. Am J Sports Med 1976;4:28–42.
5Andrews JR, Kupferman SP, Dillman CJ. Labral tears in throwing and racquet
sports. Clin Sports Med 1991;10:901–11.
6Ellenbecker TS, Roetert EP, Bailie DS, et al. Glenohumeral joint total rotation
range of motion in elite tennis players and baseball pitchers. Med Sci Sports
Exerc 2002;34:2052–6.
7Kibler WB, Chandler TJ, Livingston BP, et al. Shoulder range of motion in elite
tennis players. Effect of age and years of tournament play. Am J Sports Med
1996;24:279–85.
8Maylack FH. Epidemiology of tennis, squash, and racquetball injuries. Clin
Sports Med 1988;7:233–43.
9Hang YS, Peng SM. An epidemiologic study of upper extremity injury in tennis
players with a particular reference to the elbow. J Formosan Med Assoc
1984;83:307–16.
10 Budoff JE, Nirschl RP, Ilahi OA, et al. Internal impingement in the etiology of
rotator cuff tendinosis revisited. Arthroscopy 2003;19:810–14.
11 Sonnery-Cottet B, Edwards TB, Noel E, et al. Rotator cuff tears in middle-aged
tennis players: results of surgical treatment. Am J Sports Med
2002;30:558–64.
12 Groppel JL. The utilization of proper racket sport mechanics to avoid upper
extremity injury. In: Pettrone FA, eds. Proceedings of the symposium on upper
extremity injuries. St Louis: CV Mosby, 1986.
13 Nirschl RP, Sobel J. Conservative treatment of tennis elbow. Physician Sports
Med 1981;9:43.
14 Kibler WB. Biomechanical analysis of the shoulder during tennis activities. Clin
Sports Med 1995;14:79–85.
15 Yoshizawa M, Itani T, Jonsson B. Muscular load in shoulder and forearm
muscles in tennis players with different levels of skill. In: Jonsson B, eds.
Biomechanics X-B. Champaign, IL: Human Kinetics, 1987.
16 Fleisig G, Nicholls R, Elliot B, et al. Kinematics used by world
class tennis players to produce high velocity serves. Sports Biomech
2003;2:51–64.
17 Ryu KN, McCormick J, Jobe FW, et al. An electromyographic
analysis of shoulder function in tennis players. Am J Sports Med
1988;16:481–85.
18 Reinold MM, Wilk KE, Reed J, et al. Interval Sport Programs: guidelines for
baseball, tennis, and golf. J Orthop Sports Phys Ther 2002;32:293–8.
............. COMMENTARY 1 ............
The authors have made a valid point that current rehabilita-
tion programmes
1
are based mainly on expert knowledge of
the sport rather than on objective data. This study provides
the first in-depth breakdown and analysis of the type of
strokes played in actual matches. The results, however, are
not particularly surprising and confirm what could have been
predicted pre-study: serves are the dominant stroke and
probably contribute most to the high prevalence of shoulder
injuries in tennis.
2
Nevertheless this study should serve to
assist others explore the specific demands of the game and
how they contribute to injury.
T Wood
Glenferrie Private hospital, Australia
REFERENCES
1Reinold MM, Wilk KE, Reed J, et al. Interval sport programs: guidelines for
baseball, tennis, and golf. J Orthop Sports Phys Ther 2002;32:293–98.
2Yoshizawa M, Itani T, Jonsson B. Muscular load in shoulder and forearm
muscles in tennis players with different levels of skill. In: Jonsson B, ed.
Biomechanics X-B. Champaign, Illinois: Human Kinetics, 1987.
............. COMMENTARY 2 ............
This paper presents further information regarding the sport
specific intrinsic demands that are placed on athletes. This
helps us to understand the nature and distribution of injuries
in a sport, allows some guidance for clinicians in developing
injury prevention strategies, and also helps coaches develop
programmes to improve performance in the different strokes.
Some of the findings are intuitive. It is not surprising that
more strokes are hit when playing on a slower surface, and
that serves comprise a smaller percentage on those slower
surfaces. However, the data do emphasise the importance
and the predominance of the service motion in professional
play, and point to the need for interventions to increase the
player’s capability of withstanding these demands, which
should maintain performance with minimal injury risk.
B Kibler
Lexington Clinic, USA
Performance demands of professional male tennis players 699
www.bjsportmed.com
group.bmj.com on May 20, 2010 - Published by bjsm.bmj.comDownloaded from