See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/336221803

New Curve Sprint Test for Soccer Players: Reliability and Relationship with
Linear Sprint

Article in Journal of Sports Sciences · October 2019

DOI: 10.1080/02640414.2019.1677391

CITATIONS READS

47 7,517

6 authors, including:

Alberto Filter Ruger Jesús Olivares Jabalera

Football Science Institute University of Granada
18 PUBLICATIONS 195 CITATIONS 24 PUBLICATIONS 194 CITATIONS

SEE PROFILE SEE PROFILE

Alfredo Santalla Fabio Yuzo Nakamura

Universidad Pablo de Olavide Universidade da Maia
113 PUBLICATIONS 2,717 CITATIONS 582 PUBLICATIONS 11,917 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Jesús Olivares Jabalera on 14 October 2019.

The user has requested enhancement of the downloaded file.

 Journal of Sports Sciences

ISSN: 0264-0414 (Print) 1466-447X (Online) Journal homepage: https://www.tandfonline.com/loi/rjsp20

New curve sprint test for soccer players: Reliability

and relationship with linear sprint

Alberto Fílter, Jesús Olivares, Alfredo Santalla, Fabio Y. Nakamura, Irineu

Loturco & Bernardo Requena

To cite this article: Alberto Fílter, Jesús Olivares, Alfredo Santalla, Fabio Y. Nakamura, Irineu
Loturco & Bernardo Requena (2019): New curve sprint test for soccer players: Reliability and
relationship with linear sprint, Journal of Sports Sciences, DOI: 10.1080/02640414.2019.1677391

To link to this article: https://doi.org/10.1080/02640414.2019.1677391

Published online: 13 Oct 2019.

Submit your article to this journal

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=rjsp20
JOURNAL OF SPORTS SCIENCES
https://doi.org/10.1080/02640414.2019.1677391

New curve sprint test for soccer players: Reliability and relationship with linear sprint
Alberto Fílter a,b, Jesús Olivares b,c
, Alfredo Santalla a,b
, Fabio Y. Nakamura b
, Irineu Loturco d,e,f

and Bernardo Requena b

a
Faculty of Sport, University of Pablo de Olavide (UPO), Seville, Spain; bResearch and Development Department, Football Science Institute (FSI),
Granada, Spain; cMixed University Institute of Sport and Health (IMUDS), Granada, Spain; dNAR – Nucleus of High Performance in Sport, Sao Paulo,
Brazil; eUniversity of South Wales, Pontypridd, Wales, UK; fDepartment of Human Movement Science, Federal University of Sao Paulo, Brazil

ABSTRACT ARTICLE HISTORY

The speed performance is involved not only in linear sprints, but also in a wide spectrum of multi- Accepted 10 September 2019
directional movements, such as curve sprinting. Curved sprint can be defined as sprint with gradual and
KEYWORDS
continuous change of direction (COD). Although ~85% of the actions performed at maximum velocity in Non-linear sprint
a professional soccer league are curvilinear sprints, there is not any specific test to assess this ability. This performance; assessment;
study aimed to analyse the reliability of a new curve sprint test, and compare its results with those semi-professional soccer
obtained by soccer players in linear sprint. Forty experienced soccer players performed 3 attempts of player
curve sprint (using the penalty arc) to right and left side (17 m), and 3 linear sprints (17 m) in two different
days. The ICCs (inter-session reliability) were 0.93 for sprint curve right side (CSRS) and 0.89 for sprint
curve left side (CSLS), considered “acceptable”. The CVs (intra-session reliability) were 0.87% in CSRS and
1.15% in CSLS. The coefficient of determination (R2) between linear and curve sprinting was ~35%. The
association between curve sides was “very large” (r = 0.878; p < 0.01). In summary, we showed that “curve
sprint test” is highly reliable, and that curvilinear and linear sprints are different and independent actions.

Introduction the game, with many of these actions being characterised by

their less aggressive cutting turns and smoother angles (an arc)
Soccer is characterised as an intermittent team-sport in which (e.g., ≤15°) (Bloomfield et al., 2007). Of note, approximately 85%
high-intensity activities (e.g., maximum sprints) are regularly of the manoeuvres performed at maximum velocity in
intertwined with actions of low and moderate intensities (e.g., a professional soccer league consist of curvilinear sprints (i.e.,
walking and jogging) (Burgess, Naughton, & Norton, 2006; curve sprints), which may be defined as the upright running
Haugen, Tonnessen, Hisdal, & Seiler, 2013; Rampini, Coutts, portion of the sprint completed with the presence of some
Castagna, Sassi, & Impellizzeri, 2007). Importantly, the number degree of curvature (Caldbeck, 2019). Nonetheless, despite
of high-intensity running efforts per game has progressively their critical importance, curve sprints have been largely over-
increased across the years (Barnes, Archer, Hogg, Bush, & looked by coaches and sport scientists.
Bradley, 2014), more specifically during decisive match situations From a general perspective, curvilinear motion regularly
(Faude, Koch, & Meyer, 2012; Mohr, Krustrup, & Bangsbo, 2003). occur in soccer (Bloomfield et al., 2007; Caldbeck, 2019), regard-
This reveals the critical importance of speed-related capabilities less of playing position (except for goalkeepers), typically
for soccer performance, especially because multiple and succes- within a radius ranging from 3.5 to 11 m (Brice, Smith, &
sive accelerations are frequently observed during official games Dyson, 2008). For obvious reasons, curve sprint technique pre-
(Bloomfield, Polman, & O’Donoghue, 2007; Rampini et al., 2007). sents singular and unique kinetic and kinematic features
In addition, it is known that professional players have better (Churchill, Trewartha, & Salo, 2018). For example, when com-
performance in sprints during matches (10, 20, 30, and 40 pared to maximum linear sprints, the velocity achieved during
metres) than their less specialised counterparts (Bradley, Di curvilinear trajectories is usually lower in sprinters, which sig-
Mascio, Peart, Olsen, & Sheldon, 2010), but the duration of nificantly affects the neuromechanical demands and require-
these sprints during matches range from 2 to 4 s with respective ments during these actions (Churchill, Salo, & Trewartha, 2015;
distances lesser than 20 metres (Barnes et al., 2014). Churchill et al., 2018). Furthermore, since acceleration and
In elite soccer, the acceleration capacity is involved not speed increase with “trunk rotation” over curvilinear trajec-
solely in linear sprints, but also in a wide spectrum of complex tories, meaningful biomechanical differences between curve
and multidirectional motor-tasks (e.g., nonlinear sprints). and linear sprint techniques may be expected (Churchill et al.,
Accordingly, the “nonlinear sprint performance” appears to be 2015; Churchill, Trewartha, Bezodis, & Salo, 2016; Churchill et al.,
able to differentiate between professional and young soccer 2018). As such, maximum linear and nonlinear sprint perfor-
players (Cardoso de Araújo, Baumgart, Freiwald, & Hoppe, mances possibly embody different physical and technical cap-
2017). In this regard, it has been consistently shown that soccer abilities and, consequently, should be independently assessed
players execute hundreds of directional changes throughout and trained.

CONTACT Alberto Fílter [email protected] Faculty of Sport, University of Pablo de Olavide (UPO), Seville, Spain
© 2019 Informa UK Limited, trading as Taylor & Francis Group
2 A. FÍLTER ET AL.

There are many validated change of direction (COD) assess- Setup

ments and their relationships with other measurements such as Prior to data collection in the first testing session, each subject’s
linear sprint tests are well established and confirmed in soccer age, height and body mass were recorded. The warm-up exercises
(Brughelli, Cronin, Levin, & Chaouachi, 2008). Curvilinear displace- were: 5 min of jogging at a self-selected pace followed by a series
ments have also been assessed in soccer players (Smith, Dyson, of dynamic warm-up drills and three practice sub-maximal trials
Hale, & Janaway, 2006; Smith, Dyson, & Hate, 1997), but only at for linear and curve sprints. Subjects completed two different tests
submaximal speeds (e.g., jogging). Measuring speed and accelera- within a session: the 17 m linear sprint and curve sprint. All tests
tion during curved sprints would be interesting and critical to were performed under similar environmental conditions (13–16ºC,
understand the determinants of performance in these actions, and 35–39% of relative humidity), on artificial grass and executed
which are frequent and decisive in soccer matches. Therefore, them in a random order that was maintained across the two
the objectives of this study were to (a) test the reliability of sessions.
a new curvilinear sprint test for soccer players, and (b) compare The trial was considered “successful” when the subject ran
curved sprints with linear sprints, considering their mechanical on the guideline (arc of the area). All the trials were recorded for
demands, and using the penalty arc of a soccer field as the analysis. The subjects performed successfully three times each
“standard measure”. We hypothesised that best performances in test (the average of the three attempts was recorded), with
linear sprint tests would not be necessarily associated with best a total of nine tests with 3 min of recovery between each one
performances in curvilinear sprints. Furthermore, we expected and a minute and 30 s between sets: 3 x linear sprint + 3 x right
that the new “curve sprint test” would meet the reliability criteria. curve sprint + 3 x left curve sprint.
Linear Sprint Test. Running speed was evaluated using
17 m sprint times (0–17 m) (standing start) with 8.5 m split
Methods
times. The front foot was placed 1 m before the first timing gate
Participants in split standing start (Cronin, Green, Levin, Brughelli, & Frost,
2007).
Forty experienced soccer players (age: 22 ± 3.32 years;
Curve Sprint Test. The trajectory used to measure the sprint
height: 177 ± 4.81 cm; mass: 73.2 ± 6.2 kg) from a semi-
in curve was the arc of the area (regulation soccer field), which
professional Spanish League were recruited for this study.
has the following characteristics (Figure 1 and Figure 2):
All subjects had more than 15 years of soccer experience
and were semi-professional players in the Second B and ● Radius (from the penalty spot): 9.15 m.
Third Division Spanish League. Inclusion criteria were: train ● Distance from initial point to final point of the curve in
four times a week, not having medical contraindications
a straight line: 14.6 m.
for the tests, and have availability to carry out them. The ● Amplitude of the angle from the point of penalty spot:
methodology used was approved by the Football Science
105.84º
Institute Ethics Committee and conformed to the policy ● Total distance of the sprint curve: 17 m (obtained through
statement with respect to the Declaration of Helsinki. All
basic trigonometric analysis).
subjects were informed of the risks and benefits of the
procedure and signed an informed consent for participa-
Running curve speed was evaluated by 17 m sprint times
tion in the study.
(0–17 m) (standing start) with 8.5 m (0–8.5 m) split time. The
front foot was placed 1 m before the first timing gate and
Test setup following the line of the arch of the area. Both sides were
measured (left and right). According to the mean time spent
Procedure for the three trials for each side, we considered the slowest as
48 h after a familiarisation session with the test (Sheppard & the “weak” side, and the fastest as the “good” side.
Young, 2006), subjects participated in two test sessions to assess
the reliability of the data, with 48 h between sessions (Lockie,
Schultz, Callaghan, Jeffriess, & Berry, 2013; Sheppard & Young, Statistical analyses
2006). Soccer players attended three experimental sessions (one
familiarisation and two identical testing sessions). The players The statistical analysis was performed using the SPSS 17.0 (SPSS,
first performed the linear sprint test and then, the curve sprint Chicago, IL). Descriptive statistics are expressed as mean ± stan-
test. The only instructions before curve sprint test were: 1. Start dard deviations (SD), in addition to 95% confidence limits (95%
the sprint from behind start line (1 m distance to the first timing CL). The normality of the data distribution was checked using the
gate set), 2. Follow the arc line of the area as a guiding line, 3. If Kolmogorov–Smirnov test.
you do not sprint on the guiding line, the test will not be For the relative reliability analysis, intra-class correlation
considered valid and 4. After receiving the go ahead, you can coefficients (ICC) were used. An ICC equal to or above 0.7 was
start when you’re ready. During data collection, three sets of considered acceptable (Baumgartner & Chung, 2001). Inter-
timing gates were used over 17 m. This enables split times to session reliability was determined, using the best successfully
be measured for 8.5 m and 17 m sprints. All the players recovered (previously defined) data from each day and each variable
the same time between each sprint. Each of the repetition times obtained (8.5 m CSLS, CSRS and linear, and 17 m CSLS, CSRS,
was measured by a photoelectric cell system (Microgate Witty, and linear), via ICCs and two other separate measures
Italy) set at 90 cm height all timing gates in both sides. (Sheppard & Young, 2006).
 JOURNAL OF SPORTS SCIENCES 3

Figure 1. Characteristics of the area to measure the sprint in curve.

Figure 2. Soccer player performing new curve sprint test.

Absolute reliability is the degree to which repeated mea- Results

surements vary for individuals and the spreadsheet of Hopkins
Table 1 presents the descriptive data of both session 1 and
(2000) (Hopkins, 2000) was used to determine the typical error
session 2, which also shows data about relative reliability statis-
(TE) (s), expressed as a coefficient of variation (CV, %) (Hopkins,
tics (mean, 95% confidence limit, p value and ICC). In terms of
Schabort, & Hawley, 2001). A CV of less than 5% was set as the
relative reliability of different sprint test, the ICC of each test was
criterion for reliability.
greater than 0.7, specifically 0.93 sprint curve right side (CSRS)
Association between variables to determine the relationships
and 0.89 sprint curve left side (CSLS), considered “acceptable”.
between curve sprint speed (in both “weak” and “good” sides)
There were no significant differences between the most of test
and linear sprint speed, to help coaches and sport scientists
attempts. However, there were statistical differences between
estimate the extent of variation in straight sprint performance
session 1 and 2 in the CSRS (17 m) and CSLS (8.5 m). The effect
to a given variation in curvilinear sprint performance (i.e., shared
size (ES) showed a “small effect” in both conditions and “non-
variance). Total variance was reported by the coefficient of deter-
effect” for the rest of the tests (Cohen, 1988; Hopkins, 2004).
mination (R2) and the respective level of significance (p < 0.05).
Table 1 shows the data related to the absolute reliability of the
Additionally, a Pearson product moment correlation coefficient
different tests. Descriptive data of the two testing sessions and the
was used to test the correlation between “weak” and “good”
TE values expressed as %CV calculated from two consecutive
curve sprint performances, which was classified according to the
sessions are presented in Table 1. The mean CV was calculated
criteria defined by Hopkins et al. (Hopkins, Marshall, Batterham, &
from both the testing sessions 1 and 2. The CV of each test was less
Hanin, 2009). The level of significance was set at p < 0.05.
than 5%, specifically 0.87% in CSRS (17 m) and 1.15% in CSLS
4 A. FÍLTER ET AL.

Table 1. Descriptive data for testing sessions 1 and 2 to know relative and absolute reliability data with mean, p value for differences between the sessions, effect size
(ES), reliability statistics (intra-class correlation coefficient [ICC], typical error [TE], coefficient variations [CV]) in semi-professional soccer players.
Relative CV Absolute
Sprints Session 1 (s) Session 2 (s) Mean and 95% CL (s) p value ES ICC (CL95%) reliability TE % reliability
8.5m CSRS 1.43 ± 0.09 1.42 ± 0.09 1.43 (1.41–1.45) .245 .13 .83 (.65 – .92) Accep. .04 1.43 Accep.
17m CSRS 2.59 ± 0.12 2.57 ± 0.12 2.58 (2.54–2.59) .003* .34 .93 (.85 – .97) Accep. .03 0.87 Accep.
8.5m CSLS 1.42 ± 0.1 1.39 ± 0.08 1.4 (1.32–1.42) .005* .32 .75 (.52 – .88) Accep. .05 1.74 Accep.
17m CSLS 2.56 ± 0.13 2.54 ± 0.11 2.55 (2.5–1.57) .182 .15 .89 (.78 – .95) Accep. .06 1.15 Accep.
8.5m Lin 1.47 ± 0.13 1.48 ± 0.14 1.48 (1.4 – 1–53) .218 .14 .89 (.75 – .95) Accep. .05 1.97 Accep.
17m Lin 2.53 ± 0.13 2.54 ± 0.13 2.54 (2.54–2.57) .104 .18 .96 (.92 – .98) Accep. .03 0.5 Accep.
* Significant differences (p ≤ 0.05) between sessions; CSRS: curve sprint right side; CSLS: curve sprint left side; Lin: Linear; Accept: acceptable; TE: typical error; CV:
coefficient variations; CL: confidence limits; ES: effect size; ICC: Intra-class correlation coefficient.

(17 m). The data presented in Table 1 confirm the reliability of the part, for this similarity of reliability values with those of the
test (reproducibility and temporary changes). linear sprint.
The coefficients of determination reveal that curve sprinting
(in both good and weak sides) and linear sprinting performances
are separate and specific physical capabilities in experienced Relationship with linear sprint performance
soccer players (34% and 37% of shared variance, respectively)
Comparing the linear relationship between curvilinear and linear
(Table 3). This indicates that players who ran faster in linear
sprints, shared variances between these capabilities were 34%
sprints were not necessarily faster in curvilinear trajectories. In
and 37% for “good” and “weak” side, respectively. This indicates
contrast, a positive and “very large” (r = 0.88) correlation (Hopkins
that straight sprint can explain only a very limited variability of
et al., 2009) was found between the curve sprint performances
curve sprint (i.e., ~35%) and also confirms the hypothesis that
obtained in both “weak” and “good” sides; therefore, athletes
faster soccer players in straight sprints are not necessarily faster
who were faster to sprint to “one side” were equally able to sprint
in curvilinear trajectories. Previous research demonstrated the
faster to “the other side”.
specificity of linear sprinting when compared to COD speed,
describing moderate associations (r = 0.35) between
10 m sprint and Zigzag test (i.e., a 20 m COD test consisting of
Discussion
3 turns at 100°) (Little & Williams, 2005), and also “low” and
The main findings of this study were (a) reliability criteria of “moderate” correlations between linear sprint and different
a universally standardised curve sprint test (that uses the 9.15 COD tests (Cinarli, Kafkas, & Kafkas, 2018). In the same way that
radius penalty arc as a “guide-line”) were confirmed in experi- these studies have led to the development of specific COD
enced soccer players, and (b) performance in this curve sprint training, our results suggest that curve and lineal sprint should
test showed limited relationship with performance in a same be assessed and (probably) trained as separate physical qualities,
length lineal sprint, which suggest that curve and straight since they are different functional skills.
sprinting abilities are different physical qualities. Lastly, our results showed a “very large” (r = 0.878) correla-
tion between the curve sprint performance obtained by the
subjects in both “good” and “weak” sides (Table 2), revealing
Relative and absolute reliability that soccer players who run faster to one side (e.g., to the right
side) possibly run fast to the other side (e.g., to the left side).
Previous research has found good reliability for measures of linear
Initially, it would make sense to expect differences between
sprint and COD speed, reporting acceptable ICCs (>0.7) and CVs
both sides due to the specialisation of each player based on
(<5%) in both linear sprint and COD tests (Little & Williams, 2005).
In the present study, ICCs for the sprint CSRS and CSLS (17 m) tests
were 0.93 and 0.89, respectively, indicating higher levels of relative
Table 2. Pearson’s correlation and coefficient of determination (R2) (p ≤ 0.05)
reliability. Moreover, a small TE, expressed as %CV of 0.87% and were used to define relationship between “good/weak” sprint curve.
1.15% (for CSRS and CSLS, respectively) also shows higher abso- Sprint curve “good” side (17m)
lute reliability. Interestingly, those values are quite similar than Sprint curve “weak” side (17m) r .878
those reported in other studies regarding linear sprint assess- R2 .77
ments (CV ~0.9%) (Hopkins et al., 2001). p < .001†
Although it is first curved test is used and its comparison †Significant (p ≤ 0.01) relationship between variables.
with other studies is difficult, intermediate reliability results
between COD and linear sprint could be expected, since the Table 3. Coefficient of determination (R2) (p ≤ 0.05) was used to calculate the
curved path could imply a movement pattern less remote proportion of the curve sprint variance explained by the linear sprint.
from the linear sprint, compared to the COD. However, it Sprint curve “good” Sprint curve “weak”
doesn’t happen. Most of the nonlinear sprint trajectories in side (17m) side (17m)
a soccer match are curved (Caldbeck, 2019) with small radius Sprint curve “weak” R2 .77
side (17m) p < .001†
(from 3.5 to 11 m) (Brice, Smith, & Dyson, 2004) and less than Linear sprint (17m) R2 .34 .37
20 m long (Barnes et al., 2014). So, the specificity of our test p < .001† < .001†
(9.15 m radius, 17 m length) could be responsible, at least in †Significant (p ≤ 0.01) variance explained.
 JOURNAL OF SPORTS SCIENCES 5

their position on the pitch. However, in the analysed team there variance), and taking into account that the radius and length of
was a similar proportion of left and right side midfielders, the sprints vary in the real competition (which makes difficult
wingers and full-backs (heterogeneous sample), and the pre- a complete transfer of curve test performance to the competi-
sent study did not analyse categorising by player’s position, tion performance), strength and conditioning coaches and
which was the main limitation of this study. Future studies are sport scientists are encouraged to assess and train this capabil-
necessary to know if within the same player’s position there is ity in a separate way. These findings open new lines of research
relationship between “good” and “weak” side curve sprint regarding the biomechanical (e.g., kinematic and kinetic fac-
performance. tors), physiological and neuromuscular characteristics of max-
imal curvilinear sprinting actions in soccer players from
different levels, genders and age-categories.
Other relevant data regarding linear sprint
Another relevant data was the average of each variable. It has
been reported lower speed in curve sprints than in linear sprints Disclosure statement
regardless of radii (1, 2, 3, 4 and 6 m) in sprinters (Chang & Kram, No potential conflict of interest was reported by the authors.
2007). Briefly, when the radius decreases, the sprint performance
tends to decrease. Likewise, Churchill et al. (2015) observed higher
velocities in linear sprints than in curve sprint in track and field ORCID
athletes. Curiously, for the first time, we recorded higher speeds Alberto Fílter http://orcid.org/0000-0002-0040-0624
during curvilinear than linear trajectories in 27 of 40 subjects, Jesús Olivares http://orcid.org/0000-0003-4748-4578
specifically when considering the “good side” measurement result Alfredo Santalla http://orcid.org/0000-0001-9418-0564
within a distance of 17 m and a radius of 9.15 m (i.e., testing data: Fabio Y. Nakamura http://orcid.org/0000-0002-5336-3652
sprint curve “good” side = 2.5 s, linear sprint = 2.55 s, and sprint Irineu Loturco http://orcid.org/0000-0003-1309-2568
Bernardo Requena http://orcid.org/0000-0002-2398-5370
curve “weak” side = 2.63 s). This on the one hand suggests again
the need to investigate the influence of player’s position on
“good” and “weak” sides performance, and on the other suggest References
that these differences regarding Churchill et al. study (Churchill
Barnes, C., Archer, D., Hogg, B., Bush, M., & Bradley, P. (2014). The
et al., 2015) (with a sample of track and field athletes) could be evolution of physical and technical performance parameters in the
partially due to the specificity of the test, since this curvilinear English premier league. International Journal of Sports Medicine, 35(13),
motion regularly occur more than linear in soccer (Bloomfield 1095–1100.
et al., 2007; Caldbeck, 2019). Anyway, our limited sample (40 Baumgartner, T., & Chung, H. (2001). Confidence limits for intraclass relia-
experienced players) prevents us from understanding our results bility coefficients. Measurement in Physical Education and Exercise
Science, 5(3), 179–188.
as reference values and requires further studies to confirm these Bloomfield, J., Polman, R., & O’Donoghue, P. (2007). Physical demands of
results. different positions in FA premier league soccer. Journal of Sports
In summary, the coefficients of determination indicated that Sciences, 6(1), 63–70.
curve and linear sprint performances have a limited explained Bradley, P., Di Mascio, M., Peart, D., Olsen, P., & Sheldon, B. (2010). High-
variance of ~35%, in both “good” and “weak” sides, in experienced intensity activity profiles of elite soccer players at different perfor-
mance levels. Journal of Strength & Conditioning Research, 24(9),
soccer players. This suggests that curve and straight sprinting 2343–2351.
abilities are different physical qualities and, consequently, should Brice, P., Smith, N., & Dyson, R. (2004). Frequency of curvilinear motion
be assessed and trained in isolated conditions. As the number of during competitive soccer play. Communication to the fourth congress
maximum curve sprints throughout soccer games seems to be of science and football. Journal of Sports Science; Part II: Game Activity
influenced by matching demands (Bloomfield et al., 2007; and Analysis, 22, 502.
Brice, P., Smith, N., & Dyson, R. (2008). Curve running in soccer:
Caldbeck, 2019), additional studies should be conducted to exam- Kinematics differences between the inside and outside limbs.
ine the differences in performance between distinct playing posi- Proceedings of the XXVI International Conference of the Society of
tions. Finally, due to the importance of curvilinear sprint Biomechanics in Sports.International Society of Biomechanics in
performance in professional soccer, further studies are required Sports, Seoul, Korea.
to test the effectiveness of different training strategies in improv- Brughelli, M., Cronin, J., Levin, G., & Chaouachi, A. (2008). Understanding
change of direction ability in soccer: A review of resistance training
ing this ability in elite soccer players. studies. Sports Medicine, 38(12), 1045–1063.
Burgess, D., Naughton, G., & Norton, K. (2006). Profile of movement
demands of national football players in Australia. Journal of Science
Practical applications and Medicine in Sport, 9(4), 334–341.
Caldbeck, P. (2019). Contextual Sprinting in Football. Impress. (DSportExSci
The proposed “curve sprint test” is a specific soccer speed
thesis). John Moores University, Liverpool, UK.
measurement, performed in the arc of the area, within Cardoso de Araújo, M., Baumgart, C., Freiwald, J., & Hoppe, M. (2017).
a radius of 9.15 m (official FIFA radius) and a distance of Nonlinear sprint performance differentiates professional from young
17 m. The data reported herein confirm that, when executed soccer players. Journal of Sports Medicine and Physical Fitness, 58(9),
under these standardised parameters, this universally standar- 1204–1210.
dised test can provide a very precise and reliable measure of Chang, Y., & Kram, R. (2007). Limitations to maximum running speed on flat
curve. Journal of Experimental Biology, 210(Pt 6), 971–982.
curvilinear sprint performance. As curve sprinting is an essential Churchill, S., Salo, A., & Trewartha, G. (2015). The effect of the bend on
and decisive action during soccer games and possess a limited technique and performance during maximal effort sprinting. Sports
relationship with linear sprinting speed (i.e., ~35% of shared Biomechanics, 14(1), 106–121.
6 A. FÍLTER ET AL.

Churchill, S., Trewartha, G., Bezodis, I., & Salo, A. (2016). Force produc- Hopkins, W., Marshall, S., Batterham, A., & Hanin, J. (2009). Progressive
tion during maximal effort bend sprinting: Theory vs reality. statistics for studies in sports medicine and exercise science. Medicine
Scandinavian Journal of Medicine & Science in Sports, 26(10), and Science in Sports & Exercise, 41(1), 3–13.
1171–1179. Hopkins, W., Schabort, E., & Hawley, J. (2001). Reliability of power in physical
Churchill, S., Trewartha, G., & Salo, A. (2018). Bend sprinting performance: performance test. Sports Medicine, 31(3), 211–234.
New insights into the effect of running lane. Sports Biomechanics, 18(4), Little, T., & Williams, A. (2005). Specificity of acceleration, maximum speed,
437–447. and agility in professional soccer players. Journal of Strength &
Cinarli, F., Kafkas, A., & Kafkas, M. (2018). Relationship between linear Conditioning Research, 19(1), 76–78.
running and change of direction performances of male soccer players. Lockie, R., Schultz, A., Callaghan, S., Jeffriess, M., & Berry, S. (2013). Reliability
Turkish Journal of Sport and Exercise, 20(2), 93–99. and validity of a new test of change of direction speed for field-based
Cohen, J. (1988). Statistial power analysis for the behavioral science (Second sports: The change of direction and acceleration test (CODAT). Journal of
Edition ed.). Hillsdale: Erlbaum. Sports Science and Medicine, 12(1), 88–96.
Cronin, J., Green, J., Levin, G., Brughelli, M., & Frost, D. (2007). Effect of Mohr, M., Krustrup, P., & Bangsbo, J. (2003). Match performance of
starting stance on inicitial sprint performance. Journal of Strength & high-standard soccer players with special reference to development
Conditioning Research, 21(3), 990–992. of fatigue. Journal of Sports Sciences, 21(7), 519–528.
Faude, O., Koch, T., & Meyer, T. (2012). Straight sprinting is the most Rampini, E., Coutts, A., Castagna, C., Sassi, R., & Impellizzeri, F. (2007).
frequent action in goal situations in professional football. Journal of Variation in top level soccer match performance. International Journal
Sports Sciences, 30(7), 625–631. of Sports Medicine, 28(12), 1018–1024.
Haugen, T., Tonnessen, E., Hisdal, J., & Seiler, S. (2013). The role and devel- Sheppard, J., & Young, W. (2006). Agility literature review: Classifications,
opment of sprinting speed in soccer. International Journal of Sport training and testing. Journal of Sports Sciences, 24(9), 919–932.
Physiology and Performance, 9(3), 432–441. Smith, N., Dyson, R., Hale, T., & Janaway, L. (2006). Contributions of the
Hopkins, W. (2000). Reliability from consecutive pairs of trials (excel inside and outside leg to maintenance of curvilinear motion on a natural
spreadsheet). Available from: http://www.sportsci.org/resource/ surface. Gait & Posture, 24(4), 453–458.
stats/xrely.xls Smith, N., Dyson, R., & Hate, T. (1997). Lower extremity muscular adaptations to
Hopkins, W. (2004). How to interpret changes in a athletics performance curvilinear motion in soccer. Journal of Human Movement Studies, 33(4),
test? Sports Scicience a Peer, 8, 1–7. 139–153.

View publication stats