Curved Sprinting 101

A football movement misconception: players do not move like “robots” or “pac-man“

Currently there is a lack of sports specificity with traditional testing and training methods for movement patterns adopted in soccer. By this, some tasks and exercises lack shared specificity from a kinematic, kinematic, and even contextual perspective. Commonly, movements and assessments are typically linear, single plane movements. Movement performance is, among others, angle and vector-dependent, which determines ankle, knee, and hip joint contribution (Kotsifaki et al., 2021); if you frequently move with horizontal and linear projection, you’ll improve in this trajectory. It’s well known that performance improvement (i.e., force application scenarios) is task-specific (dependent on angle of joint, time available to apply force against the ground). This hypothesis can be extrapolated to other team sports.

One of the arguments to justify linear movement training during sprinting is the available equipment used to train loaded sprinting. For instance, devices in the market generally offer the possibility to load acceleration in linear conditions, with an attached vest which is not able to turn on itself. This leads to training like a robot, in straight planes, in contrast to the highly contextualised sprinting movements performed in soccer, which possess some degree or curvature and / or torso disassociation (Caldbeck & Dos’Santos, 2022).

In this blog, a new contributor, Alberto Fílter, with support from our own Tom Dos’Santos and Alistair McBurnie, will discuss curved sprinting in football. A topic that Alberto has researched, monitored, and researched at Football Science Institute Lab (FSI Lab) for the last 6 years.

Differences between ‘apparently similar’ high-intensity actions

Simplistically speaking, some high-intensity actions may share commonality. For example, if we assess high-school students, we will obtain a high relationship between linear and curved sprint performance. This analysis could be generally applied to most metrics in this “general population”. For instance, the boys/girls with better vertical jump in high-school or amateur soccer team are probably the ones who accelerate the best, throw the best, etc. However, if we assess more experienced soccer players (semi- or professional), with greater training ages, the shared commonality reduces. This could be explained by the specialisation process and the distinct physical, technical and tactical roles associated with sports performed. Due to position specific roles players have within sport, and their commonly locations, players will frequently perform the same actions during the matches due to the technical-tactical demands (i.e., jump to clear, run the channel with 40m sprint, curved sprint to break into the box, avoiding an off-side position, etc). 

This could be extrapolated to a curved-linear sprint relationship. Our data shows how this strength or relationship decreased as the player’s age increased (Fílter et al., 2022). This may suggest that the percentage of common influencing factors between both actions decreases with increasing sport specialisation. According to recent results, when athletes reach physical maturation and are proficiently experienced, skills that seemed similar (e.g., change of direction [COD] and agility) manifest as independent skills (Fílter et al., 2022). Therefore, in highly-experienced football players, training or testing high-intensity actions, including sprinting in all its forms (trajectories, rolling vs ‘explosive’ initial speed, etc), requires an individual physical and technical-tactical analysis (i.e., integrated approach) of each player to decide how best design S&C programmes to enhance performance .

Curved sprinting in football context

The first question is, knowing that the shortest distance between two points is a straight line, why do football players perform curved sprints more frequently than linear sprints?

85% of sprinting in football occurs with curvilinear trajectories during match play (Caldeck & Dos Santos, 2022).

First of all, football is a sport in which “effort self-regulation” plays a crucial role to maintain performance and limit fatigue during 90 minutes. For that reason, fatigue management may affect the movement patterns employed during football, whereby athletes may modulate their sprint trajectory. If players have to perform many COD actions, which involve mechanically and energetically demanding  accelerations, decelerations, and re-accelerations, players may modify their technique by “rounding” their movement trajectory and performing a curved sprint at an increased turning radius maintaining speed. These movements are more mechanically (no deceleration) and metabolic (no re-acceleration) efficient (Verheul et al., 2021; Bezodis et al., 2008; Osgnach et al., 2010), with less significant braking and associated muscle damage. So, although performed over longer distances, curved sprinting could be less fatiguinging from a biomechanical and neuromuscular perspective. This approach is adopted in the motor industry, whereby curved roads preserve speed without less abrupt “braking”, which can affect the car.. So, in football, curved sprinting is a way to change direction “i.e., deviate the path of travel” in “eco” mode whenever the circumstance allows it.

Secondly, in the context of football, particularly in forward players, the offside rule contributes to movement patterns employed by players, where curvilinear sprinting is a critical feature. The player intends to attack behind the central defenders (i.e., “break the trap”) with several intentions:

1. To penetrate the defensive line with a high velocity or maintain velocity, compared to lateral foot plant COD action which results in lower velocities. 
2. With appropriate timing, to maintain an “onside position” by running across the pitch in line with defender. 
3. The deviation in path of travel allows the player to visually scan and locate the ball or pass.

Therefore, the most effective way to deviate the path of travel while maintaining high velocities towards the opponent’s goal is a curvilinear sprint.

Assessing curved sprinting in football

In our first study, the primary objective was to devise a straightforward assessment tailored for curvilinear sprinting simply on a football pitch. We looked for a line that would serve as a guide, and to standardise performance, this way the player could accelerate along the line. Luckily, all football pitches have a curved line with the standard distance of a football sprint (17m): it is the arc of the penalty box area. There are different ways to assess curvilinear speed (figure 1): a) with timing gates at the beginning and the end of the arc, b) with a camera placed on the penalty spot to denote start and end frames using 2D analysis; or c) technology which tracks x-y coordinate data such as GPS, LPS, LiDAR, or radar to assess instantaneous velocities. The latter technology is not as readily available as timing gates or smartphones; nonetheless,  both a) and b) are user friendly methods which are both valid and reliable.

Curved Sprint biomechanical considerations

A fundamental distinction between curved and linear sprinting lies in the different roles played by each leg (i.e., inside and outside legs). In linear sprints, both legs exhibit comparable behaviour in terms of kinematics, kinetics, and spatio-temporal characteristics. However, the dynamics shift significantly in curvilinear sprinting. In my thesis, EMG results confirmed that each leg had a different muscle activation pattern. The gluteus medius and semitendinosus of the outer leg were more activated, while the semimembranosus and adductor of the inner leg were more active in curved running. Our study elucidated this divergence by highlighting that the inside leg experienced longer ground contact times compared to its counterpart on the outside. Indeed, the inside leg was more affected by the change from linear to curvilinear sprinting, per their contact times. Conversely, the outside leg appeared relatively unaffected by this transition. Therefore, suggesting that the inside leg plays a key role during curved sprinting.

We know that curved sprinting is different to linear sprinting, but why don’t we consider the curved sprint a change of direction?

Contrary to common belief, the curved sprint should not be classified as merely a COD action, due it being a cyclical action with no definitive lateral (or medial) foot plant. Rather, despite subtle differences between the inside and outside leg roles, curved sprinting comprises of cyclical steps, with medial trunk lean facilitating tangential acceleration to deviate the path of travel. Our research (Filter et al., 2021) underscores this distinction by revealing a limited correlation between single 90º COD speed and curvilinear sprint tests, sharing only about 11% of variance, indicating that these may be distinct qualities. A key disparity lies in the absence of a deceleration phase in the curved sprint, characterising it as a gradual trajectory shift designed to circumvent abrupt “cutting” actions. Additionally, our study uncovered that approximately 70% of football players (22 out of 33) exhibited asymmetries in one direction during a COD test, while displaying asymmetries in the opposite direction when performing a curved sprint test. This intriguing finding underscores the divergent roles played by each leg in these skills. While the outside leg assumes a greater role in COD tasks, the inside leg takes on a more prominent role in propelling the centre of mass forward during the curved sprint, providing an explanation for these observed differences. Therefore, curved sprinting ability may need to be treated as an independent speed quality that should be assessed and trained as such.

Toolbox to improve curved sprinting

Given the importance of curved sprinting highlighted above, below some key training modalities are recommended for enhancing curved sprinting.

1.          Unloaded curved sprinting: our study showed that training unloaded curved sprinting improves curved sprinting but not linear sprinting (Solleiro-Duran et al., 2024 [under review]). The findings indicated that both curved and linear sprinting elicit trajectory-specific adaptations.
2.          Loaded curved sprinting: unlike linear sprinting, curved sprinting is affected by anterior-posterior and medial-lateral (centrifugal) forces. From anterior-posterior load, this can be managed using a loaded sled or resistance devices, such as Battle4runⓇ or 1080motionⓇ placed at the beginning of curve. This can be applied to curvilinear sprints with large radius (low degree of turning). As radius decreases (at the same speed), medial-lateral forces play a more crucial role. In this way, equipment such as a weighted vest can be used to increase lateral momentum. In fact, lateral trunk flexion control could contribute to improved biomechanical conditions in order to apply more anterior-posterior forces.  If a player constantly deviates from the the intended arc line, it could be due to their lack of control of this centrifugal force, and thus, the radius modification would change centrifugal load. Some studies are needed to confirm how resistance device set-up (inside the arc [penalty spot] or outside the arc) may affect the enhancement of curved sprint performance.
3.          Incorporate movement skills by simulating the curved sprinting pattern: We know that the inside leg has a biomechanical disadvantage (longer contact times) in the application of antero-posterior force for several reasons: hip abduction and ankle eversion. This movement can be replicated in the gym using two pieces of equipment: landmine and slant board, which will exaggerate ankle eversion and facilitate force production and application from biomechanically similar positions.

A new metric and assessment: “curved sprint deceleration deficit”

Deceleration metrics have gained popularity in recent years. Although no studies have been published on this metric yet, based on unpublished data, the curve sprint test could incorporate a variant which provides a reliable and valid deceleration metric. This is named the “curved sprint deceleration deficit” (CSD deficit) and it is only necessary to curve sprint (Fílter et al., 2019) and incorporate a 1x1m box at the end of it in which the player has to stop. Thus, the CSD deficit (absolute data) can be calculated as follows: curve sprint time – curve sprint with stop time. This percentage equates to how much extra time the player needs considering that the player has to brake. To obtain relative data:

[(curve sprint with stop time * 100) / curve sprint time] – 100.

Take home messages and summary

• Curved sprinting has emerged as a novel component in the training and testing regimens of team sports. In fact, training and testing this skill within speed programmes is crucial to improve game-speed performance.
• Our research signifies a cost-effective and applicable method to assess and enhance curvilinear sprints across which can easily be set up on any football pitch with standardised pitch markings.
• In the context of speed development, the incorporation of curvilinear sprinting brings us a step closer to the specific physical demands of of the sport, warranting its prioritisation in both training and testing protocols.
• Sometimes players apply the curved sprint without knowing it in COD-oriented tasks, due to the presence of fatigue. Players apply the “eco” mode, performing a more efficient action (mechanically and metabolically) due to the absence of deceleration and reacceleration, maintaining speed during the turn using a curved sprint.
• Strength and conditioning programmes, in this context, should move outside of the sagittal plane to strategically focus on the hip, ankle, and foot movements to optimise performance in curvilinear sprinting ability.
• Strength programs focused on force application with ankle eversion is necessary to mimic inside leg biomechanical pattern during curved sprint. Slant board and landmine derivatives are useful equipment to target this.
• Curved Sprint Test Deceleration Deficit could provide us a valuable metric related to braking ability from curvilinear approaches.

References

1. Bezodis IN, Kerwin DG, Salo AI. Lower-limb mechanics during the support phase of maximum-velocity sprint running. Med Sci Sports Exerc. 2008; 40(4):707-15. doi: 10.1249/MSS.0b013e318162d162.
2. Bigland-Ritchie B, Woods JJ. Integrated electromyogram and oxygen uptake during positive and negative work. J Physiol. 1976; 260(2):267-77. doi: 10.1113/jphysiol.1976.sp011515.
3. Caldbeck P, Dos’Santos T. A classification of specific movement skills and patterns during sprinting in English Premier League soccer. PLoS One. 2022 11;17(11). doi: 10.1371/journal.pone.0277326.
4. Fílter A, Olivares J, Santalla A, Nakamura FY, Loturco I, Requena B. New curve sprint test for soccer players: Reliability and relationship with linear sprint. J Sports Sci. 2020; 38(11-12):1320-1325. doi: 10.1080/02640414.2019.1677391.
5. Fílter A, Olivares-Jabalera J, Santalla A, Morente-Sánchez J, Robles-Rodríguez J, Requena B, Loturco I. Curve Sprinting in Soccer: Kinematic and Neuromuscular Analysis. Int J Sports Med. 2020; 41(11):744-750. doi: 10.1055/a-1144-3175.
6. Fílter A, Beltrán-Garrido V, Dos’Santos T, Romero-Rodríguez D, Requena B, Loturco I, Madruga-Parera M. The Relationship Between Performance and Asymmetries in Different Multidirectional Sprint Tests in Soccer Players. J Hum Kinet. 2021 28; 79:155-164. doi: 10.2478/hukin-2021-0069.
7. Fílter-Ruger A, Gantois P, Henrique RS, Olivares-Jabalera J, Robles-Rodríguez J, Santalla A, Requena B, Nakamura FY. How does curve sprint evolve across different age categories in soccer players? Biol Sport. 2022; 39(1):53-58. doi: 10.5114/biolsport.2022.102867.
8. Kotsifaki A, Korakakis V, Graham-Smith P, Sideris V, Whiteley R. Vertical and Horizontal Hop Performance: Contributions of the Hip, Knee, and Ankle. Sports Health. 2021; 13(2):128-135. doi: 10.1177/1941738120976363.
9. Osgnach C, Poser S, Bernardini R, Rinaldo R, di Prampero PE. Energy cost and metabolic power in elite soccer: a new match analysis approach. Med Sci Sports Exerc. 2010;42(1):170-8. doi: 10.1249/MSS.0b013e3181ae5cfd.
10. Verheul J, Nedergaard NJ, Pogson M, Lisboa P, Gregson W, Vanrenterghem J, Robinson MA. Biomechanical loading during running: can a two mass-spring-damper model be used to evaluate ground reaction forces for high-intensity tasks? Sports Biomech. 2021;20(5):571-582. doi: 10.1080/14763141.2019.1584238.