Changing direction in sport games: are cones exercises really useful?

Slalom Agility Dog Jack Russell

There is no doubt that the ability to change direction (COD) is an important part of successful performance in sport games. It is a necessary quality in the “contact” games (e.g. football, rugby, etc.), where this ability is crucial for not being intercepted by the opposition’s players while attacking, as well as for effective playing in defence. In “non-contact” games such as tennis, COD skill helps to win “difficult balls,” which demand multiple running from one side of the court to another. It seems impossible to find coaches who don’t recognise this importance, and who are not including change direction exercises into their Strength and Conditioning training programmes. However, there are some problems in COD training, which must be considered.

First of all, it is important to understand that Change of Direction (or Agility) in real game situations includes Cognitive and Physical components. The former includes Tactical Understanding, Anticipation and Visual Reaction, while the latter involves Strength and Agility. Sometimes, athlete can compensate for a disadvantage in one component with an advantage in another. For example, an older defender can be better in 1×1 situations in comparison to a younger one, even though his physical abilities are already declining, thanks to an advantage in the Cognitive components. While both components can be trained together, it would be better to understand their relative contribution to the performance of the particular player, in order to make training intervention more precise.

Cognitive training for Sport deserves a separate discussion, and I hope to return to this exciting topic in the near future.

Here, however, I am going to discuss the physical components of COD.

Game-related COD, straight acceleration and COD tests: Are the same qualities needed?

If we try to express COD physics in two simple words, they would be deceleration and acceleration. In order to change direction of his/her running, a player needs to slow down speed in the original direction and then accelerate in a new one. Sometimes, a player completely halts his/her movement before accelerating in a new direction. In order to either decelerate or accelerate, according to the Newton’s law, a force must be applied. So then, the stronger the athlete is the better he/she has to be in COD, right? It turns out that this is not so simple, as we see when we begin to train or test COD qualities in real life.

Firstly, many of the current research studies failed to demonstrate a strong correlation between the ability to accelerate straight, and the ability to change direction (Little & Williams, 2005), though formally, both exercises demand the same quality- strength. It seems that the more turns an athlete performs at a given time, the lower the correlation is (Young, McDowell, & Scarlett, 2001).

Secondly, certain strength exercises that are believed to be useful for straight acceleration (e.g. loaded squat), probably are not useful for COD (Brughelli, Cronin, Levin, & Chaouachi, 2008).

Thirdly, different tests which aim to measure COD abilities have been found to have problems with validity. Validity means that tests have to reflect real game performance. However, it is difficult to evaluate a player’s COD in the real game objectively (although the coach will definitely have a subjective opinion about it). There are some attempts to prove validity indirectly, comparing performance in tests with player’s level. The logic behind this strategy is that, if higher-ranked players perform in the test better, it would be a demonstration of the test’s validity. Green et al. tested this idea with junior rugby players (Green, Blake, & Caulfield, 2011), while Hachana performed a similar study with U-14 footballers (Hachana et al., 2014). Although both authors claimed that they achieved validity with their new and moderated traditional tests, I am taking their result with some caution, at least, because of my scepticism about junior’s rankings. It would be much more interesting to compare COD tests between, for example, English Premier League and Championship players. Besides, COD tests often don’t correlate with each other and are thus determined by different qualities. (Chaouachi et al., 2012). Actually, it means that COD tests are too specific and often, just reflect performance within the tests themselves. Therefore, they are useless for the evaluation of abilities within a real game.

In other words, we have the situation, when the straight acceleration is not related to standard COD (tests) and the real game’s agility, while the COD tests are not related to each other. For excellent reviews of this material, see (Sheppard & Young, 2006) and (Brughelli, et al., 2008).

It is clear that there are some additional qualities that allow athletes to perform COD better in real life situations. Besides, agility is a complex skill, dependent on so many variables and initial conditions that it is very difficult to test within a controlled situation.

Strength training for game-related COD

Despite this complexity, strength and conditioning professionals still need to find workouts in addition to routine strength exercises that could help improve COD within a real game. To answer this, let’s have a look. What do we need in real COD – which is different compared to straight accelerations and tests? First of all, one key element of difference is deceleration. It isn’t present in straight running (I am not taking into account vertical displacement) and, although presented in COD tests, it is completely different there as well. Another difference is the starting body position. While fairly standard (and optimal) in the straight acceleration and COD tests, this position is highly variable in real game situations, where an athlete may find him/herself in a different position (often very inconvenient) after decelerating from an initial movement, and then having to accelerate to the new direction (see pictures 1,2, and 3).

There are some qualities which may contribute to better COD performance:

1. Developing eccentric strength of the thigh muscles. This allows decelerating more rapidly and safely. One of the examples to support the importance of strength in COD may be found in the studies comparing males and females (Brown, Brughelli, & Hume, 2014). Knee biomechanics during COD were found to be different between these groups, both in pre-planned and unplanned conditions. In addition, the amplitude of knee rotations is greater for women. This can be explained by the differences in strength and Q angle (hip – shin angle is greater in women). Having weaker leg muscles may also be one of the reasons why female athletes are generally more predisposed to knee injuries than men are (Griffin et al., 2000). 2. Having the ability to apply force horizontally and laterally, rather than vertically. 3. Being able to apply force with a single leg, rather than with the both. 4. Exerting reactive force. This is the ability to produce more force in a shorter extension-contraction period (e.g bouncing higher back after landing). As a result, this can transfer to having a quicker COD (Young, James, & Montgomery, 2002). 5. Optimizing trunk muscle strength and coordination. These are needed for sudden deceleration and starting from the inconvenient position. 6. Additionally, certain morphological characteristics may play an important role as well. Players with lesser muscle mass, though are at disadvantage in maximal strength compare to more muscular and powerful athletes, may need less force for acceleration and deceleration due to their overall lower body mass (and, if we think back to Newton, lower inertia to overcome). The shorter player, though losing in stride length, is also more agile in the turns due to lower centre of mass. These advantages in COD for short and light-weight players may be especially noticeable within a “multi-turns” situation, where there are many turns in a short time. Such advantages can disappear, however, when fewer turns and longer runs are needed instead. Often, a trade-off exists between power and agility.

It probably wouldn’t be correct to say that these qualities are needed exclusively for COD running. For example, in their resent excellent work, G. Rabita et al. showed that absolute value of Ground Reaction Force is actually not what distinguishes between elite and sub-elite sprinters. Instead, the horizontal component is higher for the elite group, which allows them to not “waste” force on vertical displacement (Rabita et al., 2015). Reactive force, without a doubt, is a necessary quality for straight sprints as well. What is probably worth to consider is the specificity of training these qualities, which should help to transfer improvements in particular quality to the gains in the agility in a real game situation. For example, one-leg jumps can be performed not only vertically or straight forward, which is useful for sprints, but laterally, backwards, with the turns (e.g. one leg “tag” games). Playing different variations of the “tag” games on the slope surface allows facilitating braking abilities (downhill), as well as acceleration (uphill). Different starting positions may be used for sprints. There are ample possibilities for the coach’s creativity. Should we give up “traditional” strength exercise? Definitely not. “Basic” strength must first be built, before implementing an intensive, COD related strength programme. For instance, ability to apply strength with one leg is important for COD, but coaching experience says that dual leg exercises should be used before for the adaptation. The same is true for plyometric workouts. Some authors and coaches recommend achieve at least 1.5 bodyweight in the loaded squat, before the athlete starts an intense jumping programme (Kutz, M.R. 2003). So really, it is not a question about the necessity of “traditional” strength for players, rather, it is a problem of when – and to what extent – it should be supplemented with a specific, game-related COD strength programme.

Pre-planned and Unplanned COD

The last (but not the least) problem that I’d like to discuss further is the difference between pre-planned and unplanned COD. Though most of the coaches include special COD exercises as well as the tests in their training sessions, however, most of these exercises include some objects (poles and cones) as the reference points for changing direction. Therefore, in these conditions, the turns are pre-planned.

As a result, this can cause significant bias for training games related COD. It isn’t only because the Visual Motor Reaction is not present (turning point is known in advance, so there is no reason to react) . Coaches can justify such an absence with the reason that they’re only aiming for the “physical” component of COD and are not going to improve a reaction. However, even the physical component is different when comparing pre-planned turns with unplanned ones, which are more often present in the real game.

So what are these differences? Let’s take, for example, downhill skiers. They can tell you that in order to achieve optimal result during the slalom, he/she needs to adjust body position, trajectory and speed before the turning point, and to make a turn as smooth as possible. The same would be true for athletes when performing pre-planned COD during exercises or tests. When approaching the turning point, they show some patterns in the selected muscles: pre-activation (Besier, Lloyd, & Ackland, 2003), prior slowdown of their speed in case of sharp turns, or smoothing trajectory in COD with the lesser angles. They have time for lowering their centre of mass in advance, thus make deceleration and the first phase of acceleration to the new direction more effective (see picture 1). Unfortunately, players don’t have the same opportunities in a real game.

During the match, the player does not know when – or if – COD will happen. Even if this is known, as in tennis for instance, it’s not always possible “to prepare” the turn, because a full effort to the initial direction is needed. Therefore, when demand for COD appears, the athlete’s body position, muscle activation, and centre of mass’s height are optimal for the current movement, but not optimal for the changes (pictures 1; 2 and 3).


2015-05-03 00_17_57-Changing direction in sport games.docx - 11Microsoft Word

Picture 1: Pre-planed and prepared turn during COD test.

Though it is still around 1.8 meters to the turning point, the athlete starts to prepare turn. Note his trunk and arm positions. They have influenced (decreased) his linear speed, however.

2015-05-03 15_30_44-Changing direction in sport games.docx - Microsoft Word

Picture 2: Pre-planed, but unprepared turn.

Though the turning point is near, Federer cannot prepare turn, because he has to run at full speed and even lean forward to reach the ball. As a consequence, his starting position for new acceleration is not optimal. Compare with the previous picture.

2015-05-03 15_33_46-Changing direction in sport games.docx - Microsoft Word

Picture 3: Unplanned and unprepared turn

Nani (white shirt) was deceived by Messi, and he is now trying to catch his opponent. He is usually a very agile player; however, here his body position looks clumsy. You can see difference between Nani’s body position and the body position of an athlete performing a COD test. His legs are parallel to each other, feet are directed in the “original” direction, though his head and eyes follow Messi. Nani’s centre of mass is outside bearing area (behind), so his position is unbalanced.

In game’s COD, the time available to make the necessary adjustments is strictly limited. Therefore, other patterns of muscle activity are needed to perform turns with the maximum possible efficiency. Producing a greater eccentric contraction force with the thigh muscles, increasing the load on the knee joint (Besier, et al., 2003), (Brown, et al., 2014), and exerting different activity mode of the core muscles are among those altered patterns which are met in unplanned COD.

Differences between planned and unplanned COD are, in my opinion, the reason why it is difficult, if possible at all, to test real COD skills for the sport games. Tests should be standardized and be the same for all players. Therefore, it would be difficult to achieve uncertainty. Nevertheless, when the movements are predictable, they don’t reflect a game’s actions. Some authors tried to do so by including “reactive” component into COD test. Veale had evaluated this by adding two-choice visual reaction on light with the Australian rules footballers (Veale, Pearce, & Carlson, 2010). Farrow et al. tried to implement a more ecologically valid design, with the netball players using pre-recorded video of the opponent’s movement as the stimuli for unplanned COD (Farrow, Young, & Bruce, 2005). Both research studies claimed to have achieved validity. However, similar to the aforementioned studies, Veale had tested junior players, whereas in Farrow’s design, two-dimensional screen images could unlikely be full-value substitution for a real opponent. Both designs are, in my opinion, both overcomplicated and impractical.

To summarise the above, it would be worthwhile to say that unplanned or unprepared turns take a considerable part of all game’s related movements, and are significantly different from pre-planned movements. These turns demand specific abilities, and therefore, specific training. Though sometimes, it may be possible to achieve validity in unplanned, or rather, reactive COD tests for some sports, it is not the same for most of the other sport games. Currently, in my opinion, games’ related agility remains a widely untestable ability.

Stimulus for the training unplanned COD

So, what stimulus for COD training should be presented instead of cones?

There are more natural methods: usages of the game-related objects and their movements. These might be a ball for tennis, or an opponent in the case of the contact games. Their unpredictability in timing and direction of movements can make COD exercises really game related. Such training develops the cognitive components of COD, simultaneously with the physical ones. Whereas the game itself is the most natural method for developing COD skills, partner and ball movements can be used in different specific drills as well, including some loaded exercises, whose aim is to develop a particular muscle group or function (e.g. eccentric quadriceps contraction, core muscles stabilization function).

The exercises with a partner sometimes don’t look so “nice” as the exercises with the cones and poles, so it’s why many coaches consider them as a childish. However, we have to remember that training should reflect and emphasise game patterns and not simply be “training for the sake of training,” because, at the end of the day, we are not playing against cones.


1. Game-related COD is the different quality, which needs specific training.

2. Unplanned and games-related conditions should be used mainly for COD running exercises.

3. Eccentric strength of the leg muscles, core muscles strength and coordination are of great importance.

4. Strength exercises, where force is applied horizontally and laterally (e.g. lateral and back-forth jumps), and with the one leg should receive significant time in training.

5. Training should reflect the different starting positions and the dynamic situations.

6. Strength programmes for the prevention of injuries should take into consideration the movement’s patterns of the unplanned COD.


Besier, T. F., Lloyd, D. G., & Ackland, T. R. (2003). Muscle activation strategies at the knee during running and cutting maneuvers. Med Sci Sports Exerc, 35(1), 119-127.

Brown, S. R., Brughelli, M., & Hume, P. A. (2014). Knee mechanics during planned and unplanned sidestepping: a systematic review and meta-analysis. Sports Med, 44(11), 1573-1588.

Brughelli, M., Cronin, J., Levin, G., & Chaouachi, A. (2008). Understanding change of direction ability in sport: a review of resistance training studies. Sports Med, 38(12), 1045-1063.

Chaouachi, A., Manzi, V., Chaalali, A., Wong del, P., Chamari, K., & Castagna, C. (2012). Determinants analysis of change-of-direction ability in elite soccer players. J Strength Cond Res, 26(10), 2667-2676.

Farrow, D., Young, W., & Bruce, L. (2005). The development of a test of reactive agility for netball: a new methodology. J Sci Med Sport, 8(1), 52-60.

Green, B. S., Blake, C., & Caulfield, B. M. (2011). A valid field test protocol of linear speed and agility in rugby union. J Strength Cond Res, 25(5), 1256-1262.

Griffin, L. Y., Agel, J., Albohm, M. J., Arendt, E. A., Dick, R. W., Garrett, W. E., . . . Wojtys, E. M. (2000). Noncontact anterior cruciate ligament injuries: risk factors and prevention strategies. J Am Acad Orthop Surg, 8(3), 141-150.

Hachana, Y., Chaab?ne, H., Ben Rajeb, G., Khlifa, R., Aouadi, R., Chamari, K., & Gabbett, T. J. (2014). Validity and Reliability of New Agility Test among Elite and Subelite under 14-Soccer Players. PLoS One, 9(4), e95773. doi: 10.1371/journal.pone.0095773

Little, T., & Williams, A. G. (2005). Specificity of acceleration, maximum speed, and agility in professional soccer players. J Strength Cond Res, 19(1), 76-78.

Rabita, G., Dorel, S., Slawinski, J., Saez-de-Villarreal, E., Couturier, A., Samozino, P., & Morin, J. B. (2015). Sprint mechanics in world-class athletes: a new insight into the limits of human locomotion. Scand J Med Sci Sports, 31(10), 12389.

Sheppard, J. M., & Young, W. B. (2006). Agility literature review: Classifications, training and testing. Journal of Sports Sciences, 24(9), 919-932. doi: 10.1080/02640410500457109

Veale, J. P., Pearce, A. J., & Carlson, J. S. (2010). Reliability and validity of a reactive agility test for Australian football. Int J Sports Physiol Perform, 5(2), 239-248.

Young, W. B., James, R., & Montgomery, I. (2002). Is muscle power related to running speed with changes of direction? J Sports Med Phys Fitness, 42(3), 282-288.

Young, W. B., McDowell, M. H., & Scarlett, B. J. (2001). Specificity of sprint and agility training methods. J Strength Cond Res, 15(3), 315-319.

Leave A Reply:

(optional field)

No comments yet.