3.10 Agility and Change of Direction

The short version

Cone drills feel like agility training. Set the cones out, run the pattern, get a time, chase a better one next week. The trouble is that the athlete knows where they are going before they start, and in a game nobody ever does. What a cone drill measures is change of direction, which is a real and trainable quality. It is just not the same thing as agility.

Agility is changing direction fast in response to something. A defender shifts his weight, a ball breaks awkwardly, a gap opens for a fifth of a second. The reading is the hard part, and it is the part that gets left out of almost every drill on almost every training ground.

This article separates the two, shows why athletes who dominate cone tests can be ordinary in a match, and covers what to do about it: the mechanics of braking and re-accelerating, and the perceptual side that decides how early you even start moving.

‌

Before anything else, get the map straight. Agility splits into two branches, and the two are trained in completely different ways.

Branch one is change-of-direction ability — the physical half. It's built from technique and footwork, braking and eccentric strength, leg strength and power, plus the anthropometry you were born with, because limb length changes the shape of a cut whether you like it or not.

Branch two is the perceptual-cognitive half: visual scanning, pattern recognition, anticipation, and the raw speed of your reactions and decisions (Sheppard & Young, 2006).

Agility is not one quality. It's those two branches multiplied together, and you can be held back by either one.

‌

Part 1 — Beginner Section: Agility Is Not a Cone Drill

The distinction that changes everything

Watch any pre-planned cone drill and you'll notice something: the athlete knows the route before the whistle goes. Nothing in the task asks them to read anything at all.

What you're timing is how well they brake, turn and accelerate. That is genuinely worth knowing. It has a name — change-of-direction ability — and it's only half the story.

Agility is a fast whole-body change of direction or speed in response to a stimulus. The stimulus is the entire point. Take it away and you're timing a choreographed sprint with corners in it.

A cone never lies to you, never arrives late, never changes its mind. Opponents do all three.

And here's the uncomfortable bit: the two barely track each other. Rank a squad on a pre-planned five-ten-five shuttle. Then rank them again on a version where they only learn the direction as they arrive.

The order changes, sometimes dramatically. The player who was third becomes eleventh.

Neither ranking is wrong, by the way. They're simply measuring different things: one is a stopwatch on the legs, the other a stopwatch on the whole athlete — eyes included.

That gap between a player's planned time and their reactive time is gold, though. It tells you how much of their problem is legs — and how much is eyes and decisions.

One-line recap: cones measure the moving, games demand the reading, and the difference between the two numbers tells you what to train next.

Reading versus reciting

A pre-planned drill is reciting a poem you already know by heart. Reactive agility is reading aloud from a page you've never seen.

Same mouth. Same vocal cords. Completely different difficulty.

One is about execution. The other is about taking in new information fast enough that execution even gets a chance.

Which is exactly why a player can look outstanding in Tuesday's drills and get turned inside out on Saturday. Nothing about their body failed. They were half a beat late reading what was in front of them — and half a beat is the whole margin.

It's also why agility ladders won't rescue anyone here. Quick feet through a ladder is rhythm and nothing else; there's no stimulus, so there's nothing to read. Lovely warm-up. Just don't file it under agility.

Why beginners still start with the mechanics

None of this makes cone drills useless. Far from it.

If you can't decelerate under control, can't get low without collapsing, can't plant a foot and push off it properly, then no amount of clever reactive work will help. You'll simply make bad decisions faster.

The mechanics come first because they're the floor everything else stands on.

So the sequence is straightforward. Teach braking. Teach getting into and out of a low position without falling apart. Build the leg strength to absorb the force of a hard cut.

Then — and only then — start removing the certainty. A coach pointing. A partner mirroring you. A ball played into space. A small game where the answer is never known in advance.

In practice, the first reactive step can be as simple as two players and a patch of grass: one leads, one shadows, and the shadow's whole job is to read and match. No kit, no lights — and it's already closer to sport than any cone pattern ever gets.

Keep those early reactive reps short, sharp and fresh, and stop while the movement still looks crisp. Sloppy, tired reps just rehearse sloppy cuts.

What a reactive drill actually needs. Three things, and none of them require equipment.

• A cue the sport would genuinely show you. A coach pointing is fine. A defender shifting their weight is better.
• No way to guess the answer in advance. Vary it enough that patterns never form.
• Game speed. A slow decision made correctly teaches almost nothing.

Where most programmes stall. They keep the cones out because cones give you a number, and a number feels like progress.

Reactive work is messier and harder to score, so it quietly disappears from the plan.

The fix isn't to throw away the cones. It's to accept that the last part of the session — the part that looks least tidy — is often the part that transfers.

Keep the scoreboard honest, too: time the planned version now and then, but judge the week by how much genuinely unpredictable work made it in.

And braking is the underrated half. Everybody wants to train acceleration. Almost nobody trains stopping.

Stopping is where cuts get lost. It's also where knees and groins get hurt.

An athlete who can absorb force fast has more usable direction changes in them than someone who can only produce it. Learn to hit the brakes before you upgrade the engine.

One last habit worth building from day one, for whenever a live partner enters the picture: watch their hips, not their feet. Feet sell the feint; hips mostly tell the truth.

Put simply: build the body first, then train the eyes — and let the equation below hang over every session you plan.

Part 2 — Advanced Section: The Perceptual-Cognitive Engine

Where the time actually goes

Total response time in a reactive movement decomposes into three intervals. Perception time is the interval from stimulus onset to recognition. Decision time is the interval from recognition to selection of a response. Movement time is everything after that.

In trained team-sport athletes, perception and decision together commonly account for a substantial share of the total, and the difference between expert and novice performers is concentrated there rather than in movement time (Williams & Ford, 2008). Two athletes with identical sprint speed can differ by hundreds of milliseconds on the field because one reads the cue earlier.

Visual search and pattern recognition

Expert athletes do not have better eyesight. They have better search strategies. Eye-tracking research consistently shows that experts fixate on fewer, more informative locations — typically proximal cues such as the hip and trunk of an opponent rather than distal cues such as the ball or the hand — and they extract usable information earlier in the opponent's movement (Williams & Ford, 2008).

They also possess richer long-term memory structures for sport-specific patterns, allowing them to recognise a developing situation from partial information. This is why occlusion studies, where footage is cut at a defined moment, discriminate experts from novices so effectively.

Anticipation, and the cost of getting it wrong

Anticipation is prediction. It buys time when correct and costs more time than it saved when wrong, because the athlete must first cancel a prepared response. This is the mechanism behind a convincing feint: The defender commits, and the cancellation penalty is larger than the original decision would have been.

Practically, this means agility training should include deception, not just speed. Drills where the stimulus is occasionally misleading train the athlete to delay commitment to an appropriate degree.

The injury dimension

Unanticipated cutting produces substantially larger knee valgus and internal rotation moments than pre-planned cutting at the same speed (Besier et al., 2001). When an athlete has time to plan, they pre-position the trunk and load the limb favourably. When they do not, the trunk is often laterally displaced over the plant leg and the knee absorbs the difference.

This gives reactive agility training a plausible injury-prevention rationale, though the case rests on the loading data rather than on a demonstrated reduction in injury rates. An athlete who has only ever cut on command has never trained the specific neuromuscular control required for the situation in which they are most likely to be hurt. Progression should be gradual: Pre-planned, then simple stimulus, then complex stimulus, then opposed.

Cognitive load and dual tasking

Adding a cognitive task — counting, remembering a sequence, tracking a second stimulus — degrades movement quality in a way that mirrors real competition. Dual-task testing is increasingly used in return-to-sport decisions because athletes who look fully recovered on isolated tests often show deficits once attention is divided. It is also a training tool: Deliberately loading attention forces movement patterns to become more automatic.

Part 3 — Testing Agility Honestly

The two-test approach

Test the same movement pattern twice: Once pre-planned, once reactive. The pre-planned time indexes physical ability. The reactive time indexes total agility. The difference between them is the perceptual-cognitive deficit, and it tells you which factor to train.

‌

• Small deficit, slow times overall. The athlete processes well but cannot move. Train change-of-direction mechanics, deceleration strength and acceleration.
• Large deficit, fast pre-planned times. The athlete moves well but reads poorly. Train perception: Video occlusion, small-sided games, reactive drills with sport-relevant stimuli.
• Large deficit and slow times. Start with mechanics, then layer perception. Do not try to fix both at once.

Practical test protocols

• Five-ten-five (pro agility) shuttle, pre-planned. The reference change-of-direction test. Record both directions.
• Reactive five-ten-five. Identical, but the first direction is signalled by a coach or light at the moment the athlete initiates.
• Y-shaped reactive agility test. Athlete sprints 5 metres, then cuts left or right at 45 degrees in response to a stimulus. The most widely used reactive protocol (Sheppard & Young, 2006).
• Sport-specific video stimulus. A projected opponent whose movement must be read. Highest validity, lowest convenience.
• Change-of-direction deficit. Shuttle time minus the time for a straight sprint of equal distance. Isolates the cost of turning from raw speed (Nimphius et al., 2018).

Reliability requires a fixed stimulus type, a fixed surface, a fixed footwear condition and a standardised warm-up. Reactive tests are noisier than pre-planned tests, so use the mean of at least three trials rather than the best (Nimphius et al., 2018).

Part 4 — Practical Section: Building Agility

The four-stage progression

Stage 1 — closed mechanics

Teach the plant. The key positions are a lowered centre of mass in the penultimate step, a wide plant foot outside the centre of mass, an inclined trunk toward the new direction, and eyes and hips turning early. Volume is low, quality is everything, and full recovery is required between repetitions.

Useful drills: Deceleration to a line and hold, lateral shuffle to sprint, crossover step to sprint, backpedal to turn and sprint, and the pre-planned shuttle.

Stage 2 — simple stimulus

Introduce a binary decision with a non-sport-specific cue. Mirror drills, coach-pointed cuts and light-gate systems all work. The cue should arrive late enough that the athlete cannot pre-plan, which usually means after they are already moving.

The single most common error at this stage is signalling too early. If the athlete can see the cue before they start, you have built a slightly noisier cone drill.

Stage 3 — complex stimulus

Move to sport-relevant cues: The hips of an opposing player, the flight of a ball, a shoulder feint. Increase the number of response options from two to three or four. Introduce deception, so that roughly one repetition in five contains a false cue.

This is also where video-based perceptual training earns its place. Temporal occlusion drills, where footage stops at a defined frame and the athlete must call the outcome, transfer measurably to on-field anticipation (Williams & Ford, 2008).

Stage 4 — opposed and constrained games

Small-sided games with manipulated constraints are the most efficient agility training available for team sports, because they train perception, decision and movement simultaneously in a representative environment (Davids et al., 2008; Paul et al., 2016). Constraints worth manipulating include pitch dimensions, player numbers, touch limits and scoring rules.

The trade-off is control. Games produce high total volume and unpredictable intensity, which makes them poor vehicles for precise dose management. Use them for development and use isolated drills when you need to control load.

Volume, intensity and placement

• Volume. Agility work is measured in total high-intensity efforts, not minutes. Six to twelve maximal reactive efforts per session is a realistic ceiling before quality falls.
• Rest. Full or near-full recovery. Twenty to sixty seconds between short efforts, two to three minutes between maximal ones. Fatigued agility training rehearses bad decisions and bad mechanics.
• Frequency. Two to three exposures per week, with at least one embedded in technical or tactical practice rather than the gym.
• Placement. Early in the session, after a thorough warm-up including multidirectional preparation.

A sample in-season week

• Day 1. Mechanics refresher: Deceleration to hold, lateral shuffle to sprint, 6 efforts. Then strength.
• Day 2. Reactive Y-drill with coach stimulus, 8 efforts, full recovery. Then technical practice.
• Day 3. Small-sided constrained games within team practice.
• Day 4. Low-amplitude reactive hops and eccentric strength maintenance. No maximal cutting.
• Day 5. Short activation only. Competition.

What each element in the week is doing

‌

• Deceleration to hold, then lateral shuffle to sprint (Day 1). Rebuilds the braking and reacceleration mechanics that every change of direction depends on. Holding the stop removes the option of cheating the position with momentum.
• Strength work following the mechanics refresher. Placed after the quality work so that fatigue does not corrupt movement patterns, and it supplies the force reserve that allows sharper cuts.
• Reactive Y-drill with coach stimulus, 8 efforts (Day 2). Introduces the perceptual and decision-making component. Full recovery between efforts protects the reaction time being trained, which is the actual target.
• Small-sided constrained games (Day 3). Deliver a high volume of genuinely unplanned agility in a game context, with the constraints steering the behaviour the coach wants without explicit instruction.
• Low-amplitude reactive hops (Day 4). Maintain ankle stiffness and elastic quality at a joint load low enough to leave the athlete fresh for competition.
• Eccentric strength maintenance. Preserves braking capacity with a single low-volume exposure rather than developing it, because development this close to competition would cost more than it returns.
• Short activation only before competition (Day 5). Primes the nervous system and rehearses the pattern without accumulating the fatigue that dulls reaction time.

‌

Supporting physical qualities

Agility rests on a physical base that is trained elsewhere. Eccentric quadriceps and hamstring strength determine braking capacity (Paul et al., 2016). Hip abductor and adductor strength determine frontal-plane control at the plant. Reactive strength determines how quickly the plant can be converted into acceleration. Trunk anti-rotation and anti-lateral-flexion strength determine whether the upper body helps or hinders the turn. Neglect these and no amount of cone work will produce a sharp athlete.

Part 6 — Periodising Agility Across a Season

Off-season: Build the physical factor

The off-season is where the change-of-direction half of the equation is built, because it requires high-force eccentric work that is difficult to recover from during competition. Priorities are eccentric quadriceps strength, frontal-plane hip strength, reactive strength, and clean deceleration mechanics at progressively higher approach speeds. Perceptual work at this stage is light and mostly video-based, because there is no meaningful opposition available.

A representative off-season structure runs mechanics twice weekly alongside a strength programme, with approach speeds increasing every two weeks and cutting angles progressing from 45 degrees to 90 degrees and finally to 180-degree turns. The angle-velocity trade-off is the governing principle: Sharper angles must be performed at lower approach speeds, and the two variables should not be increased in the same week (Dos'Santos et al., 2018).

Pre-season: Layer the perceptual factor

Pre-season is where stimulus is introduced systematically. Weeks one and two use simple binary cues. Weeks three and four move to sport-relevant cues with three or four options. Weeks five onward introduce opposed and small-sided work. Total reactive volume rises while maximal isolated cutting volume falls, because the games now supply the direction changes.

In-season: Maintain, do not develop

In-season, competition and technical practice already contain a large volume of reactive cutting, most of it unmeasured. Additional agility training should be small, sharp and early in the week. Two to three exposures of six efforts is usually sufficient to maintain quality. The main in-season job is monitoring: If cutting volume in practice rises, isolated agility volume must fall.

Methods for training the perceptual factor

• Temporal occlusion. Play footage of an opponent and cut it at a defined frame — before foot contact, at foot contact, shortly after. The athlete calls the outcome. Accuracy at earlier occlusion points is the training target.
• Spatial occlusion. Obscure a region of the display, for example the ball or the hands, forcing the athlete to use proximal cues such as the hips and trunk. This directly trains the search strategy experts already use.
• Constraint manipulation. In live play, change area size, player numbers, touch limits or scoring rules to increase the frequency of the decision you want rehearsed (Davids et al., 2008).
• Deception exposure. Include false cues in roughly one repetition in five so that the athlete learns an appropriate commitment threshold rather than reacting to the first available cue.
• Dual-task loading. Add a secondary cognitive demand — a count, a colour call, a second stimulus to track — to push movement control toward automaticity.

Return to sport after injury

Agility is usually the last quality restored and the one most often skipped. Return to sport is a clinician-led process rather than a coaching one, though: the treating clinician sets the criteria and gives the clearance, and the coach delivers the training that has been cleared. On the training side, a defensible progression after a lower-limb injury moves through straight-line running, then pre-planned change of direction at submaximal speed, then pre-planned at full speed, then simple reactive cutting, then complex reactive cutting, then opposed play — staged on criteria rather than on the calendar, in the same way that plyometric work is reintroduced in rehabilitation (Chmielewski et al., 2006). Limb symmetry under 10 per cent on single-leg hop testing is a commonly used criterion before reactive cutting is introduced, and dual-task testing is often added before clearance because deficits that disappear under focused attention often reappear when attention is divided.

This is not a medical protocol and does not replace clinical judgement; it is a training-side progression that should sit alongside the treating clinician's criteria. Stop and refer back rather than progress if there is pain during or after the session, joint swelling, giving way or a sense of instability, numbness or weakness, or a failed hop or strength test — and do not run the progression at all with an athlete who is post-operative, recovering from a concussion, or not yet cleared to return.

Common programming errors

• Signalling before movement begins. Converts a reactive drill into a pre-planned one.
• Using agility as conditioning. High heart rate and long work bouts degrade both mechanics and decision quality.
• Training only pre-planned patterns. Produces athletes who test well and play slowly, and who have never trained the movement conditions in which knee injuries most often occur.
• Ignoring practice volume. Team practice may already contain hundreds of direction changes. Isolated work must be counted on top of that, not instead of it.
• Progressing angle and velocity together. Change one variable at a time; the joint loading cost of doing both rises sharply (Dos'Santos et al., 2018).

Part 5 — Sport Applications

mixed martial arts (MMA). Agility is almost entirely reactive and occurs in a small area against a live opponent. Train level-change reactions, lateral pivots off a partner's cue, and sprawl-to-reset drills. Pre-planned footwork ladders have very little transfer.

‌

Boxing. Agility appears as ring generalship: Cutting off angles, pivoting away from the power hand, and stepping off the line under pressure. Partner mirror work with gloves on and a genuine threat is far more useful than cone patterns.

Wrestling and Brazilian jiu-jitsu (BJJ). Reactive agility is expressed in the lower body while the upper body is engaged. Train scramble reactions and directional changes from contact positions rather than free space.

Football (soccer). Very high volume of small directional changes with the ball, plus decisive reactive cuts in defending. Small-sided games are the primary tool; isolated reactive drills are the supplement.

American football. Highly position-specific. Defensive backs live in a reactive mirror world and need extensive perceptual training. Linemen change direction in a tight space with contact, favouring short, powerful lateral steps.

Basketball. Constant reactive closeouts and defensive slides on a hard surface. Deceleration capacity is the limiting physical quality; anticipation of a first step is the limiting perceptual one.

Volleyball. Agility is compressed into a very small area and driven by reading the setter and hitter. Perceptual training with occluded video has unusually good transfer here.

Hockey. On-ice direction change uses edges rather than ground friction, so off-ice cone work transfers poorly. Perceptual and decision training transfers well; train the reading, not the footwork.

Rugby. Evasion against a closing defender, usually at high speed with contact as the alternative. Emphasise late deception and the ability to cut from a genuine sprint rather than from a jog.

Tennis. The split-step timing is the single highest-value agility skill. Train reading the opponent's racquet preparation and body orientation, and pair it with lateral deceleration capacity.

Baseball. Reactive first-step quickness on the base paths and in the field. Read-and-react drills off a live batted ball outrank any pre-planned pattern.

Sprinting, Olympic weightlifting, powerlifting. Agility is not a performance quality in these sports. It remains valuable as general athleticism and injury resilience, and low-volume reactive work is a reasonable inclusion in general preparation.

Exercise Library

Reactive Mirror Drill

• Purpose. Train reaction to an opponent-based stimulus with continuous direction changes.
• Primary muscles. Gluteus maximus, gluteus medius, quadriceps. Secondary. Adductors, hamstrings, triceps surae, trunk.
• Movement pattern. Multidirectional reactive. Difficulty. Intermediate. Equipment. Partner, cones to bound the area.
• Coaching cues. Stay low with the hips loaded; read the hips, not the feet; push off the outside leg, do not step under yourself.
• Common mistakes. Watching the partner's feet; standing tall between changes; running the drill so long it becomes conditioning.
• Progressions. Add a ball, add deception, reduce the area. Regressions. Coach-pointed direction changes with fixed timing.
• Sport applications. Basketball, football, MMA, tennis.
• When to use. Early in a session, fresh. When not to use. As a conditioning finisher.
• Programming. 6 to 10 efforts of 4 to 6 seconds, 45 to 90 seconds rest.

Y-Shaped Reactive Cut

• Purpose. Develop and test reactive change of direction at speed.
• Primary muscles. Quadriceps, gluteus maximus, gluteus medius. Secondary. Hamstrings, adductors, calves, trunk.
• Movement pattern. Linear acceleration into a 45-degree reactive cut. Difficulty. Intermediate. Equipment. Cones, a stimulus source.
• Coaching cues. Lower the hips in the penultimate step; plant wide; turn the eyes and hips first; accelerate out, do not drift.
• Common mistakes. Signalling too early; planting under the body; upright trunk into the cut; taking extra chopping steps.
• Progressions. Three or four exit options, opponent-based stimulus, deception. Regressions. Pre-planned cut at submaximal speed.
• Sport applications. All field and court sports; the standard reactive agility test.
• When to use. Development blocks and testing. When not to use. When fatigued, or before deceleration mechanics are competent.
• Programming. 6 to 10 maximal efforts, 2 to 3 minutes rest.

Lateral Shuffle to Sprint

• Purpose. Bridge frontal-plane movement into linear acceleration.
• Primary muscles. Gluteus medius, gluteus maximus, adductors. Secondary. Quadriceps, hamstrings, calves.
• Movement pattern. Lateral to linear transition. Difficulty. Beginner. Equipment. Cones.
• Coaching cues. Chest over the toes; push the ground away sideways; do not click the heels together; open the hips to sprint, do not turn and step backwards.
• Common mistakes. Bobbing up and down; crossing the feet; a false step when transitioning to the sprint.
• Progressions. Reactive exit direction, resisted shuffle, longer shuffle distance. Regressions. Shuffle only, with a hold at the end.
• Sport applications. Basketball defence, football defending, tennis, boxing.
• When to use. Warm-ups and stage-one mechanics work. When not to use. With acute adductor or groin symptoms.
• Programming. 4 to 8 efforts per side, 45 to 60 seconds rest.

Recommended Viewing

• ALTIS — search for "change of direction mechanics". Precise coaching of the plant step and penultimate step.
• EXOS — search for "movement skills multidirectional". Systematic progressions for team-sport athletes.
• PJF Performance — search for "lateral quickness" and "deceleration". Practical basketball-focused progressions.
• Phil Daru — search for "MMA footwork" and "reactive agility". Combat-sport-specific reactive work.
• E3 Rehab — search for "ACL return to sport" and "cutting mechanics". The injury-risk side of unanticipated cutting.

Frequently Asked Questions

Do agility ladders improve agility? They improve foot rhythm and are a reasonable warm-up. They do not improve reactive agility, because there is no stimulus and no force application in the relevant directions.

How much of agility is trainable? Both factors improve with training. Perceptual-cognitive ability is highly trainable but largely sport-specific — it does not transfer well between sports (Williams & Ford, 2008).

Can I test agility without equipment? Yes. A partner giving a late hand signal on a Y-drill, timed with a phone at 240 frames per second, is adequate for tracking change over time.

Should agility be trained fresh or fatigued? Fresh, for development. A small amount of deliberately fatigued reactive work has a place late in a pre-season block, but it should never be the default.

Why did my athlete get slower after adding agility work? Usually volume. Reactive efforts are neurally expensive, and adding them without removing something else produces accumulated fatigue that suppresses both speed and decision quality.

Research Summary

Agility research supports a consistent set of conclusions. Pre-planned change-of-direction performance and reactive agility performance share only modest variance, so they should be trained and tested separately (Young et al., 2015). Expert-novice differences in reactive agility are located primarily in perceptual and decision-making time rather than movement time. Video-based perceptual training produces measurable improvements in anticipation that transfer to field performance in several sports. Unanticipated cutting produces greater knee loading than pre-planned cutting, which is a biomechanical argument for gradual progression from planned to reactive tasks rather than evidence that doing so lowers injury rates; in rehabilitation the same staging belongs to the treating clinician. Small-sided constrained games improve agility outcomes efficiently in team sports, though with reduced control of training load.

‌

References

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

Young, W. B., Dawson, B., & Henry, G. J. (2015). Agility and change-of-direction speed are independent skills: Implications for training for agility in invasion sports. International Journal of Sports Science and Coaching, 10(1), 159-169. https://doi.org/10.1260/1747-9541.10.1.159

Nimphius, S., Callaghan, S. J., Bezodis, N. E., & Lockie, R. G. (2018). Change of direction and agility tests: Challenging our current measures of performance. Strength and Conditioning Journal, 40(1), 26-38. https://doi.org/10.1519/SSC.0000000000000309

Besier, T. F., Lloyd, D. G., Ackland, T. R., & Cochrane, J. L. (2001). Anticipatory effects on knee joint loading during running and cutting maneuvers. Medicine and Science in Sports and Exercise, 33(7), 1176-1181. https://doi.org/10.1097/00005768-200107000-00015

Williams, A. M., & Ford, P. R. (2008). Expertise and expert performance in sport. International Review of Sport and Exercise Psychology, 1(1), 4-18. https://doi.org/10.1080/17509840701836867

Paul, D. J., Gabbett, T. J., & Nassis, G. P. (2016). Agility in team sports: Testing, training and factors affecting performance. Sports Medicine, 46(3), 421-442. https://doi.org/10.1007/s40279-015-0428-2

Davids, K., Button, C., & Bennett, S. (2008). Dynamics of skill acquisition: A constraints-led approach. Human Kinetics.

Chmielewski, T. L., Myer, G. D., Kauffman, D., & Tillman, S. M. (2006). Plyometric exercise in the rehabilitation of athletes: Physiological responses and clinical application. Journal of Orthopaedic & Sports Physical Therapy, 36(5), 308–319. Read on PubMed

Dos'Santos, T., Thomas, C., Comfort, P., & Jones, P. A. (2018). The effect of angle and velocity on change of direction biomechanics: An angle-velocity trade-off. Sports Medicine, 48(10), 2235-2253. https://doi.org/10.1007/s40279-018-0968-3