3.13 Reaction Time and Reactive Agility

The short version

Almost everything sold as "reaction training" is aimed at the wrong part of the problem. The gap between something happening and you moving is not one thing — it is a chain of steps, and they behave very differently. Some links are locked by your biology and cannot be improved. One link is enormous, varies hugely between people, and is entirely learnable.

That one link is how quickly and accurately you work out what you are looking at. Which means fast athletes are usually not faster at reacting. They are earlier at understanding.

This article takes the chain apart link by link, explains why having more options to choose from slows you down so much, separates drills that use fake cues from drills that use real ones, and lays out how to build reactive agility that still works when a real opponent turns up.

‌

What reaction time actually is

Start with the things everybody misreads. A boxer slips a punch they could not possibly have processed in time. A goalkeeper moves before the ball is even struck. A centre back gets described as "reading the game". All three get filed under lightning reflexes, and all three descriptions are wrong.

Take the chain apart and the reason is obvious. Detecting the signal and sending the command to your muscles together take somewhere around fifty to a hundred and twenty milliseconds, and that portion is basically fixed by physiology. Working out what the signal means and deciding what to do about it can take anything from eighty milliseconds to well over two hundred, depending entirely on how familiar the situation is and whether you knew where to look.

So the fixed part is small and the variable part is huge. That is the whole story in one sentence. Getting quicker is a knowledge and attention problem, not a nervous system speed problem. Nobody has ever meaningfully sped up nerve conduction with a drill.

The goalkeeper is not reacting to the ball. They are reacting perfectly normally — to the striker's plant foot, hip angle and run-up, all of which happened earlier. They bought themselves time by reading something that came first. That is a learned skill, and it is the single highest-return thing in this entire topic.

Four words that are not interchangeable. A reflex happens at the spinal cord without your brain being consulted, takes about thirty to fifty milliseconds, and cannot really be made faster. A simple reaction is one signal, one prepared answer — a sprinter and a starting gun — and it improves only slightly. A choice reaction means several possible signals and several possible answers; it is much slower and much more trainable, because the cost sits in choosing rather than in moving. Anticipation is answering before the defining signal even arrives, by reading earlier clues. That last one is where the real gains live.

More options cost you time, and not in a straight line. Every time you double the number of things that might happen, you add a roughly fixed chunk of delay. This is why a defender facing one likely move is quick and the same defender facing four is suddenly slow. Nothing about them changed.

Which is why experts do not think faster — they think less. Experience lets you bundle several possibilities into one recognisable situation, so instead of four choices you effectively have one. You are not processing each option quicker. You have quietly reduced the number of options.

Flashing lights and phone apps train an artificial world. They are not useless, and they are fun, but the cue they train is a light coming on. No sport contains that cue. Real anticipation is built by looking at real bodies doing real things.

And a cone drill you already know is not agility. If you know where you are going before you start, you are practising the mechanics of changing direction — worth doing, just not the same thing. Reactive agility needs a genuine opponent or a genuine sport cue, and something you have not already memorised.

The rest of the article goes into the detail: each link in the chain with numbers, the maths behind choice reaction, how expertise reorganises decisions, testing protocols, and drill design that produces anticipation rather than reflex trivia. Those sections are more technical, so read them when you want the mechanism rather than the map.

‌

The five links of a reactionFive stacked stages from stimulus arrival through sensory detection, decision, movement initiation and finally movement execution, with approximate timings for each.The five links of a reaction1. Stimulus occurs — 0 msThe opponent shifts weight, the ball changes direction, the starter fires. The information now exists in the environment but theathlete has not yet received it.2. Sensory detection — roughly 20 to 60 msLight or sound reaches the receptor and is transduced into neural signal. Visual detection is slower than auditory, which is whysprint starts use a gun rather than a light.3. Perception and decision — roughly 80 to 200 msThe signal is interpreted and a response is selected. This is by far the largest and most trainable component, and it is whereexpertise shows up.4. Motor initiation — roughly 30 to 60 msThe motor command is issued and travels to the muscle. Electromyographic activity begins before any visible movement occurs.5. Movement execution — variableThe limb actually moves. Everything from here on is a strength, power and technique problem rather than a reaction problem.

Figure 1. What coaches call "reaction time" is five separate processes. Training targets different links, and the link that limits a given athlete is rarely the one they assume.

‌

Look at the relative sizes. Sensory transduction and motor conduction together account for something in the order of fifty to a hundred and twenty milliseconds and are largely fixed by physiology. Perception and decision can take anywhere from eighty milliseconds to well over two hundred, depending entirely on how familiar the situation is and how well the athlete knows where to look.

‌

That is the whole story of reaction training in one observation. The fixed part is small. The variable part is large. And the variable part is a knowledge and attention problem rather than a nervous system speed problem.

‌

Reflex, reaction, and anticipation

• Reflex An involuntary response mediated at the spinal level without the brain being involved in the decision. The stretch reflex takes roughly 30 to 50 milliseconds. You cannot train a reflex to be faster in any meaningful sense, though you can change its gain.
• Simple reaction One stimulus, one prepared response. A sprinter reacting to a gun. Typically 150 to 200 milliseconds in trained athletes, and only modestly improvable.
• Choice reaction Multiple possible stimuli, multiple possible responses. Far slower than simple reaction and far more trainable, because the cost is in selecting the response rather than producing it (Schmidt & Lee, 2011).
• Anticipation Responding before the defining stimulus by reading earlier cues. This is what elite athletes actually do, and it can produce apparent response times shorter than physiologically possible reaction times.

‌

The goalkeeper who moves before the ball is struck is not reacting quickly to the ball. They are reacting normally to the striker’s plant foot, hip angle and approach line, all of which occurred earlier (Savelsbergh, 2002). That is a learned perceptual skill, and it is the single highest-return target in this entire topic.

‌

‌

Why more options mean more time

If an athlete has one possible response, they can prepare it in advance. If they have four, they must first work out which one applies. The relationship between the number of alternatives and the time taken is roughly logarithmic, and it is one of the more robust findings in experimental psychology (Hick, 1952).

‌

Choice reaction time rises with the number of optionsA logarithmic curve showing reaction time increasing as the number of possible response alternatives increases, alongside a flatter line representing a trained expert.Choice reaction time rises with the number of optionsNoviceTrained expert1234579FastModerateSlowNumber of possible response optionsRelative reaction timeExpertise flattens the curve by chunkingoptionsEvery extra option the defender mustconsider costs time

Figure 2. Reaction time increases roughly with the logarithm of the number of alternatives, a relationship known as Hick’s law. Expertise does not repeal the law, but it flattens the curve by allowing several options to be treated as one.

‌

The practical readings are immediate. Tactically, reducing the number of things a defender must consider makes them faster, which is why good defensive systems constrain options. Offensively, presenting more credible threats slows the opponent down. And in training, a drill with two options is a fundamentally easier task than one with five, which matters when you are trying to progress difficulty in a controlled way.

‌

Expertise flattens the curve rather than removing it. An experienced player does not evaluate seven options faster; they recognise the situation as one of two familiar patterns and evaluate two (Williams & Ford, 2008). This is why exposure to varied, realistic situations beats abstract speed drills.

‌

Building reactive agility that transfers

Change of direction is not agility

The distinction matters more than any other in this topic. A pre-planned cone drill measures how quickly an athlete can decelerate, reorient and reaccelerate along a known path. That is change-of-direction speed, and it is largely a mechanical and strength quality.

‌

Agility, as the term is used in the research literature, requires a response to an unplanned stimulus (Sheppard & Young, 2006). The athlete does not know where they are going until something in the environment tells them. Studies comparing the two consistently find that they discriminate different athletes: Change-of-direction time and reactive agility time correlate weakly, and reactive agility separates higher from lower level players far more reliably (Young et al., 2015; Sheppard & Young, 2006).

‌

Closed, semi-open and open drills comparedA four column table comparing five drill types by what triggers the movement, what the athlete knows in advance, what quality is trained and where it belongs in a programme.Closed, semi-open and open drills comparedTriggerKnown in advancePrimary qualityBest placementPre-planned cone drillAthlete’s own decisionEntire pathChange of directionmechanicsEarly preparation,technique workCoach call on arrivalAuditory commandOptions but not choiceSimple choice reactionMid preparationLight or app-triggered drillArtificial visual signalOptions but not choiceChoice reaction speedMid preparation, testingMirror drill versus apartnerOpponent movementNothingPerception-actioncouplingLate preparation,in-seasonSmall-sided gameFull game contextNothingDecision-making underfatigueIn-season, integrated

Figure 3. A pre-planned cone drill and a mirror drill look similar from the sideline and train almost opposite things. The trigger column is what separates them.

‌

The reason is straightforward. In a pre-planned drill the perception and decision stage of Figure 1 has been deleted. The athlete already knows the answer. Whatever the drill is training, it is not the largest component of the response.

‌

This does not make cone drills worthless. Deceleration mechanics, foot placement, and the ability to tolerate braking forces all need dedicated work, and they are easier to coach when the athlete is not simultaneously trying to read an opponent. The error is stopping there.

‌

‌

Making the stimulus resemble the sport

Once a coach accepts that the perceptual stage is the target, the question becomes what to react to. There is a hierarchy, and it runs from convenient and artificial to inconvenient and specific.

‌

1. Arbitrary signal: A whistle, a shout, or a coloured light. Easy to standardise, completely non-specific.
2. Generic directional signal: A coach pointing left or right. Slightly more relevant, still arbitrary.
3. Filmed sport footage: Video of a real opponent, paused or occluded at a decision point. Much more specific, requires setup.
4. Live constrained opponent: A partner performing restricted movements the athlete must mirror or counter.
5. Live sport situation: Small-sided games and constrained sport scenarios where cues are entirely authentic.

‌

The higher up this list a drill sits, the more of the perceptual skill it develops and the harder it is to measure cleanly. That trade-off is real and there is no way around it. The practical resolution is to use the artificial end for measurement and the authentic end for development.

‌

Light-based reactive systems deserve a specific comment because they are heavily marketed. They are excellent at producing standardised, repeatable, quantifiable reaction data, and they are genuinely useful for testing and for return-to-play benchmarking. What they do not do is train an athlete to read a hip drop or a shoulder feint, because a light contains no such information. An athlete can become very fast at responding to lights and no faster at responding to opponents.

‌

Where the delay actually comes fromA branching diagram with three main branches representing perceptual, decisional and motor sources of delay, each with specific sub-causes.Where the delay actually comes fromSlow responsePerceptual limitsLooking at the wrong cue, such as theball rather than the hipsNarrow visual search patternLate pickup because of poor bodypositioningGenuine visual processing limitationsDecisional limitsToo many options being considered, perHick’s lawWeak pattern recognition from limitedexposureHesitation caused by fear of beingwrongUnclear tactical instruction from thecoachMotor limitsPoor starting position or stanceInsufficient eccentric strength tostop and redirectLow rate of force development in thefirst stepFatigue degrading movement quality

Figure 4. Before adding reaction drills, diagnose which branch is limiting the athlete. Adding light-based drills to an athlete whose problem is a slow first step will not help.

‌

Occlusion and cue training

The most direct method for improving anticipation is temporal occlusion. The athlete watches footage of an opponent that is cut off at a defined moment, and must predict the outcome. Cut earlier and the task becomes harder; cut later and it becomes trivial.

‌

This method is used in research to establish which cues experts actually use, and the results are consistent across sports (Mann, 2007). Experts extract information from proximal, earlier-occurring sources such as the trunk and hips, while novices fixate on distal, later-occurring sources such as the hand or the ball (Abernethy & Russell, 1987; Savelsbergh, 2002). The expert is not seeing more; they are looking somewhere better.

‌

• Film a partner or opponent performing the relevant action from the athlete’s perspective.
• Cut the clip at a point before the outcome is obvious, such as the moment before a plant foot lands.
• Ask the athlete to call the direction, then reveal the answer.
• Progressively move the cut earlier as accuracy improves.
• Debrief on which cue they used, because making the cue explicit accelerates learning.

‌

This can be done with a phone and a laptop. It is unglamorous, it produces no sweat, and it targets the largest component of the reaction chain directly. For sports with a dominant one-on-one perceptual demand, it is arguably the highest-value training method available and it is almost universally neglected.

‌

‌

The motor side: You still have to be able to move

Improving perception raises the ceiling on how early an athlete can start moving. It does nothing about what happens after that. If the athlete reads the cue perfectly and then cannot decelerate, reorient and produce force through the first step, the advantage evaporates.

‌

The motor branch of Figure 4 is where strength and conditioning does its ordinary work. Eccentric strength determines how quickly the athlete can stop (Spiteri, 2014). Rate of force development determines the first step. Stance and starting position determine how much of the movement is wasted getting into a position from which force can be produced.

‌

• Eccentric braking capacity The limiting factor in most sharp direction changes. Trained through decelerative running, eccentric-emphasis strength work and controlled landing exposure.
• Rate of force development How quickly force rises from the moment of initiation. Trained through ballistic and plyometric work and through heavy lifting performed with maximal intent.
• Postural readiness The stance and weight distribution from which the athlete waits. An athlete standing tall with locked knees has added a wasted step before any movement can begin.
• Fatigue resistance Decision quality degrades with fatigue faster than movement quality does (Paul et al., 2016). Reactive drills performed only when fresh overestimate what an athlete can do late in a match.

‌

A useful diagnostic sequence: Test change-of-direction time and reactive agility time separately. If the change-of-direction time is poor, the limit is mechanical and belongs in the gym. If change-of-direction time is good but reactive agility time is poor, the deficit is perceptual and belongs in occlusion and live-opponent work.

‌

Programming reaction work across a season

Reactive work is cognitively demanding and fatigue-sensitive, which places constraints on when it can be done well.

‌

In general preparation, weight the work towards closed drills and mechanics. The athlete has capacity to absorb technical coaching, and building the motor foundation now means later reactive work is not limited by an inability to stop.

‌

In specific preparation, shift the balance. Introduce coach-called and light-triggered drills, then partner mirror drills. Keep the volume low and the quality high; twelve to twenty high-quality reactive efforts in a session is plenty.

‌

In-season, the sport itself provides most of the reactive stimulus. Supplementary work should be small, specific, and placed where it does not compete with match recovery. Occlusion work is particularly valuable here because it costs nothing physically.

‌

1. General preparation: 70 per cent closed mechanics, 30 per cent simple reactive.
2. Specific preparation: 40 per cent closed, 60 per cent reactive with increasingly sport-like cues.
3. In-season: 20 per cent closed maintenance, 80 per cent integrated within sport practice plus video-based perceptual work.
4. Always place reactive drills early in the session, before fatigue degrades decision quality.

‌

One deliberate exception: Occasionally run reactive drills under fatigue on purpose. Matches are decided late, and an athlete whose decision-making collapses at eighty minutes needs to know that and to have trained against it.

‌

A practical six-week reactive block

This block assumes an athlete with a reasonable strength base who is already competent at basic change-of-direction mechanics. Two sessions a week, placed early in the training day.

‌

Weeks one and two: Mechanics and simple reaction

• Deceleration drills: 6 repetitions of a controlled stop from a 15 metre approach.
• Pre-planned five-cone drill: 4 repetitions, timed, focus on foot placement.
• Coach-called left or right on arrival at a gate: 8 repetitions.
• Simple auditory start reaction from a two-point stance: 8 repetitions.

‌

Weeks three and four: Choice reaction and constrained opponents

• Four-gate coach-called drill with three options: 8 repetitions.
• Mirror drill against a partner in a 5 by 5 metre box: 6 rounds of 8 seconds.
• Reactive gate drill with a visual signal: 8 repetitions, recorded.
• Occlusion video session: 20 minutes, once a week, off the field.

‌

Weeks five and six: Live and fatigued

• Live one-versus-one in a constrained area: 8 rounds of 10 seconds.
• Small-sided game with a numerical overload: 4 rounds of 90 seconds.
• Reactive drill performed immediately after a conditioning bout, twice per session.
• Retest both change-of-direction time and reactive agility time at the end of week six.

‌

Retesting both measures separately at the end is the point of the whole block. If reactive agility time improved and change-of-direction time did not, the perceptual work did its job. If neither moved, the limitation is probably motor and the next block should be a strength block.

‌

‌

Sport applications

Combat sports depend almost entirely on anticipation from proximal cues. Occlusion work using footage of actual opponents, and constrained sparring where one partner has restricted options, are more valuable than any light-based system.

‌

Invasion sports such as football, rugby and hockey need both branches. The perceptual demand is high but so is the mechanical demand of decelerating a large athlete at speed, and neglecting either produces a player who is slow in a different way.

‌

Racquet sports have a very short window and depend heavily on reading the opponent’s racquet preparation and body position. Temporal occlusion training has been studied extensively in tennis and badminton and remains one of the better-supported interventions (Farrow & Abernethy, 2002).

‌

Sprinting is the one case where simple reaction genuinely matters, since the start is a single prepared response to a single known stimulus. Even there, the gains available are small relative to the gains available in the acceleration that follows.

‌

Goalkeepers and defensive specialists across sports benefit disproportionately from cue-based training because their role is defined by responding rather than initiating.

‌

Common mistakes

• Calling pre-planned cone drills agility training. They train change-of-direction mechanics. That is worth training, but it removes the perception and decision stage that makes agility what it is.
• Assuming light-based systems train sport reaction. They train reaction to lights. Excellent for standardised testing, weak for developing anticipation, because a light carries no advance information.
• Trying to train nerve conduction speed. It is essentially fixed. Time spent chasing it is time not spent on the perceptual stage, which is where the available improvement actually lives.
• Only running reactive drills when fresh. Decision quality degrades with fatigue faster than movement quality. An athlete tested only when fresh has an inflated picture of their game-day capability.
• Adding options faster than the athlete can handle them. Hick’s law means each additional option costs time. Progress from two to three to four options deliberately rather than jumping to chaos.
• Neglecting the motor side entirely. Reading the cue early is useless if the athlete cannot decelerate or produce a fast first step. Diagnose which branch is limiting before choosing the intervention.
• Never debriefing the cue. Athletes improve much faster when the cue they should have used is made explicit. Silent repetition of reactive drills is a slow way to learn something that can be taught directly.
• Measuring reaction drills without standardising them. Trigger type, starting stance, distance and surface all change the number (Nimphius & Callaghan, 2018). Comparing a light-triggered time to a coach-called time tells you nothing.

‌

Coaching cues

• "Watch the hips, not the ball" for most one-versus-one defending.
• "Stay in a position you can move from" to eliminate the wasted preparatory step.
• "Be early, not fast" to shift attention from speed to anticipation.
• "Small feet before the cut, big feet through it."
• "Decide, then commit" to reduce the hesitation cost of choice reaction.
• "Load the outside leg" to prepare the braking limb before it is needed.
• "What told you?" as the standard debrief question after every reactive repetition.
• "Quality over chaos" when a drill degenerates into random flailing.

‌

FAQs

Can you actually improve reaction time?

You can improve response time substantially, but mostly by improving the perception and decision stage rather than the sensory or motor conduction stages. Simple reaction time in a trained athlete improves only marginally. Sport response time can improve a great deal, because most of it is anticipation, and anticipation is learned.

‌

Do reaction training apps and light systems work?

They reliably improve performance on the task they train, and they are useful for standardised testing and return-to-play benchmarking. The evidence that this transfers to sport reaction is weak, because an artificial signal contains none of the advance information an athlete would use in a game. Use them for measurement, not as your main development tool.

‌

What is the difference between agility and change of direction?

Change of direction is a pre-planned movement along a known path, and it is largely mechanical. Agility requires responding to an unplanned stimulus, so it includes the perceptual and decisional stages. The two correlate weakly and should be trained and tested separately.

‌

Why do older athletes often react well despite being slower?

Because they are anticipating rather than reacting. Years of exposure build pattern recognition that lets them start moving earlier from earlier cues. This compensates for slower movement, sometimes completely, and it is a good illustration of where the trainable variance really sits.

‌

Does vision training improve sport performance?

General visual-skills training that is not tied to sport context has limited support in the research. Sport-specific perceptual training, particularly occlusion-based work using real footage from the athlete’s own sport, has considerably better support (Farrow & Abernethy, 2002). The distinction is whether the training preserves the relationship between the cue and the action.

‌

How much reactive work should I do per session?

Between twelve and twenty high-quality efforts is plenty. It is cognitively demanding and quality falls off quickly. If the athlete starts guessing rather than reading, the useful part of the session is over regardless of how many repetitions remain on the sheet.

‌

Should reaction drills be done fresh or fatigued?

Mostly fresh, so that quality is high and learning occurs. But include some deliberately fatigued exposure, because matches are decided late and decision quality degrades before movement quality does. Treat fatigued reactive work as a specific, occasional stimulus rather than the default.

‌

Is reaction time genetic?

The fixed physiological components, such as nerve conduction velocity, have a substantial genetic contribution and are not meaningfully trainable. The perceptual and decisional components, which make up the majority of the interval, are dominated by experience and deliberate practice. In practice this means most athletes have far more room to improve than they assume.

‌

Recommended videos

Each video below was chosen because it covers a specific part of this article in more depth than text alone allows.

‌

‌

‌

‌

‌

‌

Related reading on FitXplor

• 3.10 Agility and Change of Direction
• 3.7 Deceleration: Braking Mechanics and Eccentric Strength
• 3.6 Acceleration: Mechanics, Physiology, and Programming
• 1.5 The Nervous System and Athletic Performance
• 3.4 Sprint Mechanics and Maximum Velocity
• 3.9 Reactive Strength and Elastic Qualities
• 2.5 Power Development and Explosive Strength
• 1.7 Recovery Science and Adaptation

‌

References

Hick, W.E. (1952). On the rate of gain of information. Quarterly Journal of Experimental Psychology.

Sheppard, J.M., Young, W.B. (2006). Agility literature review: classifications, training and testing. Journal of Sports Sciences.

Young, W.B., Dawson, B., Henry, G.J. (2015). Agility and change-of-direction speed are independent skills. International Journal of Sports Science & Coaching.

Abernethy, B., Russell, D.G. (1987). Expert-novice differences in an applied selective attention task. Journal of Sport Psychology.

Williams, A.M., Ford, P.R. (2008). Expertise and expert performance in sport. International Review of Sport and Exercise Psychology.

Farrow, D., Abernethy, B. (2002). Can anticipatory skills be learned through implicit video-based perceptual training? Journal of Sports Sciences.

Savelsbergh, G.J.P. et al. (2002). Visual search, anticipation and expertise in soccer goalkeepers. Journal of Sports Sciences.

Nimphius, S., Callaghan, S.J. et al. (2018). Change of direction and agility tests: challenging our current measures of performance. Strength and Conditioning Journal.

Spiteri, T. et al. (2014). Contribution of strength characteristics to change of direction and agility performance in female basketball athletes. Journal of Strength and Conditioning Research.

Mann, D.T.Y. et al. (2007). Perceptual-cognitive expertise in sport: a meta-analysis. Journal of Sport and Exercise Psychology.

Schmidt, R.A., Lee, T.D. (2011). Motor Control and Learning: A Behavioral Emphasis. Human Kinetics.

Paul, D.J., Gabbett, T.J., Nassis, G.P. (2016). Agility in team sports: testing, training and factors affecting performance. Sports Medicine.