For nearly 20 years, I thought I understood what consistency meant in tennis.
A consistent player had good technique and could reproduce that technique over and over again. My job as the coach was to help shape the stroke, groove the movement and give players enough repetitions that eventually the technique would become automatic. Then, when they found themselves under pressure in a match, hopefully that technique would hold together.
Unsurprisingly, this influenced almost everything I did on court. There was lots of basket feeding, isolated practice, repetitive cross-court hitting and long cooperative rallies. Players were challenged to make 20, 30 or 50 balls without missing. Mistakes were something we tried to eliminate. If players could become technically consistent, mentally consistent and make one more ball than their opponent, I believed they would become better tennis players.
It sounds completely logical.
But eventually I started asking a fairly obvious question.
Is that actually what happens when people play tennis?
The more I watched the game, studied match data and later became interested in skill acquisition, the more uncomfortable I became with my definition of consistency.
Tennis is not consistent.
The ball changes. Your opponent changes. Your position changes. The score changes. Your intention changes. Even when two balls look almost identical, you are not in exactly the same place, moving at exactly the same speed or interacting with exactly the same information.
So perhaps becoming consistent at tennis does not mean learning to reproduce the same movement.
Perhaps it means becoming consistently capable of adapting.
The match changed how I thought about consistency
One of the first things that really challenged my thinking was the work of tennis analyst Craig O’Shannessy.
For years, much of my coaching had revolved around long rallies. We might rally to four balls before playing competitively, or challenge players to reach 20, 30 or 50 shots.
Then match analytics started highlighting just how much tennis occurs in the opening exchanges.
O’Shannessy has reported from his analysis of professional tennis that around 70% of points fall within the first four shots, with only a relatively small proportion developing into the long rallies that often dominate traditional practice.
That created an uncomfortable question for me.
Why was so much of my practice centred around a version of tennis that represented such a small part of competition?
It also made me question why the serve and return had historically received so little attention in many coaching sessions. I can remember spending almost the entire lesson developing groundstrokes and then saying, “Right, ten minutes left, let’s serve.”
Yet the match begins with the serve.
The return follows.
And many points are largely shaped by what happens immediately afterwards.
This did not mean long rallies suddenly became useless. It meant I had to become clearer about what the practice was actually preparing players for.
Research examining common practice activities in elite junior tennis has found something similar. Krause and colleagues compared several practice tasks with match play and found differences in shot and movement characteristics. Practice tended to produce more cooperative behaviour, longer exchanges and fewer winners than competition.
That matters when we talk about consistency.
A player can become extremely consistent at something that does not resemble the problem they eventually need to solve.
Watch a real forehand and it is never really repeated
There was another obvious clue sitting in front of me.
Watch tennis.
Not a slow-motion technical video containing one perfect forehand. Watch an actual match.
A forehand hit three metres behind the baseline does not necessarily look like a forehand hit inside the service box. A player defending under pressure may organise their movement very differently from when they are stepping forward trying to finish a point.
Ball height changes.
Speed changes.
Spin changes.
The player’s balance changes.
Their position changes.
Their intention changes.
And therefore the movement changes.
This is closely related to the idea often associated with Nikolai Bernstein: repetition without repetition.
Skilled performers can achieve reliable outcomes while continually adjusting how they move. Contemporary motor-control literature still uses Bernstein’s phrase to describe the fact that even highly practised actions contain movement variability, and that behavioural flexibility is one of the characteristics of expertise.
This completely changes what consistency can mean.
Instead of asking:
“Can the player repeat the same swing?”
perhaps we should ask:
“Can the player continually reorganise their movement and still solve the problem?”
That is a very different version of consistency.
Consistency of movement or consistency of behaviour?
Today, when I talk to players about consistency, I am probably more interested in their behaviours than whether every forehand looks the same.
Can they consistently be brave?
Can they consistently commit?
Can they continue showing up when things become difficult?
Can they recognise an opportunity and try to create something rather than simply waiting for the opponent to miss?
When I think about the best competitors I have watched, that is often what stands out. Under scoreboard pressure they may not execute every shot perfectly, but they repeatedly demonstrate behaviours that allow them to compete.
I often describe it to players this way:
Great players don’t wait. They create.
This also means that a player could potentially become less effective while appearing more “consistent.”
Imagine a player who becomes nervous and starts rolling every forehand safely through the middle of the court. They make fewer errors. Their rally count increases.
Are they now playing better tennis?
Maybe.
But perhaps they have simply become consistently passive.
If our intention was to stretch the opponent, create space or attack an available opportunity, repeatedly putting the ball harmlessly through the middle is not necessarily functional consistency.
Consistency therefore needs a target.
Consistent at doing what?
Why I stopped trying to eliminate mistakes
For nearly 20 years I tried hard to eliminate mistakes from practice.
The logic was understandable. If players experienced success repeatedly, they would become confident. If they missed less, they would become consistent. So I created environments in which success was relatively easy to reproduce.
I now see misses very differently.
If a player hits long, that tells them something.
If they hit wide, that tells them something.
If they leave the ball too short, that tells them something.
If they put the ball into the net, that tells them something.
The human system has to become calibrated to what it can and cannot currently do. Exploration inevitably involves discovering boundaries.
This does not mean every miss is automatically good learning. A player can repeatedly fail without becoming more skilful. Coaches still need to design appropriate challenges, guide attention, provide information and sometimes change the task.
But removing failure is not the same thing as creating learning.
There is a well-established distinction in motor-learning research between performance during practice and learning that persists beyond practice. Conditions that make practice look easier and more successful can produce good immediate performance without necessarily producing the strongest retention or transfer. Conversely, more challenging practice can sometimes look worse while practice is happening but produce better longer-term learning.
That does not justify simply making every practice random or difficult either. A 2024 meta-analysis of contextual interference found benefits of more randomised practice overall, but those effects were considerably less convincing in applied sporting settings and were negligible in younger participants.
The practical message is therefore more nuanced.
Do not mistake clean practice for learning.
And do not mistake messiness for learning either.
Judge practice by what the player is becoming capable of doing later.
A miss is not necessarily a mistake
I have also changed the language I use.
I rarely describe a player hitting the ball out as making a “mistake.”
They lost the point.
That is different.
Suppose a player recognises a short ball, moves forward, sees space behind the opponent and commits to attacking it. They hit the ball five centimetres long.
Was the intention wrong?
Not necessarily.
Was their behaviour inappropriate?
Not necessarily.
Their action simply did not produce the required outcome on that occasion.
That matters because words shape how players interpret what happened.
If every lost point becomes “What did I do wrong?”, players can become increasingly internally focused and increasingly afraid of failure.
Sometimes the better discussion is:
What were you trying to do?
What information did you notice?
What opportunity did you see?
What happened?
What might you explore next time?
This is also why I am increasingly uncomfortable with how casually tennis uses the terms forced and unforced errors.
They can certainly function as useful statistical categories. But they should not automatically be treated as explanations for why a point ended.
The opponent’s previous shot matters. Court position matters. Score matters. Fatigue matters. Intention matters. The player’s current capabilities matter.
If I played Roger Federer tomorrow, I would miss plenty of balls.
That would not suddenly mean I had forgotten how to hit a forehand.
My current skill and action capabilities would simply be interacting with a far more demanding opponent.
The effectiveness of a movement is always relational.
Why I no longer ask players to rally to four before competing
One of my old favourites was:
“Rally to four, then play the point.”
I used it constantly.
But imagine Carlos Alcaraz and Jannik Sinner walking onto court and agreeing:
“Okay, let’s hit four cooperative balls first and then we’ll try to beat each other.”
It does not happen.
The server is already trying to create an advantage.
The returner is already trying to affect the server.
If neither player can finish the point immediately, they are continually trying to change the interaction in their favour.
Tennis is not four neutral balls followed by competition.
It is an interaction in which advantage and disadvantage are continually changing.
This is why I have largely stopped using the word rally in my own sessions. We tend to play points.
The word rally can subtly carry an intention: keep the ball going.
That intention is sometimes useful. There may be occasions when cooperative hitting serves a purpose.
But it is not the same intention as trying to win a point.
And intention changes behaviour.
If we constantly train players to hit the ball back towards each other and praise them for keeping it going, we should not be surprised if some eventually become very good at keeping opponents comfortable.
Pressure may require exposure, not simply a more deeply grooved technique
Another traditional assumption I held was that technique “breaks down” under pressure.
The solution therefore seemed obvious: groove the technique more deeply.
Repeat it so many times that when the scoreboard becomes uncomfortable, the movement survives.
I now think that explanation is incomplete.
Competition changes the environment.
Players experience their own expectations, expectations from parents or coaches, rankings, selections, match score, weather, opponents and consequences.
A player encountering these situations for the first time simply has less experience operating within them.
This is why I increasingly think exposure matters enormously.
A player’s first tournament feels very different from their twentieth.
That does not mean repeated tournament play automatically creates better competitors. Players can repeat unhelpful behaviours too. But repeated experience gives them opportunities to become sensitive to information they previously did not recognise and to explore different behaviours within those situations.
Take serving at 40–love.
Perhaps the first time a young player reaches that score they suddenly become cautious. They hesitate, push their serve in and lose the initiative.
Then it happens again.
Eventually they may recognise the pattern.
“Whenever I reach this score, I stop committing.”
Now they have something they can explore.
Next time perhaps they strike the serve more positively. Maybe they serve wide. Maybe into the body. Maybe down the T.
Same problem.
Different potential solutions.
That is another form of repetition without repetition.
What about muscle memory?
“Muscle memory” is one of the most common explanations used in tennis coaching.
Repeat the movement enough times and eventually the body remembers it.
As a coaching shorthand, I understand what people mean.
But tennis presents an interesting problem for that explanation.
Which movement are we trying to remember?
If I hit 100 forehands today, none will actually be identical. Even relatively controlled repetitions contain movement variability. Put me into a match and the range becomes substantially larger because ball flight, opponent, court position, intention and pressure all change.
Skill therefore cannot simply be reduced to reproducing one ideal swing regardless of context.
From an ecological perspective, skill exists in the relationship between the individual, task and environment. The player’s job is continually to organise an action that works within that relationship.
This is why repetition remains hugely important, but I think we need to be clearer about what we are repeating.
We are not necessarily repeating a movement.
We are repeatedly encountering problems.
Research in tennis has questioned highly repetitive practice that gives little consideration to the context in which the skill will ultimately be performed, and representative-learning research has demonstrated that changing the competitive information available in practice can change the behaviours that emerge.
The challenge for the coach is therefore not to remove repetition.
It is to create better repetitions.
Build a sandbox, not a script
The analogy I increasingly use with younger players is Minecraft or Roblox.
The designer creates a world.
There are boundaries. There are rules. There are tools available. Certain actions are possible and others are not.
But the player is not handed an exact script explaining every action they must perform.
They explore.
They solve problems.
They discover what works.
That is close to how I now think about practice design.
As coaches, we can create the sandbox.
We can manipulate court dimensions, scoring systems, starting positions, ball types, opponents and intentions. We can increase or reduce difficulty. We can ask questions. We can guide attention. We can demonstrate when appropriate.
But ultimately the player has to learn how to solve the problem themselves.
If I want a player to become better at stretching opponents, I do not necessarily need to prescribe one perfect cross-court forehand and make them reproduce it 50 times.
I can repeatedly create situations in which stretching the opponent becomes useful.
Sometimes the ball will be high.
Sometimes low.
Sometimes fast.
Sometimes slow.
Sometimes the player will be balanced.
Sometimes they will be moving.
They are repeatedly solving a similar problem, but the solution continually has to adapt.
That is the consistency I am interested in.
So how do you become more consistent in tennis?
If you found this article because you want to become more consistent, I would first ask what you actually mean by consistency.
If you mean you want every forehand to look identical, I am not sure that is either possible or desirable.
If you mean you want to become more reliable at solving tennis problems, then I think we can have a much more useful conversation.
Play more tennis in environments that require you to adapt. Experience different opponents, scores, balls and situations. Do not judge every miss as evidence that your technique has failed. Use misses to help calibrate what you can currently do.
If you are a coach, look carefully at the intention behind your practice. If players spend most of their time cooperatively keeping the ball going, ask whether that behaviour is helping them solve the problems they encounter in competition. If they look beautifully consistent in a basket drill, ask what happens when uncertainty, an opponent and a meaningful score return.
And rather than demanding consistency only from the player’s strokes, perhaps demand consistency from their behaviours.
Can they repeatedly commit?
Can they repeatedly explore?
Can they keep showing up when things become difficult?
Can they remain brave enough to act when an opportunity appears?
Can they learn from losing rather than becoming frightened of it?
That is a form of consistency worth developing.
For years, I tried to make tennis players more consistent by removing inconsistency from practice.
Now I think that may have been the wrong target.
Tennis itself is variable, uncertain and continually changing. Good players do not succeed by eliminating that variability. They become more capable of operating within it.
So if I had to reduce the whole article to one idea, it would be this:
Stop trying to become consistent by repeating the same technique. Become consistent at adapting, solving problems and bringing useful behaviours to an inherently inconsistent game
