Why your small-sided games are not working

It is a Wednesday evening in October. Fourteen U14s, one pitch, ninety minutes. You have set up a 4v4 in a box near the 23, and for about six minutes it looks glorious. Heads up, quick touches, one kid actually plays a proper third man ball. You are thinking, this is it, this is the session.

Then Saturday comes around, and the same fourteen kids play the same 4v4 problem in the middle of the pitch and none of it shows up. They queue. They wait. They look at you.

So what happened between Tuesday and Saturday? Nothing mysterious. The game on Tuesday was fun and busy, but it was not actually asking the same question that Saturday asks. That is the whole problem with small-sided games. Everybody agrees they are better than cone drills. Far fewer of us are honest about the fact that a badly designed or executed small-sided game is just a busier cone drill with more sweat and less learning.

Small-sided games only transfer to match day if they are designed to ask a match-day question. That means keeping all four phases of the game in the exercise, picking one focus and one only, using space, numbers and scoring as the coaching tools rather than your voice, and being clear about which decision window you are actually training: the look before the ball arrives, or the decision once it is on the stick. The coaching craft is not in inventing the game. It is in what you watch, what you say, and how often you keep your mouth shut.

Small Sided Games by Andreu Enrich is the philosophical backbone: what a small-sided game actually is, what constraints do, and why drills fail.

Basic Skills through Small Sided Games with Lisa Letchford is the counterweight, showing how you keep technical development alive inside game-based work instead of losing it.

Practice Session Design with Fede Tanuscio covers where these games sit in a week and a season, and why a game with no plan behind it is wasted pitch time.

Boosting Field Hockey IQ: Training On-Ball Decision Making with Robert Noall is the most practically useful masterclass on separating the prescan from the on-ball moment.

What is game intelligence with Mathias Becher gives you the diagnostic tool for working out why a decision went wrong.

Practical Approaches for Fostering Creative Field Hockey Players with Tin Matkovic handles the tension between individual brilliance and team structure.

And When to Stop Coaching and Let Them Play: The Hardest Skill in Youth Coaching deals with the bit nobody wants to talk about, which is us.

Enrich does not hedge on this. His definition of a small-sided game is simply a training task where the game is played by fewer players and under certain constraints, and the crucial word in there is game. Not a simplified version of hockey. Hockey, with fewer bodies.

His strongest point, and the one that should change how you look at your own session plans, is about the four phases. Attack, defence, and both transitions. All four, all the time. As he puts it: "how many times we've seen training tasks, drills where the four natural faces of the game are not represented... every training task that you practice during the week has to contain these essential four phases."

Go and audit your last three sessions against that. Most of us have at least one exercise where the attackers attack, and when the defenders win it, the ball goes dead or gets pushed to the side. Enrich's view is that this is not a small compromise. It removes the thing the players most need to learn, which is what happens in the two seconds after a turnover. If your training never contains that moment, your team will never be good at it, no matter how many times you shout about counter press on Saturday.

Lisa Letchford comes at the same conclusion from the technical side. Her argument for game-based work is not that it is more fun, although it is. It is that the player feels why the skill exists: "So by playing in a game and recognizing why you're doing it. It really helps kids to just to grow and develop a little bit quicker." Receiving across the body around a cone is a movement pattern. Receiving across the body with a defender arriving on your right shoulder, so that you can play the line pass, is a reason.

But here is where Letchford is more useful than the purists. She is explicit that this is not an either/or, and she is describing a national programme that swung hard one way and then had to come back: "there's two extremes. There's running around cones, just doing drills, just doing basic skills and there's just doing gameplay with nobody talking about basic skills. And there's somewhere in the middle where you might choose gameplay to help deliver basic skills." Her working pattern is start with the game so they see the why, break the skill out and fix it in isolation if it is genuinely broken, then put it straight back in the game and let them feel the difference.

That, for my money, is the honest position. The game is the primary teacher. It is not the only one.

There is a useful way to think about why both are needed. Some skills in hockey are closed and repeatable, where there is essentially one correct pattern and the job is to groove it until it is reliable under any conditions. The mechanics of a hit, the shape of a drag flick, the grip change on the reverse. Others are open, where there is no single correct answer and the whole skill is reading, recognising and reacting to whatever picture the game gives you. Almost everything interesting in hockey is the second kind, but the second kind sits directly on top of the first. If a player cannot reliably stop a ball, no amount of game exposure will teach them to decide well, because every scrap of their attention is going into the stop. Match your method to the kind of skill you are building and the drills versus games argument mostly dissolves on its own.

The most common failure mode with small-sided games is not over-coaching. It is under-designing, and then over-coaching to compensate for it.

Fede Tanuscio has a phrase for the underlying sin, and it is blunt:

His position is that a session that is not linked to a weekly, monthly and yearly plan is not a session, it is an activity. "Everything should be prepared before, everything should be linked with a weekly program, monthly program and year program." His junior week in Luxembourg splits it cleanly, with one evening built around technical work and one built around small-sided game principles and game situations, so that in a single week the players have touched the skill, the principle and the live problem.

There is an old coaching line that luck is the residue of design, and it applies uncomfortably well to small-sided games. The sessions that look loose and spontaneous, where the players appear to be solving everything for themselves, are almost always the ones that were planned in the most obsessive detail. The freedom they feel on the pitch is a direct product of the constraints you set before they arrived.

Once you accept the game needs designing, the design levers are where the actual coaching lives. Enrich walks through them: space and zones, time, numbers and jokers, the scoring system, technical constraints, tactical constraints, goals, rules, even the umpiring and the senses.

The pitch shape example is the one that catches most coaches out, because it is completely counter-intuitive. If you want your team to use the width, the instinct is to make the pitch wide. Enrich says the opposite: "if you design a field that is really, really, really flat, the incentive for the players to go for the vertical line to the goal is higher... So paradoxically, if you want that players use the width, you make it narrow." Make it long and thin, and the width becomes valuable because it is scarce. Make it short and fat, and everyone just runs at the goal, because why would you not.

Scoring is the other lever nobody uses enough. If your team keeps entering the circle and then panicking, do not talk about it. Make circle entry worth a point, a shot worth two, and a goal worth three. You have now paid them to do the thing you want, and they will organise themselves around it without you saying a word.

And then there is the discipline of one focus. Enrich again: "the focus, the main focus of the small sided game must be one of these dimensions... The feedback must go towards this dimension." Letchford's version of the same rule is more quotable and, frankly, more brutal: "It's really important that you're watching what you should be watching. So we're not trying to coach hockey 101 every session."

One game. One question. If you are giving feedback on the press, the receiving skill, the body shape and the width in the same twelve-minute block, none of it is landing.

And this is not only about your attention, it is about theirs. Hand a group everything you know in one go and they absorb almost none of it. Deliver one small, correctly sequenced piece at a time and it sticks. The temptation, especially when you can see four things going wrong at once, is to fix all four. Fix one properly and the other three often improve on their own, because they were symptoms rather than causes.

This is the part of Robert Noall's work that I think is genuinely underused, and it is the single biggest lever on transfer.

He splits the decision into two distinct windows. There is the pre-scan, which is the period while the ball is travelling towards the player and they are reading the picture. And there is the on-ball decision, which is what happens once it is on the stick and the picture has changed.

Most of us train the first one and assume the second one comes free. It does not. Noall's video breakdowns show a young player who prescans correctly, decides to play wide, and is then intercepted, because between the decision and the execution the defender moved. The read was right. The problem was that she did not change her mind.

His fix in the exercise is deceptively small. The defender must move late. "they want to move while the player is on the ball." If your defender commits while the ball is still travelling, you are training the prescan. If your defender commits after the first touch, you are training the on-ball adjustment. Same exercise, completely different skill, and the difference is a two-second timing instruction to one player.

He takes it further at the higher end, coaching the defender to actively deceive: sit in one place during the prescan phase so the attacker commits, then shift the moment the ball arrives. Nasty, and exactly what Saturday does to your players.

The second Noall idea worth stealing is lowering the technical bar to raise the cognitive one. His warm-ups are frequently not hockey at all. Handball, rugby ball, tennis ball, bibs. "if you take the level of technique needed lower, the possibility to train decision making is higher." If a player is fighting the ball, they have no attention left over for the picture. Take the stick away for ten minutes and you can work on scanning and space at a level they cannot yet reach with a ball on the stick.

And for the youngest groups, his rule on confidence is worth writing on the back of your session card. "make it so easy to succeed that you can't fail." Six against two. Five against one. Ridiculous overloads, on purpose, so that the habit of looking gets rewarded every single time before you make it hard.

The underlying principle is worth naming, because it applies at every level and not just to beginners. Learning happens fastest in the gap just beyond current ability, close enough that the player has to reach for it, not so far that they thrash. Too easy and they coast on autopilot. Too hard and they stop trying to solve the problem and start trying to survive it. Almost every complaint about a small-sided game not working is really a complaint about that gap being set wrong, and the fix is nearly always numbers or space rather than a whole new game.

So the game is running. What are you doing?

Mathias Becher's contribution here is a diagnostic checklist, and it stops you from giving the wrong feedback confidently. When a decision goes wrong, he works through four possibilities in order. Does the player not know the option exists? Do they know it but lack the skill to execute it? Do they have both but fail to recognise the situation as it happens? Or can they do all of that in training and not under match pressure?

Four completely different problems. Four completely different responses. And most of us skip straight to the first one and explain something the player already knew.

His warning about video is one of the most useful things in the archive for anyone who reviews footage with juniors: "this is the view from the video tower. But at the same time, this is a view from the player in this situation. And it is a big difference." From above, the free player is obvious. At stick height, with a defender on your shoulder, it is not. If you have ever wondered why a player agrees with you on Monday and does the same thing again on Saturday, that is your answer. They did not have a knowledge problem. They had a perception problem, and you coached the wrong one.

Becher is also refreshingly honest about the limits of all this. On the whole topic of teaching game intelligence: "If you're waiting for some simple explanations or magic exercises to train Game Intelligence easily, you will be a little bit disappointed." There is no drill. There is only environment, principle and repetition.

The principles he does offer are simple enough to survive contact with a U14 brain. Forward first: "go forward first, whenever you can, and side roads second." And the one I like most, the boss and the manager. The player on the ball is the boss and makes the decision. Everyone else is a manager, and the manager's job is to give the boss a good option. "the players without the ball, they have to be really, really active to give the boss, the player with the ball, a really good option that they get the ball and they become the new boss."

That single reframe fixes more small-sided games than any constraint I know. When a 4v4 goes flat, it is almost never because the boss is bad. It is because the three managers have stopped working.

There is also a way of giving live feedback inside these games that is worth practising deliberately, because most of us are bad at it. Short, vivid, in the moment, and specific enough that it locates the mistake and the fix in the same breath. Not a paragraph. A phrase. And the flip side, which coaches forget almost universally, is to mark the moment a player gets it right. A loud "there it is" at the exact instant something works does more for the next twenty minutes than any amount of explanation afterwards. Then, the second they can do it, raise the bar. Good, okay, now do it with a defender arriving late.

There is a nagging worry underneath all game-based coaching, which is that if we design everything and constrain everything, we sand the edges off the players who see things nobody else sees.

Tin Matkovic's answer is that structure and creativity are not rivals. He frames the shift in the game itself: "I think in the past people were much more keen on having this clarity of everything what's being done and also having a control. I would say then now in these times when we have a little bit of a more of okay, I'm gonna let my players also find their own way of doing it. As long as it's like working in a big picture."

The qualifier at the end is doing a lot of work. Freedom inside a framework, not instead of one.

His practical example is the sacred cow of youth coaching, the left foot. Most of us teach attack the left foot and leave it there. Matkovic's version is more honest: "you need to let them understand that if they attack the left foot they might have bigger percentage of success rate because it's much more common and harder to defense. But you should encourage them to do other stuff as well." Teach the percentage play, then explicitly ask them to try the other side, because a player who only ever goes one way is a player the opposition solves in ten minutes.

Noall lands in the same place from a different angle, and pushes back hard on the idea that decision training kills instinct. His view is that scripted exercises kill instinct, and decision-rich ones feed it: "Not every exercise is an a, b, c, d exercise... No. Let them decide."

Which brings us to feedback. Noall's is the line I would put on the wall of every clubhouse: "I really like to hear coaches... who shout, great decision. Unfortunately, it didn't work or great decision. Next time, it will work." If you only ever praise outcomes, you are training your team to take the safe option. Praise the read, and they keep looking.

There is a self-fulfilling quality to this that is easy to underestimate. Players tend to grow into the version of themselves you reflect back at them. Coach a group as though they need telling, and they will learn to wait to be told. Coach them as though they are capable of reading the picture for themselves, and a surprising number of them start doing exactly that.

Here are two designs that sit at opposite ends of the space and time spectrum. Run them in the same session and you cover both decision windows.

Objective. On-ball decision making under immediate pressure. Skill execution when there is no time to think, and the ability to change your mind after the first touch.

Setup. 4v4 in roughly 25 by 20 metres, with two small goals on each end line, no keepers. Small goals rather than a full goal, because you want them looking for the pass and the angle rather than smashing at a target. All four phases live: turnover means play continues immediately, no resets, no dead balls. Coach feeds a new ball the instant one goes out.

The constraint that does the work. A goal counts for one point. A goal scored within five seconds of winning the ball counts for three. Nothing else. You do not need to say the word transition once.

Progressions. Start there and let it settle for four minutes. Then add: the receiving player may not take more than two touches inside the central third of the box. Then, for older or stronger groups, add a single floating joker who always plays for the team in possession, which turns every moment into a 5v4 and forces the defending four to organise rather than chase.

What to look for. Body shape on the receive, because in this density there is nowhere to hide a bad first touch. Whether players receive on the open side or square up and get closed down. Whether anyone actually changes their mind after the touch, or whether they commit to the pass they decided on two seconds ago and get intercepted.

What the coach watches, and only this. The first touch and what it makes possible. Not the pass, not the finish, not the press. If the first touch opens the pitch, say so loudly. If it closes it, say nothing and watch whether they work it out. Enrich's warning about noise applies here more than anywhere: "just let them play under principles and under some guidance. But be careful of not being too noisy from the bench." In a game this fast, a coach shouting is simply a coach adding another defender to the picture.

Duration. Sets of about two minutes with equal rest, four to six sets. Enrich's guidance on total volume is worth respecting: "normally use the small sided game for a period of 15 minutes. I would say 15 minutes, no more than that, 15, 20 maximum." Past that, they stop solving and start repeating.

Objective. Prescanning, forward-first decision making, and the long transfer. The opposite problem: too much space, too many options, and the discipline to pick the right one early.

Setup. 5v5 on a long, narrow area, roughly 55 by 25 metres, running the length of the pitch rather than across it. Goals with keepers if you have them. The shape is deliberate and comes straight from Enrich's point about narrow pitches creating the incentive to actually use the width, because here the width is genuinely scarce and therefore genuinely valuable.

The constraint that does the work. Progressive scoring. Entering the attacking quarter with the ball under control is one point. A shot on goal is two. A goal is three. Points accumulate, so a team that enters four times and scores nothing still beats a team that scores once from a scramble. You are paying for process.

Progressions. Begin with a free game to establish the shape. Then add the prescan trigger: the receiving player must call the name of the teammate they intend to play to before the ball reaches them, and if they change their mind after the touch, that is fine and expected. This is the Noall separation made literal. They commit to a prescan read out loud, then have to live with, or consciously override, that read once they have the ball.

For the third progression, add a single defender who is instructed to hold position while the ball travels and only commit once the receiver has taken a touch. That one late-moving body converts the whole exercise from a prescan game into an on-ball adjustment game.

What to look for. Head turns before the ball arrives. Whether the forward pass is even considered before the square one. Whether the far-side player is offering width or drifting in to watch the ball. This is where Becher's boss and manager language earns its keep, because in this much space, poor off-ball work is instantly visible and instantly fixable.

What the coach watches, and only this. Off-ball movement. Not the pass quality, not the finish. Ask yourself one question repeatedly: at the moment the ball carrier looked up, how many genuine options had been offered? If the answer is one, your managers are asleep, and that is your feedback for the whole block.

Duration. Longer sets than the box, six to eight minutes, because the physical density is lower and the tactical picture takes longer to develop.

Both of those games fall apart the moment you start narrating them.

The when to stop coaching piece offers a test I now use constantly, and it is embarrassingly effective. Before you blow the whistle, ask yourself whether the game would self-correct in the next two or three minutes without you. If the honest answer is probably yes, stay quiet.

Robert Noall's answer, when asked directly how much information he gives players during decision-making work, was two words: "Zero. Before and after a lot." During the game he uses trigger words only, mostly about ball position. He does not tell them where the pass is.

The reason this works is not just that silence is polite. Learning sticks when the player generates the answer rather than receiving it. One solution a player works out for themselves is worth a dozen you hand over, because the act of retrieving it is what builds the memory in the first place. Which is why the most useful thing you can do at the end of a block is not a summary but a question. What did you see? What changed as the game went on? What would you do differently next time? Let them do the reconstructing. The better the question, the better the answer, and the answer they build themselves is the one that is still there on Saturday.

That does not mean silence is always right. Letchford is pointedly against the coach who stands and watches and never intervenes, calling it not particularly helpful and not great coaching. The distinction is between stopping because something specific and repeated genuinely needs addressing, and stopping because you saw something and felt the urge to say it.

And if what you want is players who keep trying, Becher's cultural point is the foundation everything else sits on: "When there's a culture where there's no right or wrong, there's good and better, it helps to make the players more confident." Good and better, not right and wrong. Players who are frightened of the wrong choice play the safe ball every time, and the safe ball never transferred to anything.

1. Design the question, not the drill. Before you draw a single cone, decide what decision you want the player to face. Then build the space, the numbers and the scoring so that decision comes up over and over, and keep all four phases of the game inside the exercise. If your game does not naturally produce the behaviour you want, that is a design problem and not a player problem.

2. Train both decision windows on purpose. Know whether the exercise is asking players to read the picture before the ball arrives or to change their mind after the first touch. The difference is often nothing more than when your defender moves. Train the prescan, then deliberately break it, because Saturday will.

3. Pick one thing to watch and let the rest go. One focus per game, feedback only on that focus, and praise the read rather than the result. The urge to fix everything you can see is the single biggest reason good small-sided games stop working.