Teach It or Let the Game Teach It? When Direct Instruction Is the Right Call

The instruction that helps a beginner can make a more experienced learner worse. It's one of the better-established findings in education research, and it has an awkward consequence for coaches on both sides of the drills-versus-games argument: there's no fixed right amount of telling. It depends on who's in front of you.

Researchers call it the expertise reversal effect. Detailed, step-by-step guidance that beginners need becomes unnecessary for learners who already know the basics, and can actively get in their way (Kalyuga, Ayres, Chandler & Sweller, Educational Psychologist, 2003). That's classroom learning, not sport. But anyone who has coached a mixed group of eight-year-olds has seen a version of it: the explanation that finally clicks for one player is noise for the player next to them who could already do it.

This is the second step in the series. The scanning piece was about whether a player sees the picture. This one is about whether they can do what they see, and what to do when they can't.

Two arguments that are both partly right

The rest of this site leans toward game-based coaching: build the environment so the right decision is the obvious one, and talk less. There's decent evidence for that. A 2025 systematic review of the nine studies that compared game-based, constraints-led practice with traditional drill-based practice in team sports found the game-based approach ahead on tactical skills in most outcomes, and on technical skills, no outcome favored drills (Bromilow et al., Sports Medicine – Open, 2025).

There's also a well-known argument pointing the other way. Beginners, it says, learn better from explicit guidance than from being left to figure things out, and the advantage of guidance only fades once learners know enough to guide themselves (Kirschner, Sweller & Clark, Educational Psychologist, 2006).

Both can be true, because they're mostly about different players. A game teaches decisions to players who can already do enough to make them. It struggles to teach a skill to a player who can't yet do it at all, because the game never gets to the decision. The ball bounces off their shin first.

The working-memory research on this site says the same thing from another angle. In the study behind Why Your Soccer Practice Feels Like Chaos, five instructions before each block of practice helped children with more working memory and made children with less of it worse (Buszard et al., Frontiers in Psychology, 2017). Same instructions, same age, opposite results.

Choosing badly, or can't do it yet?

When a game keeps breaking down, one question is worth asking before changing anything: are players choosing badly, or can they not yet do the thing?

It usually shows if you watch a few breakdowns in a row.

It's a choosing problem when the player can do the action in a quieter moment, the option was there and they took a different one, or they succeed as soon as the game gives them a bit more time. That's a STEP problem. Change the space, the numbers or the scoring so the choice you want is the one that pays (the STEP piece covers how).

It's a can't-do-it-yet problem when the action fails even with nobody near them, the behavior you designed for never appears because the technique ends every attempt first, or the whole group breaks down at the same technical moment. No change to the scoring fixes that. That's a teaching problem.

In the Switch of Play game from Why Your Soccer Practice Feels Like Chaos, the two look the same from the sideline: the ball never gets switched. But a player who sees the open far side and plays it short anyway is choosing, and making switched goals count for more is one way to change that choice. A player whose pass can't reach across the 25 yards can't do it yet, and no scoring rule will fix that. A few minutes of passing over that distance, then back into the game, might.

That split is a practical rule of thumb, not something a study has tested. But it helps avoid two common mistakes: drilling a group that could already do the skill and needed a better game, and redesigning a game for a group that needed the skill taught.

What "teach it directly" doesn't have to mean

Direct teaching gets pictured as lines, laps and a long explanation. It doesn't have to be any of those.

• Simplify the Task a long way first. More space, fewer defenders, a bigger target. Several of the game-based programs in the Bromilow review used task simplification themselves, so this isn't a departure from game-based coaching. It's part of it.
• Show it. Demonstrate is one of the twelve interactions in How Much Should You Actually Say?, and for a skill a player has never done, it's often the fastest route. It also doesn't depend on a young player holding a description in their head while they move.
• Give one cue, not five. A plain cue can do as much as a clever one. In a pooled set of experiments with 173 boys under 18, a neutral cue like "sprint as fast as you can" did as well as crafted external-focus and analogy cues in seven of eight comparisons (Moran et al., PLOS ONE, 2023).
• Repeat without repeating. Movement scientist Nikolai Bernstein described practice as "repetition without repetition": the goal stays the same while the movement adapts each time (Bernstein, The Co-ordination and Regulation of Movements, 1967; a 2020 review of the idea explains it well). For coaches, that's a case for varying the pass speed, angle and distance in a focused block rather than feeding the identical ball twenty times.
• Keep it short, then put it back. The point of a focused block is to get players to the level where the game can take over. Then go back to the game, so the skill gets used for a reason.

Where the evidence runs out

The limits are worth having in one place.

• The two sources arguing for guidance, Kirschner and Kalyuga, are about classroom learning. The logic travels to sport, but the studies weren't done on a field.
• The Bromilow review is nine studies, mostly small, and only two with players around ten or younger. Its authors say plainly that more research is needed, including across different ages.
• A 2023 systematic review looked specifically at the soccer research on feedback, focus of attention, demonstration and practice schedules, the everyday choices a coach makes when teaching a skill. Its conclusion was that the findings, in most cases, aren't unanimous (Pacheco et al., International Journal of Sports Science & Coaching, 2023).
• The choosing-versus-can't-do split is practical judgment. I haven't found anyone who has tested it as a rule.

So neither camp gets to call it settled. What the evidence supports is narrower and more useful: games hold up well against drills, even on technique, and beginners need more guidance than players who already know the basics. Where a given player sits between those two is something you judge in the moment.

Next session

One thing to try: when a game breaks down, hold off changing anything straight away. Watch the next three breakdowns and ask of each one: couldn't, or didn't? If it's mostly didn't, change the game. If it's mostly couldn't, take five minutes, teach it, and put them back in.

Common questions about teaching skills directly

Is it wrong to use drills with young players? No. In the studies that compared the two head to head, drill-based practice didn't come out ahead of game-based practice, even for technique. But a short, focused block of repetition can be exactly right for a player who can't yet do a skill well enough for a game to teach them anything about it. The problem is drilling by default, not drilling at all.

How do I know whether a player needs a skill taught directly? Watch whether the action fails even without pressure. If a player can do it when nobody's near them and only fails in the game, they probably need a better game. If it fails every time, pressure or not, they probably need it taught.

What is the expertise reversal effect? A finding from education research: guidance that helps beginners learn can become unnecessary, or even unhelpful, for learners who already know the basics. For coaches, it's a reason to adjust how much you tell each player rather than picking one amount for the whole group.

Does game-based coaching mean never demonstrating or instructing? No. Demonstrating and instructing are both on the menu of interactions this site uses, and several of the game-based programs studied in the research still used task simplification and focus cues. Game-based coaching is about where most of the learning happens, not a ban on telling.

This is part two of Before the Decision. Start with Scanning: Why Information Comes Before Decisions, or continue with First Touch and Receiving. It also builds on Why Your Soccer Practice Feels Like Chaos, which introduces the "what a game can't teach on its own" problem, and The Fundamentals of Soccer Coaching, the pillar for this topic.