Why your players look great in drills (and fall apart in matches)

The drill looks perfect.

Twenty cross-court forehands in a row. Clean contact, good depth, a rhythm you could set a metronome to. You watch from the baseline and think, they’ve got it.

Then the match starts, and it’s gone.

If you're a tennis coach like me, you probably recognize this. And you start asking yourself questions:

I feel like the answer is rarely in the player. When I look at how people learn movement, something clicked: the problem is often in the drill. Here are a couple of reasons why.

1. Performance: a good drill looks like learning

Soderstrom and Bjork (2015) point out that during instruction, what we can observe and measure is performance, and that performance is often an unreliable index of whether learning has taken place. They review research showing that improvements in performance can fail to yield significant learning.

A tidy drill is the clearest example. The ball comes to the same place, the player knows what’s coming, and the success rate is high. It looks like progress. And to be fair, part of it is. But it tells you how well the player hits the drill, not how well they’ll hit in a match.

2. Blocked practice: same ball, same place

Hitting the same shot over and over is what researchers call blocked practice, and Schmidt and Bjork (1992) note that it resembles what we usually call drill. In an experiment they describe, people practiced three tasks either blocked or in random order. The blocked group performed better during practice. But on a retention test 10 days later, the random group did better, whether that test was blocked or random.

That is uncomfortable for a coach, because the blocked session is the one that looks best on the day. A match is, however, much more unpredictable and random.

3. The information is missing

In a match, the ball tells the player what to do. Where it’s coming from, how fast, how high. In a basket drill, much of that is removed: the feed is predictable and the decision is already made.

Pinder et al. (2011) propose the term representative learning design: practice should keep the functionality of the performance setting and the fidelity of the actions the player will need. I’d argue that’s a useful test for any drill. If the player doesn’t have to read something and choose, you are practicing the hit without the match.

4. Nothing is at stake

Drills usually have no score and no consequences, and matches have both. A meta-analysis of 14 studies found that practicing under simulated pressure improved performance under pressure, with a medium effect (g = 0.67 after removing one outlier). The authors suggest that coaches create pressurized training environments rather than relying on more training (Low et al., 2021).

I wrote about the player’s side of this in why players play worse in matches than in practice.

Together, these four make the drill a good place to look good, and a poor place to prepare. And that is a darn difficult thing to see from the baseline.

The environment you build changes what players learn

In a five-week study with beginning tennis players, those who practiced with a standard ball on a standard court had fewer hitting opportunities, and lower hitting success as a result, than those who practiced on scaled courts (Farrow & Reid, 2010).

The point is not that you should scale every court. It’s that the practice environment you set up shapes what your players learn, in ways that are easy to miss. For coaches working with young players who are aiming high, I’d say that matters even more: every session either brings them closer to the demands of match play, or it doesn’t.