Your Drill Session Is Lying to You

Picture the Saturday morning drill session. You and your buddy, crosscourt dink rally, both of you locked in. Ball after ball into the same square foot of kitchen. Forty in a row. Sixty. Eighty. You’re in a trance. You look at each other after and go, “man, our hands are on today.”

Tuesday night league. First real dink exchange of the first real game. Ball comes in a foot wider than you expected and with a little sidespin on it, and you pop it up like a beach ball at a Jimmy Buffett concert.

Here’s the thing nobody wants to hear: that Saturday session didn’t fail you. It did exactly what it was designed to do, and what it was designed to do was make you feel great on Saturday. Whether you could still hit that ball three days later under stress, off a shot you hadn’t seen coming, was never part of the deal.

Performance Is Not Learning

Motor learning researchers have a distinction that ruins a lot of people’s week when they first hear it. Performance is how well you execute right now, under current conditions. Learning is a durable change that shows up later, in different conditions, when nobody warned you it was coming.

These two things are not just different. They frequently point in opposite directions. The practice conditions that make you look best today are often the ones that leave the least behind tomorrow.

The classic demonstration goes back to a 1979 experiment where college students learned a set of movement patterns. One group practiced them in blocks — all of task A, then all of task B, then all of task C. The other group got them scrambled in random order. During practice, the blocked group looked dramatically better. Smoother, faster, more accurate. It wasn’t close.

Then everyone came back later for a retention test. The scrambled group won. And on a transfer test — a version of the task nobody had practiced — the scrambled group won by more.

The blocked group had spent their practice getting good at practice. The random group had spent their practice getting good at retrieving the skill, which turns out to be the thing you actually need to do in a game.

This is called the contextual interference effect, and it has been reproduced hundreds of times in labs.

The Honest Caveat, Because We’re Not a Supplement Company

You are going to see people online turn this into “BLOCKED PRACTICE IS USELESS, RANDOMIZE EVERYTHING.” Don’t be that person, because the applied sports research is a lot messier than the lab research.

When you take contextual interference out of the lab and into real sports settings, the effect gets smaller and less consistent. It depends heavily on two things: whether the learner already has a workable version of the movement, and how complex the task is. Dump a total beginner into fully randomized practice and you don’t build a resilient skill — you build flailing. They never get enough clean reps to form the movement in the first place.

So the actual rule is not “random good, blocked bad.” It’s:

Blocked to build it. Mixed to keep it.

Most rec players get the first half right and skip the second half entirely. They found the blocked phase comfortable — because it is comfortable, that’s the whole problem — and then they just… stayed there. For four years.

The Four Rungs of the Ladder

Think of practice organization as a ladder, not a switch:

Blocked — the same shot, same spot, same feed, over and over. Great for the first week of a new movement. Great after a technical change. Terrible as a permanent home.
Serial — a fixed repeating sequence. Dink, dink, speed-up. Dink, dink, speed-up. You know what’s coming, but you’re switching between skills.
Interleaved — a small menu of shots, and the incoming ball tells you which one. Now you’re making a decision before you make a swing.
Random — genuinely unpredictable. Location, pace, spin, height, all live. This is the closest thing to a match that isn’t a match.

The reason interleaved and random work is that in a real game, the swing was never the hard part. The hard part is the 200 milliseconds before the swing, where you read the ball and pick a shot. Blocked drilling deletes that part of the skill entirely and then you wonder why it’s missing on Tuesday.

The 70% Rule

Here’s a calibration tool that will change how you run a session: aim to make roughly 70% of your drill balls.

Not 95%. If you’re making nearly everything, congratulations, you’ve found a difficulty level your nervous system has nothing to learn from. Make it harder — add pace, add spin, add movement, add a target, add a consequence.

Not 40%, either. If you’re missing more than half, your errors have stopped being informative. You’re not adapting, you’re just spraying. Back down a rung on the ladder for ten minutes and rebuild.

Seventy percent feels bad, by the way. It feels like a session that isn’t going well. That feeling is the point. If you leave the court feeling like a hero every single time, you’ve been running a highlight reel, not a practice.

Rebuild Your Hour

Here’s a 60-minute session structured the way the research actually suggests, instead of the way it feels good:

0–10 min — Blocked, on purpose. Pick one thing you’re technically fixing. Crosscourt dink, or third shot drop, or nothing. Clean reps. This is the only part of the session that should feel smooth, and it has a hard stop.

10–25 min — Serial. Fixed patterns with switching. Third shot drop → move to kitchen → two dinks → reset. Same sequence every time. You’re now stringing shots together, which blocked reps never teach.

25–45 min — Interleaved with constraints. This is the meat. Constraints are the cheat code — they force better decisions without anyone having to coach. A few that punch way above their weight:

String over the net. Run a string 12–18 inches above the tape. Dinks and drops must go under it. Instantly kills the floaty attackable dink.
Tape line in the transition zone. Returns must land past it. Short returns are a top-three rec player disease and this fixes it in one session.
Three-dink rule. You may not speed up until three dinks have crossed. Cures the itchy trigger finger.
Feeder chaos. The feeder mixes depth, pace, and spin without announcing it. Their job is to be a jerk, politely.

45–60 min — Random, and keep score. Play out points from a specific start — everything begins with a third shot, say, or everyone starts at the kitchen. Score them. Score changes physiology, and physiology is what you’re training for.

Your Drill Partner Is Going to Hate This

Real talk: cooperative dinking is pleasant. It’s meditative, it’s social, it feels like teamwork. Interleaved constraint drilling at 70% success is none of those things. It’s stop-start, it’s frustrating, and someone has to be willing to hit their partner a ball they can’t handle.

Sell it like this: you’re each allowed to be bad at this for three weeks. That’s the deal. Because the other thing the research says is that the benefit shows up later. During the practice block, you will look worse than you did doing 80-ball dink rallies. That is not a bug, that is the mechanism. If your new practice structure isn’t making you look worse right now, it probably isn’t doing anything.

Then you go play Tuesday league, that ball comes in a foot wider than you expected with a little sidespin on it, and this time your hands already know what to do — because you’ve been in that exact uncomfortable spot forty times, on purpose, in a drill you didn’t enjoy.

That’s the trade. Smooth practice and a fragile game, or ugly practice and a game that holds up when it counts.

Go make it worse.