Blocked vs random practice: which one helps an adult learner more?

The bucket of balls and the real game

Say you play tennis on Sunday mornings at the park. Before your doubles game, you take a bucket of balls to an empty court. Twenty forehands, then twenty backhands, then twenty serves.

By the end of each set, the ball goes where you aim. It is a good feeling.

Then the game starts. A ball comes to your backhand, then a short one, then a lob, then one at your feet. Shots that landed twenty times in a row an hour ago now sail long or hit the net.

You did hit them, with the same ball from the same place and nothing else to think about. The game never once asks for that. That is the gap between blocked and random practice, and it is one of the most studied questions in learning a movement.

Mixing wins big in the lab

The classic study is from 1979, by Shea and Morgan. People learned three movement tasks in either a blocked or a random order. Retention was tested after 10 minutes or 10 days. Random practice led to better retention when the test order differed from the order practiced. It also led to better transfer to new tasks, most of all for the hardest one.

A 1994 study took the idea to a baseball field. Hall, Domingues and Cavazos gave 30 college players 12 extra batting sessions over six weeks. The pitches were fastballs, curveballs and change-ups. The random group improved 56.7 percent. The blocked group improved 24.8 percent. The group with no extra practice improved 6.2 percent. The book is careful: that is ten players per group, and it is more than thirty years old.

There is also the catch you feel in your own sessions. In a 2001 keypad study by Simon and Bjork, blocked practice looked better while people did it and worse on a test a day later. People still predicted blocked would win.

Out on real courts, the edge shrinks

If the story stopped there, everyone should shuffle everything. It does not stop there.

In 2004, Brady pooled 61 studies and found an overall effect of mixing of .38. It split sharply by setting: .57 in basic lab research and .19 in applied research, the kind closer to real teaching. A 2007 review by Barreiros and colleagues looked only at applied settings. Of those studies, 60 percent failed to find a benefit from mixing.

The most recent analyses, by Czyż and colleagues in 2024, pooled 54 studies of retention. Lab studies showed a large benefit from random practice. In applied studies the effect was small and not significant. In their words, the benefit in applied settings "was almost negligible."

Age changes the picture too. Brady found an effect of .50 for adults and .10 for younger learners. A 2019 review of 25 papers on children found "mainly no or conflicting evidence." For older adults, the two newest lines of work point in opposite directions, and nobody has settled it.

So the fair summary is plain. The lab effect is large and well replicated. On courts, in gyms and in practice rooms, every recent analysis finds it small and uncertain.

Block it until it works, then shuffle it

There is a way to use both. In 2010, Porter and Magill tested a schedule that started blocked, moved to a fixed rotation (A, B, C, A, B, C) and ended in random order. For golf putting, novices on this rising schedule kept more. For basketball passes, they did better on retention and transfer than people who stayed blocked or stayed random.

A 2024 sprint hurdles study by Makaruk and colleagues found a mixed result: the same retention, but better transfer for the rising schedule. Two more studies hint at when to switch. In a 2008 study of 48 people by Simon, Lee and Cullen, moving to a new task after one success gave results like random practice. And a 2022 darts study of 120 people found mixing helped when practice kept errors low.

The Practice Lab turns this into two drills in its chapter Blocked or Mixed?, both badged PROMISING:

• Block First, Then Mix. Repeat a new skill in a block until you can do it three times in a row. Then add it to a fixed rotation. When that feels steady, go random. After a good attempt, switch skills. After a miss, stay and try again.
• The Shuffle Session. Pick two or three related skills you can already do sometimes. Mix them so the same one rarely comes twice in a row. Expect it to feel worse. Compare a week of mixed sessions with a week of blocked ones, using cold scores the next day.

Both drills lean on the Next-Day Test from the chapter Practice Is Not Performance, because the session itself will mislead you. The book also flags music: a 2024 study of advanced violinists found blocked practice worked best at a delayed test. If mixing wrecks a passage, send it back to blocks.

What we like is that the chapter refuses to oversell. Mixing is a reasonable, cheap experiment. Your own next-day scores decide whether it stays. For why a smooth session can fool you in the first place, see practicing it perfectly and then messing it up.

Some drills ask for more reps or a faster tempo, so the book's one safety rule applies here too. Pain is not part of the method. Stop and ask a professional.