Harnessing the spacing effect to creating more efficient gaps

Why quiet intervals, sleep and early returns beat cramming in both training and teaching

Since becoming obsessed with running,¹ one of the counter intuitive truths of training I’ve had to get to grips with is that having a least one rest day per week makes me better at running. That’s right, on the day that I do nothing, my endurance and speed improves. Why is this? Well, my fairly weak understanding of the biology is this: I stress my system, then muscles rebuild stronger during the recovery period. If I run again before the muscles have rebuilt, I get weaker, not stronger. On a similar note, one of the pieces of advice I’ve been given is that 80% of my running should be ‘easy’ - that is to say, a low-intensity effort of a short to moderate duration. At first I was tempted to pick up the pace but since coming to understanding that going slower builds endurance, loosens up sore muscles and flushes out waste, I’ve been trying to rise the challenge of doing less to get better.²

What if the process of learning works in a similar the same way? Not - obviously because ‘the brain is like a muscle’ (it’s really not) but because taking time out from learning reduces interference and grants the conditions in which new information can be more strongly integrated into prior understanding? This is anything but a new idea; it is the very essence of the spacing effect, the robust finding that leaving intervals between trying to master information and processes improves our ability to retain and transfer whatever we’re trying to learn.

In a recent post, Carl Hendrick interrogates the two main competing theories about how spacing works. The Rest & Recovery Theory treats spacing - the gap between learning events - as an opportunity to top up drained mental reserves. The other, The Mental Rehearsal Theory sees spacing as an opportunity, possibly unconscious, for the mind quietly turns the material over and strengthen its place within existing schema.

The distinction seems to matter because it can shape how we timetable learning. Each account predicts a trade-off: if gaps are too short, you get little benefit from pausing; too long and you pay a forgetting penalty when you return. Somewhere between those extremes sits a sweet spot of ‘desirable difficulty’³ where the next learning interval is just hard enough to strengthen memory but not so difficult that our ability to retrieve collapses. This just how training blocks are organised: running a hard interval session too soon and you blunt the adaptation and risk injury, wait too long and you detrain.

On a rest and recovery view, optimal spacing is the point at which resources have rebounded and interference has fallen, but the representation remains accessible, whereas a mental rehearsal view suggests that the gap gives time for consolidation and replay. Those processes are not instantaneous. They need a stretch of low-interference wakefulness - and, crucially, sleep - to stabilise and integrate what was learned. If you return before that machinery has done its work, you interrupt it and gain little. If you leave it so long that retrieval fails, you also gain little, because strengthening depends on successfully bringing the trace back to mind. The best interval is the one that allows enough offline processing to make the next retrieval effortful and fruitful without tipping into guesswork.

Both stories imply a ridgeline.⁴ Early on, lengthening the gap helps by reducing fatigue and allowing rehearsal. Beyond the peak, further delay hurts because access falters and relearning dominates. Prior knowledge, task complexity and the ‘quality’ of the gap all shift the peak. As schemas grow and interference drops, longer gaps become viable. When demands are high and knowledge is shallow, shorter, more frequent returns keep retrieval possible while the system does its quiet work between times.

Time - the space between learning intervals - is only a proxy; what matters is what happens in the gap. Quiet intervals that permit replay and sleep are likely to boost future performance; noisy, unrelated tasks are more likely to depress it, even if the length of the space is the same. As knowledge chunks into schemas, the peak shifts right and widens. Novices need shorter, earlier returns; with fluency, longer gaps become both safe and useful.

the same is true with running. Two days between hard sessions are not equal if one of them is spent sleeping well, eating properly and jogging easily, while the other is spent on your feet all day and partying hard into the night. Although the clock reads the same, your body will feel very different. Likewise, twenty-four hours between lessons are not equal if one contains sleep and a calm review, while the other is rammed with unrelated demands and dopamine fizz.⁵

If we think about spacing as if we had to pick between two mechanisms, we might be in danger of creating a meaningless opposition. Either learning stalls because attention and working memory are depleted, or it accelerates because the brain quietly rehearses. But the classroom is not obliged to honour our dichotomies. Just as muscles rebuild when we take a break from exercise, so our minds need recovery time in order for rehearsal to produce more flexible and durable memory.

The key point - both for physical and mental exercise - is not to undermine the benefits of recovery time by filling it with stuff that risks undoing the benefits of the spacing effect. If we clog the gap with busywork, novelty or stress, we’re likely to swap consolidation for interference. The same is true in training. Junk miles blur the stimulus you were trying to adapt to, just as cramming an unrelated task into the last five minutes of a lesson may blunt the gain from what you have just taught. Protect the interval. Keep it quiet and aligned. Let sleep do its work. Then return while retrieval is still possible and effortful. If we respect the gap, the system improves. If we cannot leave it alone, spacing collapses into massing and you end up rebuilding rather than strengthening.

Practical implications

Running has taught me that progress is made in the gaps. As I stress my system, I need to then get out of the way: cool down, keep the signal clean, sleep, and come back before the adaptation fades. Trying to cram extra effort into every gap risks over training and injury. Building memory seems to works similarly. Here are five suggestions for effective spacing adapted from what I’ve learned from trying to be a better runner:

1. Memories are most fragile after new content has been taught. Treat the minutes after new teaching like the cool down after a hard session. You would not finish intervals and then sprint to the car park. You jog, breathe, let the system settle. A single, quiet retrieval prompt is that jog. It keeps the signal clean and stops ‘junk miles’ of admin from trampling over what you want to keep.

2. Return early, then stretch, just as you would in a training block. After a hard workout you take an easy run the next day to lock in the adaptation, then you lengthen the gap between quality sessions as fitness builds. Do the same with curriculum content. Revisit within a day or two while recall is possible, then increase the spacing as knowledge firms. Beginners need frequent, gentle outings. Stronger runners can handle longer intervals between efforts.

3. Set homework like a recovery run, not a race. Ten calm minutes that keep form and rhythm beat a sprawling project that spikes effort and steals recovery. Keep it related to what you taught so you nudge the same pathways rather than dragging attention up a new hill.

4. Use sleep like an endurance athlete uses the night before a long race. Sleep is where the body cashes in the gains from training and the mind does the same for memory. If you can, place the next learning interval after an overnight break so consolidation has had time to do its quiet work.

5. Plan spacing in the curriculum the way a coach periodises a season. Map the big races, set the key sessions, and tune the rest of the week around them. Track success rates as you would heart rate or pace. If retrieval fails, the gap is too long and needs trimming. If it is effortless, the gap can grow. The aim is the same on road and in class: stress, recover, return, improve.

1. A reminder that you can sponsor my forthcoming half marathon for which I’m raising money for Cancer Research.

2. When I say ‘less,’ I’m currently running six days per week but only doing an intense interval session once as well as a weekly longer run. Most of my running is deliberately slower than feels comfortable. (I’m not joking when I say I’m a bit obsessed as I’m averaging well over 50km a week.)

3. A reminder that a desirable difficulty induces struggle but must result in success.

4. Cepeda et al. (2008) taught 1,350 participants facts, inserted a single review after gaps of up to 3.5 months, then tested up to a year later; performance traced a non-monotonic ridgeline in which the optimal interstudy gap grew with the test delay but shrank as a proportion of it, from about 20–40% at a one-week delay to roughly 5–10% at a one-year delay. For teachers, this yields a simple rule of thumb: pick your target retention date and back-plan the first recap near the ridge. If you want recall in a week, revisit after about 1–3 days; if you want recall at year’s end, revisit after about 2.5–5 weeks, then lengthen intervals as knowledge firms.

5. In case anyone cares about my extended running analogy, there is one practical difference. Running has hard lower bounds set by tissue repair. Tendons and collagen remodel on their own slow timetable, which limits how far recovery quality can speed things up. In learning the main constraint is interference. Improve the quality of the gap and gains can appear quickly, especially once some knowledge is in place.