The Spacing Effect: Why Cramming Fails

In the 1880s, the German psychologist Hermann Ebbinghaus spent months memorizing lists of nonsense syllables and testing how fast he forgot them. The curve he drew — fast forgetting at first, then a slow tail — is still the first thing many people learn about memory. Less famous is his second observation: when he spread his relearning out over time instead of bunching it together, the forgetting slowed down.

More than a century of experiments has turned that side note into one of the most dependable findings in all of learning science. It is called the spacing effect, and it is the reason cramming is such a bad deal.

What distributed practice does

In 2006, Nicholas Cepeda and colleagues published a quantitative synthesis of the distributed-practice literature, pooling results across a large body of experiments spanning different materials, ages, and time scales. The verdict was consistent: for the same amount of total study, spreading sessions apart produced better long-term retention than packing them together.

There is a catch that explains why cramming survives. Massed practice — many repetitions back to back — often produces better performance during the session itself. You feel fluent, so you conclude it worked. Then the knowledge drains away far faster than it would have if you had spaced the same effort out. Dunlosky and colleagues' 2013 review reached the same conclusion from the applied side, naming distributed practice one of only two techniques worthy of their top high-utility rating.

How far apart should the gaps be?

The natural question is how big the gaps should be. Cepeda's group found that the best spacing depends on how long you need to hold the knowledge: the longer the retention interval, the longer the optimal gap between study sessions. Preparing for something months away rewards gaps of days or weeks, not hours.

For most learners the precise schedule matters less than the headline: almost any spacing beats none. Two one-hour sessions on different days will usually outperform a single two-hour block, even though the block feels more productive while you are in it.

Interleaving: spacing's useful cousin

A closely related tactic is interleaving — mixing different problem types within a single practice set rather than drilling one type at a time. In a 2007 study, Doug Rohrer and Kelli Taylor had students practice mathematics problems either in blocks (all of one kind together) or shuffled. The shuffled group performed worse during practice but markedly better on a later test.

Interleaving is not the same thing as spacing, but it shares the same counter-intuitive shape: the arrangement that feels harder and messier in the moment is the one that pays off later. Both work partly because they force you to reconstruct and discriminate, rather than coasting on momentum.

How Amistio paces a plan

Amistio Learn turns this into scheduling you do not have to manage by hand. When the planner builds a curriculum from your goal — and from any deadline you mention to Ami — it sequences skills so earlier material resurfaces in later assignments, instead of being taught once and abandoned. The progress agent tracks what you have mastered, and re-planning pulls weak or aging skills back into rotation.

Deadline-aware pacing is the same idea applied to time. Tell Ami your exam is six weeks out and the plan distributes practice across those weeks rather than betting everything on a final push — which is exactly the arrangement the spacing research favours.