Spaced Repetition 101: How It Works and Why It Beats Cramming

Spaced repetition is one of the few study techniques with genuinely strong evidence behind it, and it's also one of the most misunderstood; people hear "spaced" and assume it just means "study a little bit every day," which is true but misses the actual mechanism. Here's what's really going on, and why it matters for anything you're trying to remember longer than a week.

The forgetting curve is the whole story

In the 1880s, Hermann Ebbinghaus measured how quickly he forgot lists of nonsense syllables and found something that's held up remarkably well since: memory doesn't decay at a steady rate, it drops off fast at first and then slows down. Most of what you'll forget from something you studied once, you forget within the first day or two. The forgetting curve is steep early and flattens out later, which means the timing of your review matters enormously, and reviewing too early or too late both waste the effort.

Why rereading and cramming both fail the same way

Rereading your notes feels like studying because the material feels familiar, but familiarity is a recognition effect, not a recall one. You recognize the sentence because you just read it thirty seconds ago; that tells you almost nothing about whether you could produce the answer cold, on an exam, a week from now. Cramming has the same problem at a larger scale: massed practice (several exposures back to back, close together in time) produces a short-term bump that mostly evaporates within days, because the brain never had to do the effortful work of retrieving something it was at risk of losing.

What actually works: retrieval right before you'd forget

The technique that reliably beats both is deceptively simple: test yourself on something right around the point you'd otherwise forget it, not immediately after learning it and not so late that you've already lost it. Doing this repeatedly, with the gap growing each time you succeed, is spaced repetition. The "testing" part matters as much as the "spaced" part: actively trying to recall an answer and getting it right is a fundamentally stronger memory event than passively rereading the same fact for the fourth time.

• •Each successful recall should push the next review further out: a card you get right should come back in days, not tomorrow.
• •A failed recall should reset the interval, not just push it back slightly: "Again" means start close to the beginning, not "try in three days instead of five."
• •The scheduling algorithm most spaced-repetition software is built on (SM-2 and its variants) adjusts an "ease" factor per card based on how hard you found it, so an easy fact naturally drifts toward longer and longer intervals while a hard one keeps coming back more often.

How to approximate it without an app

If you're not using flashcard software, you can get most of the benefit with a simple manual schedule: review new material the next day, then three days later, then about a week later, then roughly every month after that, moving anything you get wrong back to the start of that sequence instead of skipping it forward. It's less precise than a real scheduler that tracks per-item ease, but it captures the core insight: review timed to just before you'd forget beats review timed by convenience.