You study something today, know it cold tonight, and by next week almost nothing is left. That's not your memory failing — it's exactly how every human brain works. In 1885, the German psychologist Hermann Ebbinghaus measured it for the first time by memorizing nonsense syllables and tracking how many he could recall over time. The result was the forgetting curve: a steep drop showing that we lose most of what we learn within days unless we do something to stop it.
But Ebbinghaus discovered something else, and that finding is the key to everything. Each time you review material just before forgetting it completely, the curve flattens: forgetting slows down and the memory lasts longer. This is the phenomenon we now call the spacing effect, and it's one of the most robust and replicated findings in all of learning psychology.
Why Cramming Doesn't Work
The way most of us study — packing all our review into the night before an exam — is called cramming, and it's the opposite of what the brain needs. It works for passing the test the next morning, but the information evaporates days later because it never consolidated into long-term memory.
Spaced repetition does the reverse: it distributes reviews over time, with ever-lengthening intervals. You review a fact the next day, then after three days, then a week, a month, three months. Each successful review reinforces the memory trace and lets you stretch the next interval. You study less in total and remember far more.
The goal isn't to review when it's convenient, but at the exact moment you're about to forget. There, each repetition is worth ten.
The Algorithm That Decides When to Review
Working out the optimal moment to review each fact by hand would be impossible: with a hundred cards, tracking what to review and when becomes unmanageable. That's why the SRS (Spaced Repetition System) exists — software that automates the scheduling.
The most influential is the SM-2 algorithm, developed by Piotr Woźniak for his SuperMemo program in the late 1980s. Its logic is simple and elegant: every time you review a card, you grade yourself on how well you recalled it. If you knew it fluently, the program multiplies the interval and won't show it again for a long time. If you hesitated or failed, it shrinks the interval and brings the card back soon. That way, hard material appears often and easy material gets out of the way.
Anki: SRS Within Everyone's Reach
SuperMemo was the pioneer, but Anki — created by Damien Elmes in 2006 — democratized spaced repetition. It's free on desktop and Android, open source, and has become the go-to tool for medical, language, and exam students worldwide. Its mechanics are straightforward:
- You make cards with a question on one side and the answer on the other.
- Each day, Anki shows you only the cards due according to the algorithm.
- After seeing the answer, you grade yourself: "Again," "Hard," "Good," or "Easy."
- That grade adjusts the next interval, which grows with each success.
The piece that makes Anki so powerful isn't only the spacing, but that it rests on active recall. Each card forces you to retrieve the information from memory before seeing the answer, instead of rereading it passively. The combination of active recall and spaced repetition is, according to the scientific literature, one of the most effective effective study techniques that exist.
How to Make Cards That Actually Work
The most common mistake is turning Anki into a dumping ground for whole paragraphs. A card with a five-sentence answer is impossible to grade honestly and exhausts you instead of teaching you. Good cards follow a principle Woźniak himself formulated: the minimum information principle.
- Atomic: one card, one fact. Instead of "List the causes of the French Revolution," make one card per cause.
- Concrete: avoid vague questions. "What does the mitochondrion do?" beats "Tell me about the cell."
- In your own words: rephrase the content in your own language. The effort of writing the card is already part of the learning.
- Bidirectional with care: in languages, a "dog → perro" card and a "perro → dog" card train different retrievals, but don't overdo it by making pairs of everything.
Where It Shines: Languages and Medicine
Two fields have adopted spaced repetition en masse, and not by chance. Language learning consists largely of memorizing thousands of words and conjugations — terrain where an SRS performs spectacularly against traditional vocabulary lists.
The other is medicine. The sheer number of facts a student must retain — drugs, doses, metabolic pathways, symptoms — is enormous, and it has to last for years. Huge collaborative decks for exams like the USMLE in the United States are now part of the standard study method at many schools, precisely because the forgetting curve is merciless with that kind of volume.
Daily Consistency Is the Real Secret
This whole architecture collapses without one condition: using it every day. Spaced repetition schedules reviews for specific dates; if you skip days, cards pile up and the system loses its meaning. A weekly binge recreates the very cramming problem you were trying to escape.
The good news is that daily sessions are short: fifteen or twenty minutes usually suffice once the system settles. Slotting them into a fixed block — at the start of the day, alongside coffee — helps them become a habit. Doing those reviews within a timed interval using a timer like Pomodomate keeps them from turning into an endless session and keeps the rhythm sustainable. The difference between knowing something for a week and knowing it for life fits inside those twenty daily minutes.
FAQ
How much time per day do I need for it to work?
For a reasonable card volume, fifteen to thirty minutes a day usually suffice. What matters isn't the length of each session but its regularity: twenty minutes every day beats two hours every other day. The system rewards consistency, not occasional intensity.
Why does Anki show me cards I already know?
Because the algorithm doesn't wait for you to forget before reinforcing the memory; it shows you material just before it starts to fade. Finding a card easy is a good sign: it means you're ahead of forgetting. Grade those cards "Easy" and the interval will grow, spacing them out further and further.
Should I download premade decks or build my own?
Building your own cards is almost always better, because the act of writing them is itself learning and ensures you understand the material. Premade decks can save time in fields with standard vocabulary (languages, anatomy), but they often contain poorly formulated cards. If you use one, prune it and adapt it to the way you think.
Does spaced repetition work for concepts, not just isolated facts?
Yes, but it takes more care in designing the cards. Asking for a concept's definition isn't enough; include cards that ask you to apply it, give an example, or explain the why. Spaced repetition consolidates what you already understand: first grasp the concept thoroughly, then use it to keep from forgetting.
