Spaced Repetition: The Science Behind Never Forgetting a Word

You learned 50 Spanish words last week. How many can you recall right now? If you are like most people, the answer is somewhere between "a few" and "almost none." This is not a failure of effort or intelligence. It is a failure of timing.

Spaced repetition is the technique that solves this problem. It is the single most researched, most validated method for moving information from short-term to long-term memory. And if you are serious about learning a language, it is non-negotiable.

The Forgetting Curve: Why You Forget

In 1885, German psychologist Hermann Ebbinghaus ran a series of experiments on himself, memorizing lists of nonsense syllables and tracking how quickly he forgot them. His discovery, now called the Ebbinghaus Forgetting Curve, showed that memory decay is predictable and exponential.

Within 20 minutes of learning something new, you forget about 40% of it. After one hour, 50%. After one day, 70%. After a week, you are left with roughly 20% -- if you did nothing to reinforce the memory.

This curve is not a flaw. It is a feature. Your brain aggressively prunes information it deems unimportant. The way it decides what is important is simple: repetition. If you encounter a piece of information multiple times at specific intervals, your brain flags it as worth keeping.

How Spaced Repetition Works

The core principle is elegant: review information right before you would forget it. Not too early (wasted effort), not too late (you have to relearn it from scratch).

Here is the basic mechanism:

1. You learn a new word today. Your brain will forget it within 1--2 days.
2. You review it tomorrow. Now your brain holds it for about 4 days.
3. You review it again on day 5. Now it sticks for about 2 weeks.
4. You review it on day 19. Now it holds for a month.
5. Each successful review roughly doubles the interval.

After 5--7 reviews spread over a few months, the word is in your long-term memory. You can recall it months or years later with minimal effort.

This is not theory — it is active recall in action. It has been validated by over a century of cognitive science research, including large-scale studies by Pimsleur (1967), Leitner (1972), and more recently by Wozniak, the creator of the SuperMemo algorithm.

Why Cramming Fails

Cramming -- studying intensively right before you need the information -- feels productive. You can recall everything right after the session. But cramming produces almost zero long-term retention.

The reason is biological. Long-term memory formation requires a process called consolidation, which happens during sleep and over multiple days. When you cram, you create strong short-term traces but skip the consolidation process. The memories are vivid today and gone next week.

For language learning, cramming is particularly destructive. Languages are cumulative -- every new word and structure builds on what came before. If your foundation keeps dissolving because you crammed instead of spacing, you are perpetually rebuilding from scratch.

SRS Algorithms: From Leitner to FSRS

Not all spaced repetition is equal. The intelligence of the system depends on the algorithm that calculates your review intervals.

The Leitner System (1972): The simplest SRS. You sort physical flashcards into boxes. Correct answers move to a higher box (longer interval). Wrong answers go back to box one. It works, but it is crude -- it doesn't adapt to individual word difficulty.

SuperMemo / SM-2 (1987): Piotr Wozniak created the first computer-based SRS algorithm. SM-2 assigns an "ease factor" to each card and adjusts intervals based on your rating of each review (easy, good, hard, again). Anki, the most popular flashcard app, uses a modified version of SM-2.

FSRS (Free Spaced Repetition Scheduler): The current state of the art. FSRS uses a machine learning model to predict your probability of forgetting each individual card. It considers not just your rating but your entire review history, the card's difficulty pattern, and your personal learning speed. Studies show FSRS reduces the number of reviews needed by 20--30% compared to SM-2 while maintaining the same retention rate.

The difference matters at scale. If you are learning 30 words per day across two languages, a 25% reduction in reviews saves you 15--20 minutes daily. Over a year, that is over 100 hours.

💡 Try it now: Hyperpolyglot's Flashcards feature uses FSRS spaced repetition to schedule your reviews automatically — no manual interval tracking, just daily practice that adapts to your memory. Available on iOS, Android, and Web.

How to Implement Spaced Repetition Daily

Knowing the science is not enough. Here is how to actually use spaced repetition for language learning:

Step 1: Choose your tool. You need an SRS-powered flashcard system. Anki is free and powerful but has a steep learning curve. Hyperpolyglot uses the FSRS algorithm and is built specifically for language learners managing multiple languages -- cards are created from your own phrases and sentences, with audio and translations generated automatically.

Step 2: Learn words in context. Never put a single word on a flashcard — read our guide on why flashcards work best with full sentences. Use full sentences. "I would like a coffee" teaches you vocabulary, word order, and verb conjugation simultaneously. Isolated words are harder to remember and less useful in conversation.

Step 3: Set a daily new-word limit. Start with 15--20 new words per day. As your review pile grows, you may need to reduce new cards temporarily. The priority is always reviews first, new cards second. Skipping reviews defeats the entire system.

Step 4: Never skip a day. This is the hardest part and the most important. SRS works because it catches words at the exact moment they are about to fade. If you skip a day, you miss that window for dozens of cards, and they all need relearning. Even five minutes of reviews on a bad day is better than zero.

Step 5: Trust the algorithm. If a word keeps coming back, it is because you keep forgetting it. That is the system working correctly. Don't get frustrated -- get curious. Find a better mnemonic, create a vivid mental image, or connect the word to something you already know.

Optimal Intervals: What the Research Says

The optimal spacing intervals depend on your target retention rate. For language learning, most researchers recommend aiming for 85--90% retention. Here is what that looks like in practice:

• Review 1: 1 day after learning
• Review 2: 3--4 days after Review 1
• Review 3: 7--10 days after Review 2
• Review 4: 21--30 days after Review 3
• Review 5: 60--90 days after Review 4

After five successful reviews, most words are stable in long-term memory and will only need occasional refreshing every few months.

Modern algorithms like FSRS adjust these intervals dynamically for each individual card. A word that is phonetically similar to your native language might graduate to long intervals quickly. An abstract word in a distant language might need extra repetitions. The algorithm handles this complexity so you don't have to.

Keep Reading

• Why Flashcards Are Still the #1 Tool for Learning Vocabulary
• Active Recall: Why Testing Yourself is the Fastest Way to Learn
• The Perfect Daily Language Learning Routine

The Bottom Line

Spaced repetition is not a shortcut. It is the most efficient path. Every minute you spend reviewing at the right interval is worth ten minutes of random study.

If you are learning a language without SRS, you are working harder than you need to and retaining less than you should. Set up your system, commit to daily reviews, and let the algorithm do what a century of cognitive science designed it to do: make forgetting nearly impossible.