You've probably heard someone say "just use Anki" as if it's magic. Medical students swear by it. Language learners credit it with fluency. Law students, engineers, and certification preppers all report the same thing: Anki works better than any other study method they've tried.
But why? What's actually happening when you review flashcards on a schedule? And why does this particular approach to studying produce dramatically better results than rereading notes, highlighting textbooks, or cramming before exams?
The answer lies in over 130 years of memory research, and it turns out the science is remarkably clear.
The Forgetting Curve: Where It All Started
In 1885, German psychologist Hermann Ebbinghaus published a study that would eventually reshape how we think about learning. He memorized lists of nonsense syllables — meaningless combinations like "DAX," "BUP," and "ZOL" — and tested himself at various intervals to see how much he retained.
What he found was devastating for anyone who relies on cramming: memory decays exponentially. Within 20 minutes, he'd forgotten 40% of what he'd learned. After a day, 67% was gone. After a month, nearly 80% had vanished.
This exponential decay is called the forgetting curve, and it applies to nearly everything we learn. It doesn't matter how well you understood the lecture or how carefully you read the chapter. Without reinforcement, most of that knowledge will fade within days.
Here's the critical insight, though: Ebbinghaus also discovered that each time he reviewed the material, the forgetting curve flattened. The first review might keep the memory alive for 2 days. The second review extended it to a week. The third pushed it out to a month. Each repetition made the memory more durable, and the intervals between necessary reviews grew longer.
This is the core principle behind spaced repetition: review at the moment you're about to forget, and you'll remember for longer each time.
The Testing Effect: Why Flashcards Beat Rereading
Before diving deeper into spacing, we need to address something fundamental: why flashcards at all? Why not just reread your notes on a schedule?
The answer comes from what cognitive scientists call the testing effect (also known as retrieval practice). Decades of research, including landmark studies by Roediger and Karpicke in 2006, have demonstrated that the act of retrieving information from memory strengthens that memory far more than passively reviewing it.
When you read a fact in your notes, your brain recognizes it. "Oh yes, I've seen this before." That recognition feels like learning, but it's an illusion — psychologists call it the fluency illusion. You feel confident because the material looks familiar, but familiarity isn't the same as recall.
When Anki shows you a card and you have to produce the answer before flipping it, you're forcing your brain to reconstruct the memory from scratch. This effortful retrieval creates stronger neural pathways than passive review. It's the difference between recognizing someone's face in a crowd and being able to draw their face from memory.
A 2011 study in Science by Karpicke and Blunt found that students who practiced retrieval retained 50% more material than students who used elaborate concept mapping — a method that already outperforms rereading. Testing yourself isn't just better than rereading; it's better than almost every other study strategy researchers have tested.
The Spacing Effect: Why Timing Matters
The spacing effect is one of the most robust findings in all of cognitive psychology. Distributing your study sessions over time produces dramatically better long-term retention than massing them together, even when the total study time is identical.
Imagine you have 60 minutes to study pharmacology drug mechanisms. You could spend all 60 minutes today (massed practice), or you could spend 20 minutes today, 20 minutes in three days, and 20 minutes next week (spaced practice). The spaced approach will produce superior retention on a test given a month later — often by 30-50%.
Why does spacing work? Several theories contribute:
Contextual variability: When you study the same material on different days, your brain encodes it with different contextual cues — different moods, environments, and mental states. This creates multiple retrieval pathways, making the memory more accessible.
Desirable difficulty: Spacing makes retrieval harder, which sounds bad but is actually beneficial. When you struggle to recall something, the eventual success of remembering it creates a stronger memory trace. Easy recall doesn't strengthen memory much.
Consolidation: Your brain consolidates memories during sleep and rest. Spacing gives your brain time between sessions to move information from fragile short-term storage to durable long-term memory.
How Anki's Algorithm Implements the Science
Anki takes these principles — the testing effect, the spacing effect, and the forgetting curve — and automates them. Here's how:
The SM-2 Algorithm (Classic Anki)
Anki's original algorithm is based on SM-2, developed by Piotr Wozniak in 1987 for his SuperMemo software. It works like this:
- You see a card and try to recall the answer
- You rate your recall: Again (forgot), Hard, Good, or Easy
- Based on your rating, Anki calculates the next review date using an ease factor — a multiplier that determines how quickly intervals grow
If you press "Good" consistently, your intervals might look like: 1 day → 3 days → 8 days → 21 days → 55 days → 143 days. Each successful review multiplies the previous interval by approximately 2.5.
If you press "Again," the card resets to a short interval and the ease factor decreases, meaning future intervals grow more slowly. This is how Anki adapts to difficulty — cards you struggle with appear more frequently.
SM-2 is elegantly simple, and it works remarkably well for a 1987 algorithm. But it has known issues. The ease factor can spiral downward (a problem called "ease hell"), it doesn't account for individual learning differences, and it treats all "Good" ratings the same regardless of how close you were to forgetting.
FSRS: The Modern Evolution
In 2023, Anki adopted FSRS (Free Spaced Repetition Scheduler) as an optional replacement for SM-2. Developed by Jarrett Ye based on the DSR (Difficulty, Stability, Retrievability) model, FSRS represents a significant leap forward.
FSRS uses machine learning trained on millions of real review records to model three key variables:
- Difficulty: How inherently hard is this card for you?
- Stability: How long can you go before your recall probability drops below your target?
- Retrievability: Right now, what's the probability you could recall this card?
Instead of a simple ease factor, FSRS builds a mathematical model of your memory for each card. It knows that a card with stability of 30 days has about a 90% chance of being recalled at day 30, and uses this to schedule reviews at the optimal moment.
The result? Studies show FSRS reduces total review time by 20-40% compared to SM-2 while maintaining the same retention rate. It achieves this by scheduling reviews more precisely — neither too early (wasting time on cards you'd remember anyway) nor too late (forcing you to relearn forgotten material).
The Interleaving Effect: Why Mixed Review Works
When you review Anki cards, you don't see all your anatomy cards, then all your pharmacology cards, then all your pathology cards. Instead, they're interleaved — you might see a pharmacology card, then anatomy, then biochemistry, then back to pharmacology.
This feels harder, and students often prefer blocked practice (studying one topic at a time). But research consistently shows that interleaving produces better long-term learning. A 2014 meta-analysis by Dunlosky and colleagues identified interleaved practice as one of the most effective learning strategies available.
Why? Interleaving forces your brain to identify which strategy or knowledge set applies to each problem. In blocked practice, you always know the context — "I'm studying pharmacology, so this must be a drug mechanism question." In interleaved practice, you first have to recognize what kind of question you're facing, which mirrors how exams and real-world application actually work.
Anki's default behavior of mixing cards from different decks and topics automatically gives you this benefit. If you've been separating your cards into isolated study sessions by subject, you're leaving performance gains on the table.
Active Recall vs. Passive Recognition: The Neural Difference
Neuroscience research using fMRI and EEG has begun revealing what happens in your brain during active recall versus passive review.
During passive review (rereading notes), activity is primarily in the visual cortex and areas associated with familiarity detection. Your brain says "I've seen this before" and moves on.
During active recall (Anki-style retrieval), you see widespread activation across the hippocampus, prefrontal cortex, and the specific cortical regions where the memory is stored. Your brain is literally reconstructing the memory, strengthening the connections between neurons that encode it.
Think of it like a path through a forest. Passive review is like looking at the path from a helicopter — you can see it exists, but you haven't walked it. Active recall is like walking the path yourself, wearing it deeper with each trip. Eventually, the path becomes so well-worn that finding your way is effortless.
The Minimum Information Principle
One of the most important principles for effective Anki use comes from Wozniak's "20 Rules of Formulating Knowledge": the minimum information principle. Each card should test one atomic piece of information.
Bad card: "List the 12 cranial nerves and their functions."
Good card: "Which cranial nerve innervates the lateral rectus muscle?" → "CN VI (Abducens)"
Why does this matter scientifically? When a card tests multiple things, a failure could mean you forgot any one of them. Anki can't tell which piece you're struggling with, so it can't schedule optimally. Atomic cards let the algorithm do its job — hard cards get reviewed more, easy ones less.
This principle also leverages the generation effect: producing an answer (even a short one) creates a stronger memory than recognizing or selecting one. The more specific the prompt, the more precise the retrieval, and the stronger the resulting memory.
How Much Should You Review? The Research on Daily Load
A common question is how many new cards to learn per day and how many reviews to do. Research and community experience suggest some guidelines:
New cards: 10-30 per day is sustainable for most people. More than 40-50 new cards daily leads to review pile-ups that become unmanageable within weeks. Starting with 10-15 and adjusting upward is the safest approach.
Reviews: The number of daily reviews is determined by your past learning. If you've been adding 20 cards per day for six months, you might see 100-200 reviews daily. This is normal and takes 30-60 minutes for most users.
Retention target: With FSRS, you can set a desired retention rate (default: 90%). Higher retention means more frequent reviews. Research suggests 85-90% is the sweet spot — high enough to maintain knowledge, low enough to avoid excessive review burden.
The key insight: Consistency matters more than volume. Reviewing 50 cards every day for a year produces enormously better results than doing 200 cards sporadically. Spaced repetition only works if you maintain the schedule.
Common Mistakes That Undermine the Science
Understanding the science helps you avoid patterns that sabotage your learning:
Pressing "Easy" too often: If you inflate your intervals by always pressing Easy, you'll push cards past the optimal review point and forget them. Rate honestly — "Good" should be your default for cards you recalled correctly with moderate effort.
Reviewing in order: Some students add cards in lecture order and review them the same way. This creates artificial context cues — you remember the answer partly because you know what came before it. Anki's random ordering eliminates this, so don't override it.
Too many cards, too little understanding: Anki strengthens memories, but you need to understand the material first. Using Anki on material you don't comprehend is like trying to memorize a language you don't speak — the cards won't stick because there's no conceptual framework to anchor them to.
Skipping days: Missing a single day doubles your review load the next day, and the cards you miss are now past their optimal review point. The science works because of precise timing. Consistency is non-negotiable.
Not using images: Dual coding theory (Paivio, 1986) shows that information encoded both verbally and visually is retained better than either alone. If you're studying anatomy, histology, pathology, or anything with visual components, include images on your cards.
From Slides to Cards: Applying the Science Efficiently
The biggest barrier to using Anki isn't the science — it's the time it takes to create good cards. Manually converting lecture slides or textbook chapters into properly formatted, atomic flashcards can take hours.
This is where tools like SlideToAnki come in. Instead of spending your evening typing out cards, you can upload your lecture slides and get AI-generated flashcards that follow the minimum information principle — one concept per card, proper cloze deletions, and image occlusion for visual material.
The time you save on card creation is time you can spend on what actually matters for learning: reviewing your cards consistently and letting the spacing algorithm do its work.
The Bottom Line
Spaced repetition isn't a study hack or a productivity trick. It's the direct application of some of the most well-established findings in cognitive science:
- The forgetting curve tells us memory decays exponentially without reinforcement
- The testing effect tells us active recall strengthens memory far more than passive review
- The spacing effect tells us distributed practice beats massed practice
- Interleaving tells us mixed review beats blocked review
- The minimum information principle tells us atomic cards produce the best results
Anki automates all of this. You just need to show up every day, answer your cards honestly, and trust the algorithm. The science is on your side — 130 years of it.