Anki and Spaced Repetition: How the System Actually Works
Anki is the most widely used implementation of spaced repetition, but how does the algorithm actually schedule your reviews? This guide covers SM-2, FSRS, card design, common mistakes, and how to build a sustainable daily study habit.
What Are Anki and Spaced Repetition, and Why Do Students Pair Them?
Spaced repetition is a learning strategy built on one of the most replicated findings in cognitive science: memory fades predictably over time, and reviewing information just before you would forget it produces far more durable retention than reviewing it at fixed intervals or in a single cramming session. The principle traces back to Hermann Ebbinghaus's forgetting curve, which he described in the 1880s. Without active review, most new information loses the majority of its detail within 24 to 48 hours. Each successful retrieval resets the forgetting clock and extends how long the memory remains accessible.
Anki is the software that turned this principle into a practical daily study tool. Released in 2006, Anki is an open-source flashcard application that uses an algorithm to schedule when each card in your deck should appear for review. You rate how well you recalled each card, and the algorithm adjusts the next review date accordingly. Cards you know well are pushed weeks or months into the future. Cards you struggle with return tomorrow.
The reason anki and spaced repetition are mentioned together so frequently is that Anki is the canonical implementation of the concept. When educational researchers study spaced repetition, they often use Anki as the study tool. When medical students describe their board exam preparation, they nearly always mean Anki decks reviewed on a spaced schedule. The software has become so central to the practice that the two are nearly synonymous in many study communities.
For a broader comparison of spaced repetition software beyond Anki — including newer tools and AI-powered alternatives — see our guide on spaced repetition apps.
Anki and spaced repetition have such a long shared history that in many study communities, learning someone uses one means the other is assumed.
How Does Anki Turn Spaced Repetition Theory into a Daily Review Schedule?
The scheduling algorithm at the heart of Anki is called SM-2, developed by Piotr Wozniak in the late 1980s. It operates on two variables for each card: the interval (days until the next review) and the ease factor (a multiplier controlling how quickly the interval grows). When you review a card, you rate your recall on a four-point scale.
**Again**: You could not recall the card. The interval resets to minutes, and the ease factor decreases, slowing future interval growth. **Hard**: You recalled the card with significant effort. The interval increases slightly; the ease factor decreases slightly. **Good**: You recalled correctly with manageable effort. The interval multiplies by the current ease factor — typically around 2.5. A card due in four days, recalled as Good, next appears in roughly ten days. **Easy**: You recalled instantly. The interval multiplies by a larger factor and the ease factor increases.
Over many review cycles, well-learned cards accumulate intervals measured in months. Cards you consistently struggle with stay on short intervals. Your daily review queue contains exactly the cards most at risk of being forgotten — no more, no less.
Anki also offers FSRS (Free Spaced Repetition Scheduler) as a built-in alternative to SM-2. FSRS builds a personal memory model for each card based on your full review history rather than just the most recent rating. Independent testing by medical students and language learners suggests FSRS reduces daily review time for the same retention level, particularly for large decks maintained over many months. Enabling it requires only a toggle in deck options and is worth doing for any consistent Anki user.
For the specific intervals to use when building a review calendar manually rather than relying on an algorithm, see our guide on spaced repetition schedules.
SM-2 was developed in 1987 and remains the foundation of Anki's scheduling. FSRS, available as a built-in option since 2023, improves long-term retention estimates measurably for large, sustained decks.
How Do You Set Up Anki and Spaced Repetition from the Start?
Getting started with Anki is quick, but a few early decisions determine whether the system stays manageable long-term. The most important is your daily new card limit, which controls how many new items enter your queue each day and therefore how large your review sessions become as the deck grows.
The daily new card limit is the most consequential setting in Anki. Set it too high and the review queue grows faster than you can clear it; keep it sustainable and the system runs itself.
- 1
Download Anki and create an AnkiWeb account
Anki is free on Mac, Windows, Linux, and Android. The iOS app is a one-time $24.99 purchase. Create a free AnkiWeb account immediately after installing to enable cross-device sync. Mobile review sessions during a commute, between classes, or over lunch are where the daily review habit actually forms for most students.
- 2
Create one deck per subject rather than one large deck
Build a separate deck for each subject or exam. A single deck of 500 cards covering an entire semester is harder to review sustainably than five decks of 100 cards each. Smaller decks allow you to complete a full review session in under 20 minutes, which is the target length for sessions you will actually sustain every day.
- 3
Set your daily new card limit before adding any cards
Before adding your first card, open deck options and set the new card limit. A sustainable starting rate is 10 to 20 new cards per day. At that pace, review sessions 30 days in stay under 25 minutes. Students who begin at 50 or more new cards per day build a review queue that outpaces available time within three weeks and typically abandon the system before the method has time to show results.
- 4
Enable FSRS in deck settings
In deck options, find the FSRS toggle and enable it. FSRS requires several weeks of review history to calibrate your personal memory curve. During the initial period the difference from SM-2 is minor, but after two to three months of consistent use, FSRS noticeably reduces daily review time by scheduling longer intervals for material you have reliably retained.
- 5
Add your first cards using the minimum information principle
Each card should test exactly one fact, definition, or relationship. Avoid questions that accept a range of answers, cards where the question contains the answer, and cards that require producing a long list. If the correct answer runs longer than two sentences, it is probably two cards. This one-card-one-fact rule is the single most important guideline for Anki card design.
What Makes a Good Anki Card for Spaced Repetition?
Card design is where most Anki and spaced repetition workflows quietly fail. Students who convert notes directly into question-answer pairs often produce a deck full of recognition tasks rather than recall tasks. The cards look like quizzes but do not require actually retrieving information from memory — and spaced repetition's effectiveness depends entirely on genuine retrieval.
The distinction matters because of how the scheduling works. You rate each card based on how well you recalled it, and the algorithm schedules the next review accordingly. A deck of poorly designed cards produces inflated recall ratings, cards get pushed to long intervals before the underlying knowledge is actually retained, and you end up with false confidence that collapses during an exam.
Effective Anki cards follow the minimum information principle: each card tests one fact, one definition, one relationship, or one application. This makes the card easier to answer precisely, easier to assess honestly, and easier to retrieve across different exam contexts.
For building decks more efficiently from lecture notes or PDFs, see our guide on turning notes into flashcards.
A card requiring a paragraph-length answer is almost always two or three cards in disguise. The most common Anki card quality problem is questions too broad to assess clearly.
- 1
Use cloze deletion for factual content
Cloze deletion hides a word or phrase within a sentence and asks you to recall the missing part. A card reading 'The mitochondria uses {{c1::oxidative phosphorylation}} to produce ATP' tests specific recall within context. Anki supports cloze cards natively and they work particularly well for anatomy, pharmacology, and any subject with precise technical terminology.
- 2
Avoid questions where the phrasing contains the answer
If reading the question would let someone with no prior knowledge guess correctly, rewrite it. A question like 'What enzyme does the mitochondria use to produce ATP via oxidative phosphorylation?' nearly answers itself. A cleaner version: 'What process does the mitochondria use to generate most of its ATP?' That version requires actual knowledge to answer.
- 3
Test application, not only definition
For conceptual subjects, include cards that require applying a concept to a new situation, not only defining it. 'Define osmosis' is a recognition task. 'If a cell is placed in a hypertonic solution, what happens to its water volume and why?' requires applying the concept. Cards that only test definitions leave gaps in practical understanding and in exam performance for application-style questions.
- 4
Keep the answer short enough to assess in under ten seconds
If the answer takes 30 seconds to read, the self-assessment step becomes ambiguous — you cannot tell whether you actually recalled the card correctly until you have read the whole answer. Split long-answer cards into two or three specific cards, each testing one piece of the full explanation.
What Are the Most Common Mistakes in an Anki and Spaced Repetition Workflow?
Most students who try anki and spaced repetition and abandon it are not failing because the method does not work. They are making one or two consistent errors that compound over weeks until the system feels unmanageable. The patterns below are predictable enough that knowing them in advance is more useful than troubleshooting after the fact.
Anki review sessions longer than 30 minutes are almost always caused by too many new cards added too quickly — not by a problem with spaced repetition itself.
- 1
Adding new cards faster than the review queue can absorb them
At 10 new cards per day, the review queue stays manageable. At 50 new cards per day, the daily backlog outpaces available time within three weeks. If your review sessions are running more than 30 minutes, halve your new card rate before the backlog becomes discouraging enough to quit. Session length is the clearest signal that the card rate needs adjusting.
- 2
Stopping entirely rather than clearing a backlog
Missing two or three consecutive days creates a review pile. The correct response is to temporarily suspend new card additions and work through the backlog at a lower pace over several days. Students who stop using Anki entirely rather than clearing the backlog typically restart with a fresh deck, lose all accumulated scheduling history, and repeat the cycle. Backlog is normal; abandonment is the failure mode.
- 3
Building cards from material you do not yet understand
Adding a card to Anki before you understand the underlying concept encodes the misunderstanding in long-term memory. Spaced repetition reinforces whatever you first encoded, including errors. Learn the concept first from a lecture, textbook, or discussion, then write the card. The deck should contain knowledge you already have, scheduled for retention — not material you are hoping to understand through repeated exposure to a question you cannot yet answer.
- 4
Rating recall too generously
SM-2 and FSRS rely entirely on your self-report to schedule reviews. Marking a card Good when you hesitated or partially guessed pushes it to a longer interval based on inaccurate data. The only consequence of marking Hard or Again honestly is a shorter review interval — which is exactly what a weak memory needs. Consistently generous ratings produce false mastery that becomes visible only during the exam.
How Notelyn Removes the Setup Friction in an Anki and Spaced Repetition Workflow
The biggest practical barrier in an anki and spaced repetition system is not the algorithm or the daily review habit. It is building the initial deck. A student processing a 40-page textbook chapter or a two-hour lecture recording needs to extract key concepts, phrase them as questions, and format them as cards before any review scheduling can begin. For students carrying five subjects, that deck-building overhead often exceeds the time available for actual studying.
Notelyn addresses this by generating study material directly from source content. Import a PDF, upload a lecture recording, paste typed notes, or drop in a video link, and Notelyn produces a structured AI summary, a flashcard deck, a quiz, and a Q&A mode — all from the same import. The deck-building step that prevents many students from reaching consistent daily review is removed from the workflow.
This matters most for students processing their own original material: university lectures, research papers, online course videos. For subjects with well-maintained shared Anki decks already available — first and second year medical curricula, language vocabulary systems, bar exam rules — downloading an existing deck may still be faster. But for anyone generating review material from their own sources, Notelyn's capture-to-review pipeline saves hours per week that go back into actual retrieval practice.
Notelyn's quiz mode presents questions without visible answers, which is the correct format for retrieval practice. The AI Q&A feature lets you ask questions about your imported content and receive answers grounded in that specific source, useful when a flashcard references something you need to understand more deeply before it goes into long-term review rotation.
For a direct comparison of how different AI tools handle the step from raw study material to reviewable flashcards, see our guide on how to make flashcards with NotebookLM.
The highest-friction step in any Anki and spaced repetition workflow is building the initial deck. Notelyn's automatic generation from PDFs, lectures, and audio removes that barrier so study time goes toward retrieval practice rather than card formatting.
- 1
Import your source material and use the AI summary as a retrieval cue
Upload a PDF, paste lecture notes, or import audio or video into Notelyn. Before reading the generated AI summary in full, close it and write down everything you remember about the topic from memory. Then open the summary and compare. This turns what would be passive review into a retrieval practice session and identifies your actual knowledge gaps before any flashcard review begins.
- 2
Generate flashcards, a quiz, and Q&A from the same import
From the same imported content, generate a flashcard deck and enable quiz and Q&A modes. Review the generated flashcards and replace recognition-style questions with recall-format questions. Use quiz mode for structured retrieval sessions with answers hidden. Use Q&A when a flashcard references a concept you need to understand more fully before placing it into long-term review rotation.
- 3
Use the edited deck as the foundation for your spaced review schedule
Once the deck is edited for quality, begin spaced review sessions. Cards you recall confidently in the first session get scheduled further out. Cards you struggled with return sooner. This follows the same core logic as Anki's algorithm, applied to content generated directly from your own study material without the manual card-creation overhead that causes most students to abandon the system before it delivers results.
Building a Long-Term Study System with Spaced Repetition
A spaced repetition system only produces its full benefits when maintained consistently over months. The compound effect is real: a deck started at the beginning of a semester and reviewed daily will have most of its material in long-term memory by mid-term, requiring only a fraction of the original weekly time to sustain. A deck started one week before finals produces familiarity that fades within days of the exam.
The single most reliable predictor of whether anki and spaced repetition will work for a given student is whether they review on days when the queue is small and the material feels easy. Those are the sessions that prevent backlog from forming and keep long-interval cards from slipping back to short ones. The habit is built during the low-stakes sessions, not only before an exam.
Start with one subject and one deck. Build 10 cards from your most recent lecture and review them the next morning. Add 10 more and repeat for two weeks. After two weeks, the habit either exists or it does not — and a two-week experiment is far easier to diagnose and adjust than a semester of accumulated cards and mixed motivation.
For further reading on the retrieval science that makes spaced repetition effective, see our guide on active recall studying. For how to structure review windows around a fixed exam date, the spaced repetition schedule guide covers manual interval planning from first principles.
Powiązane artykuły
Wypróbuj te funkcje
Odkryj przypadki użycia
Rób lepsze notatki z AI
Notelyn automatycznie przekształca wykłady, spotkania i pliki PDF w uporządkowane notatki, fiszki i quizy.