Flashcards and spaced repetition, done right
Cramming decays within days. Spaced repetition is the most evidence-backed way to make knowledge stick — if the cards are effortless to make.
The short answer
Spaced repetition works by scheduling each review just before you would have forgotten the material, so every recall lands where it strengthens memory most. It is one of only two techniques rated high utility in the Dunlosky review of ten common study methods — the other being practice testing, which flashcards also deliver. A meta-analysis by Cepeda and colleagues pooled 254 studies and found spaced practice beat massed practice on later recall. AuroraNote generates decks from your own uploaded material at four difficulty levels and schedules them with FSRS-6, the same algorithm Anki moved to: it models how long a memory lasts and how hard a given card is for you as two separate quantities, then targets a 90% chance you still remember a card when it comes up. Reviews sync across devices, so a card graded on a laptop is due at the same moment on your phone.
Cramming feels productive but decays within days. Spaced repetition — reviewing a fact just before you would forget it — is the single most evidence-backed technique in learning science. The catch has always been effort: making good cards by hand is so tedious that most people never keep it up.
When the cards generate themselves from your material, the technique finally becomes practical. Here is how to use it well.
Why does spaced repetition work?
Every time you successfully recall something, the memory strengthens and the forgetting curve flattens. Reviewing at expanding intervals — a day, then a few days, then a week — puts each review at the moment it does the most good, so you spend the least time for the most retention.
This is not a folk theory. Cepeda and colleagues pooled 254 separate studies of distributed practice and found the spacing advantage held across essentially every condition they could compare. And in the Dunlosky review of ten common study techniques, distributed practice and practice testing were the only two rated high utility; highlighting, rereading and summarising were all rated low.
The active part matters as much as the spacing. Retrieving an answer from memory beats rereading it, because the effort of recall is what builds the memory. Roediger and Karpicke showed the effect and the trap in the same experiment: students who were tested recalled more a week later than students who reread, while the rereaders were the ones who felt more prepared. Difficulty during study feels like failure and predicts success. Flashcards are simply the most efficient way to force that retrieval.
What algorithm decides when a card comes back?
AuroraNote schedules reviews with FSRS-6 — the Free Spaced Repetition Scheduler, the algorithm Anki itself moved to. It matters more than it sounds, because most flashcard apps still use SM-2, an algorithm published in 1987 that tracks a single "ease factor" per card and nudges it up or down after each review.
FSRS instead models two things separately: the stability of a memory, meaning how many days it will last, and the difficulty of that particular card for you. From those it schedules the next review to hit a target probability that you will still remember it — 90% by default. Due dates are also deliberately fuzzed, so a deck you built in one sitting does not all come due again on the same day.
| SM-2 (1987) | FSRS-6 | |
|---|---|---|
| What it tracks | One ease factor per card | Memory stability and card difficulty, separately |
| How intervals are chosen | Fixed multipliers, nudged after each review | Fitted to how memory actually decays |
| Target | None — intervals simply grow | A chosen recall probability, 90% by default |
| Cards built in one sitting | All come due again together | Due dates fuzzed so the load spreads |
How should you use the difficulty levels?
Aurora generates complete decks at four difficulty levels. Start at "Medium" for a first pass to lock in the basics; move to "Hard" or "Expert" once they feel automatic. Those higher levels ask application and reasoning questions — the kind professors actually put on exams — instead of simple recall.
- Easy / Medium: definitions, facts, core vocabulary.
- Hard: apply a concept to a new example.
- Expert: compare, reason, or solve — exam-style.
How do you turn cards into an honest score?
Use Quiz mode for self-testing and finish with a Study Session that ends in a percentage. "Am I ready?" then gets a number instead of a feeling — which matters precisely because, as the retrieval research shows, the feeling is the unreliable part. Below 70%, reread only the sections behind the cards you missed, not the whole chapter.
Frequently asked questions
How many cards should I review per day?
Short daily sessions of 10–15 cards are more effective than long infrequent ones. Consistency across days is what drives retention.
Where do the flashcards come from?
They are generated automatically from your uploaded material — lectures, notes or PDFs — so the deck matches exactly what you need to know.
Is active recall better than rereading?
Yes, and it is one of the most replicated findings in learning research. Roediger and Karpicke found that students who took a memory test on a passage recalled substantially more of it a week later than students who spent the same time rereading — even though the rereaders predicted they would do better. The Dunlosky review reached the same conclusion across ten techniques: practice testing rated high utility, rereading rated low.
Does AuroraNote use the same algorithm as Anki?
Yes. Reviews are scheduled with FSRS-6 through the ts-fsrs library — the same Free Spaced Repetition Scheduler that Anki adopted in place of SM-2. Review state is stored per card and synced to your account, so grading a card on one device updates its due date everywhere.
Sources
- Cepeda, Pashler, Vul, Wixted & Rohrer — Distributed Practice in Verbal Recall Tasks: A Review and Quantitative Synthesis — Psychological Bulletin 132(3), 2006.
- Roediger & Karpicke — Test-Enhanced Learning: Taking Memory Tests Improves Long-Term Retention — Psychological Science 17(3), 2006.
- Dunlosky, Rawson, Marsh, Nathan & Willingham — Improving Students’ Learning With Effective Learning Techniques — Psychological Science in the Public Interest 14(1), 2013.
- Open Spaced Repetition — FSRS, the Free Spaced Repetition Scheduler — GitHub, 2026.