Why a term comes back today
Reviews are scheduled by FSRS‑6 (Free Spaced Repetition Scheduler), an open algorithm developed by the open‑spaced‑repetition community. It is the same method the app describes on its own “The science behind it” screen.
Half a century of hunting for the right moment
The idea is old: to keep something for good you have to review it — but not too early (wasted effort) and not too late (you start from scratch). Successive systems differ only in how well they hit that moment.
The Leitner system. A card moves between boxes on a fixed schedule: every card in a given box comes back after the same delay, no matter how well you know it.
SuperMemo's SM‑2. Used for years by Anki as well, it went further and gave each card a single ease factor to multiply successive intervals by — but it still does not separate how hard the material is from how durably you have already learned it.
FSRS‑6. A memory model instead of a multiplier. For each card separately it answers the question “when will I forget this?” — and only then brings it back.
Three quantities instead of one
FSRS models memory with three quantities at once: the difficulty of the card, the stability of the memory trace, and the chance of recalling it right now. That lets it fit a forgetting curve to each card separately and schedule the review for the moment your probability of recall drops to a target you choose yourself (90% by default, adjustable under advanced settings). Instead of multiplying an interval by a constant, it answers the question “when will I forget this?”.
Where the parameters come from
The model's twenty-one parameters were fitted against a very large body of real review histories shared by the community. The goal is simple: fewer reviews for the same recall — or better recall for the same effort.
How you see it in the app
Every study mode runs on it:
- Review — classic self-assessment (Again / Hard / Good / Easy). Your rating moves the next appearance of the card directly.
- Typing mode — you type the answer, and the result maps onto an FSRS rating. The strictest check there is, because “almost remembering” does not count.
- Topic review — deliberately leaves the schedule alone. It is there for a quick sweep before an exam and will not damage your plan.
This is why it pays to rate honestly: inflated ratings thin out the reviews, and the term slips your mind exactly when you need it.
You set the recall target yourself. The 90% default is a sensible compromise between how many reviews you do and how sure you can be that the term will be there. Before an exam session you can raise it — at the cost of more cards per day.
What else the memory research says
- Active recall beats passive reading. That is why a card asks first and only then shows the answer — even when it is tempting to peek.
- Spacing beats cramming. Ten minutes a day for a week leaves a more durable trace than seventy minutes the night before the exam.
- Difficulty helps. Effortful recall strengthens the trace — hence the typing mode and asking for a term from several sides: English, Latin, Polish.
Who is behind MedFlashcards
MedFlashcards: Medical English (medyczny.app) comes from the makers of InstaLing, the most widely used language-learning platform in Polish schools, and podreczniki.app. The teaching method is backed by, among others:
Author of more than a dozen language textbooks and editor of the largest English–Polish dictionary ever published (2014, Wiedza Powszechna).
MA in psychology, linguistics and intercultural education.
World champion in team programming.
Over a decade of work on daily vocabulary practice in Polish schools — and on getting students to come back to it.
Test the method on 60 terms
Free, with no account and no time limit.
Google Play and the Google Play logo are trademarks of Google LLC. Apple and App Store are trademarks of Apple Inc.