FSRS Retention Calculator
How much does remembering more actually cost? Explore the trade-off at the heart of the FSRS spaced repetition algorithm: your desired retention setting versus your daily review workload.
The retention–workload trade-off
FSRS schedules each review for the moment your predicted recall falls to your desired retention, so the interval scales by ln(retention)/ln(0.9) relative to the 90% default. This calculator combines that with the hidden cost of lapses — every failed card needs relearning — to estimate your true daily workload. The curve makes the decision concrete: pushing from 90% to 97% retention roughly triples your reviews, while dropping to 85% cuts them by about a third.
In RemNote, FSRS is built in and fits its parameters to your real review history — this page is for building intuition about what the retention dial actually does.
Join 1,000,000+ students getting higher grades!
More study tools
Frequently Asked Questions
What is desired retention in FSRS?
Desired retention is the probability of recall you want to have when a card comes up for review. FSRS schedules each review for the moment your predicted recall drops to that level. At 90%, you'll answer about 9 in 10 reviews correctly; setting it higher means shorter intervals and more reviews, lower means longer intervals and more lapses.
Why does workload explode above ~93% retention?
Because forgetting is exponential, keeping recall near-perfect requires disproportionately short intervals: the interval scales with ln(retention)/ln(0.9), which shrinks rapidly as retention approaches 100%. Going from 90% to 97% roughly triples your daily reviews while adding only 7 points of retention. That non-linear cost is why 85–92% is the usual sweet spot.
What retention setting should I use?
Start at the 90% default. Raise it toward 93–95% only for high-stakes material close to an exam, where extra reviews are worth it. Lower it toward 80–85% for large low-stakes collections you want to maintain cheaply. The right setting is the one whose daily workload you will actually sustain.
Does RemNote use FSRS?
Yes. RemNote uses the FSRS algorithm to schedule flashcard reviews, adapting to your actual review history per card. You get the retention-versus-workload trade-off described here automatically, without tuning anything — and you can adjust desired retention in your scheduler settings if you want to.
