Forgetting Curve Calculator: When to Review

This forgetting curve calculator is a text tool, not a widget. It maps how many days have passed since your last review to an estimated retention percentage, then tells you when to review next so that recall does not slip below your target. Use it to answer the practical question behind most study planning: when should I review this so I do not forget it? The estimates come from the Ebbinghaus forgetting curve and modern spaced repetition research, and they are educational approximations, not guarantees about your own memory.

If you want the background story of why memories fade, read our guide on the forgetting curve first. This page is the actionable part. Feed it a number of days and a target retention, and it hands you a retention estimate plus a recommended review window. For the wider explainer on Ebbinghaus and the science, the guide at /guides/forgetting-curve goes deeper.

How This Forgetting Curve Calculator Works

The calculator rests on one idea: the longer you go without recalling something, the less of it you keep, and the drop is steep at first and gentle later. To estimate when to review, you need three things. First, the days since your last active review, meaning the last time you actually tested yourself rather than reread the page. Second, how many real reviews you have already done, because each one makes the memory more durable. Third, your target retention, the recall level you are willing to let a memory fall to before topping it up.

From those inputs the calculator does two jobs. It estimates roughly how much you still remember today, and it works out the day when your recall is about to drop below your target so you can review just before that point. The model assumes you did no sneaky reviews in between. If you have been quizzing yourself, your real retention is higher than the table shows.

The Forgetting Curve in 60 Seconds

Hermann Ebbinghaus described the forgetting curve in 1885 after testing himself on lists of nonsense syllables. He found that without review, memory drops sharply soon after learning and then forgets more slowly over time. Roughly half of fresh, meaningless material can be gone within a day or two, and only a fraction survives weeks later.

Two caveats matter. First, Ebbinghaus used nonsense syllables, which are forgotten faster than meaningful material you understand. If you truly grasp a topic, your personal curve is much flatter than these generic numbers. Second, these are historical averages, not laws. The point of a calculator is to give you an order of magnitude and a review window, not a precise reading of your brain.

The Forgetting Curve Formula in Plain Language

A widely used simple model is R = e^(-t/S). Here R is the probability you can recall the material, t is the number of days since your last review, and S is the stability of the memory, sometimes called the time constant. A bigger S means a flatter curve and slower forgetting. Stability grows every time you successfully review, and it is larger for easy or meaningful material than for hard or arbitrary content.

To decide when to review for a target retention R*, you rearrange the same formula: t = -S times the natural log of R*. For a target of 90 percent, the natural log of 0.9 is about -0.105, so the review day is roughly 0.105 times S. In plain words: we look for the moment your recall is about to slip below your target, for example 90 percent, and recommend a review just before that point. The calculator does not try to measure your personal S clinically. It uses reasonable defaults based on difficulty and how many reviews you have done.

This is the same logic behind modern algorithms like SM-2 and FSRS, which also model a recall probability that decays over time and trigger a review when it crosses a threshold. FSRS tracks three values, difficulty, stability, and retrievability, and uses a power-law curve to predict due dates more precisely than SM-2. Think of this calculator as a teaching version of that idea for a single topic, and an app like Betta as the full per-card version that targets around 90 percent recall automatically.

When to Review With Spaced Repetition

The single-exposure table above is the worst case: one look, then nothing. The whole point of spaced repetition is to break that curve. Every time you successfully recall the material, retention jumps back toward 100 percent and stability rises, so the next safe gap is longer. After a few reviews the same topic that used to vanish in days can hold for weeks.

The table below interprets a days-since-last-review input when you have reviewed more than once. The retention bands are conceptual ranges, not exact readings, and they assume active recall rather than passive rereading. The action column is what to do if your target is around 90 percent.

Worked Example 1: A Chapter You Saw Yesterday

You studied a chapter once yesterday, so it is 1 day since your single review, difficulty is medium, and your target is 90 percent. With only one exposure, stability is still low, so the calculator estimates retention around 70 to 85 percent today. You are right at the edge of your target.

The recommendation is to review today or tomorrow to solidify the memory before it slides further, then follow the sequence: next review around day 6 to 7, then around day 15, then day 30. That first early review is the one that flattens your curve the most, so do not skip it.

Worked Example 2: A Hard Topic Left for 10 Days

You reviewed a topic three times, it is hard material, and you last touched it 10 days ago with a 90 percent target. Three reviews have built decent stability, but 10 days is a long gap for hard content, so the estimate lands around 60 to 80 percent. You are entering the risky zone.

If this is exam-relevant, review in the next 1 to 2 days. After that successful review, stability climbs again and you can stretch the next gap toward two or three weeks. If you keep letting hard topics drift past their window, they cost you a full relearn each time.

Worked Example 3: Vocabulary Reviewed Five Times, 30 Days Ago

You have language vocabulary you reviewed five times, the material is easy and familiar, you last reviewed 30 days ago, and your target is 85 percent. Five reviews on easy material build high stability, so even after a month the estimate stays relatively high, perhaps 60 to 80 percent.

The recommendation is a maintenance review within the week to keep mastery, then you can space the next one out to two or three months. This is the payoff of consistent spacing: the same 30-day gap that would wipe a hard, once-seen topic barely dents a well-practiced easy one.

Worked Example 4: Fitting Reviews Around an Exam Date

You last reviewed 3 days ago, you have done 2 reviews, and your exam is in 12 days. The calculator suggests one review today, around day 3 to 4, then another around day 9 to 10, just before the exam. That places a strong, recent review close to test day while keeping the earlier spacing that built your stability.

For a near deadline, lean toward a higher target retention and one extra review rather than fewer. The cost of an early review is small; the cost of forgetting on exam day is not. To turn this into a day-by-day plan across a whole deck, the spaced repetition calculator at /tools/spaced-repetition-calculator and the daily review calculator at /tools/daily-review-calculator both work from the same spacing idea.

How to Use This Tool Step by Step

Start by picking one specific topic, chapter, or small deck, not your whole syllabus. Estimates are far more meaningful for a focused unit than for everything at once. Then estimate your inputs: days since your last real recall session, how many reviews you have done, and how hard the material feels.

Read the retention estimate as a range with a short interpretation, for example we estimate you remember roughly 60 to 80 percent, so expect to recall the main ideas but forget some details if quizzed now. Then act on the recommended window: write a specific date in your planner or calendar inside that window rather than a vague someday. Finally, redo the estimate after each big review to watch the windows stretch as your stability grows.

A forgetting curve calculator is only useful if you actually do a solid, retrieval-based review when it tells you to. Rereading is not enough. If you would rather not compute curves for hundreds of cards, that is exactly what Betta automates: it schedules each card just before you are likely to forget it, targeting around 90 percent recall, and you can photograph notes or slides to build the deck in seconds.

How Spaced Repetition Breaks the Curve

The spacing effect is one of the most robust findings in learning research: reviews spread over time produce stronger long-term retention than the same study crammed into one sitting. Each successful, spaced recall slows your future forgetting, which is why the intervals in the schedule table can keep growing.

The reviews should be active. Testing yourself with flashcards or quizzes strengthens memory and lengthens stability far more than rereading, which mostly creates a false sense of familiarity. Algorithms like FSRS formalize this by aiming for a target recall probability, often around 90 percent, and rescheduling each card the moment its predicted recall is about to dip below that line. Betta does this work automatically, card by card, where this calculator gives you the big-picture intuition for a topic. For how spacing fits a full exam timeline, see /guides/spaced-repetition-exam-prep, and to compare box-based scheduling, the Leitner box calculator at /tools/leitner-calculator covers the classic 1/2/4/7/15 day version.