Skip to content
K Knidox Search…
Study · Memory

Forgetting Curve Calculator

Estimate how much you still remember as time passes without review.

How slowly you forget — larger means the memory lasts longer.
How long since you last learned or reviewed it.
Try a scenario
Estimated retention (R)
36.8%

Drops to 50% after about 1.386 days (t = S·ln2). Each review resets the curve.

Retention over time
Memory retention decaying over time100%0%Day 0Day 8

Without review, retention falls fast at first, then levels off. A review lifts the line back toward 100% and flattens the next decline.

Memory decays exponentially without review. The forgetting curve estimates retention as R = e^(−t ÷ S), where t is time elapsed and S is memory strength. With S = 2 days, after t = 2 days you retain e^(−1) ≈ 36.8% — and you cross 50% at about 1.39 days.

What the forgetting curve shows

In the 1880s Hermann Ebbinghaus tested his own recall of nonsense syllables and found that memory fades fast at first, then slows down. That shape is the forgetting curve: retention drops steeply in the hours and days after learning, then flattens as the memories that survive become more stable.

This calculator uses the common exponential form R = e^(−t ÷ S). R is the fraction of material you still hold (shown as a percent), t is the time since you learned it, and S is the memory strength — a larger S means the curve falls more slowly. At t = S the exponent is −1, so R = 1/e ≈ 36.8%. This is a simplified model: real retention varies by person, by how meaningful the material is, and by how well you sleep.

R = e^(−t ÷ S)

R = fraction retained · t = time elapsed · S = memory strength (same time units)

Worked example

You set memory strength S = 2 days and check retention after t = 2 days:

  1. 1
    Pick a memory strength S. Use days. Fresh, unrehearsed material is roughly S = 1; well-learned material holds longer, around S = 7.
  2. 2
    Enter the time elapsed t. How long since you last studied or reviewed, in the same units as S — here t = 2 days.
  3. 3
    Divide the exponent. −t ÷ S = −2 ÷ 2 = −1.
  4. 4
    Read the retention. R = e^(−1) = 0.368, so you retain about 36.8% of the material.
  5. 5
    Find the half-life. Retention hits 50% at t = S·ln2 = 2 × 0.6931 ≈ 1.39 days — review before then to stay ahead of the drop.

Retention at multiples of S

Because R depends only on t ÷ S, the same fractions apply whatever your memory strength is.

Time elapsed (t)Retention (R)
0100%
0.5 · S60.7%
S36.8%
2 · S13.5%

Why review beats cramming

Each time you successfully recall something, the curve resets higher and, crucially, gets flatter — the next decline is slower, so the memory lasts longer before it needs another review. Reviewing at widening intervals as the curve falls is the core idea behind spaced repetition: a handful of well-timed reviews holds material far better than one long cramming session.

Treat the numbers as a guide, not a guarantee. The exponential curve is a model, not an exact law — actual forgetting depends on how deeply you understood the material, how it connects to what you already know, and how much you slept between study and recall. Meaningful, well-organised material and good sleep both stretch S, flattening the curve on their own.

What is the forgetting curve?
It is Hermann Ebbinghaus’s finding that memory fades quickly at first and then more slowly. This tool models it as R = e^(−t ÷ S), giving the fraction of material you still retain after time t.
How does reviewing change the curve?
Each successful recall resets retention back toward 100% and flattens the next decline, so the memory decays more slowly afterward. Well-spaced reviews are what make spaced repetition effective.
What does memory strength S mean?
S sets how fast you forget, in the same time units as t. A larger S means a slower drop. At t = S retention is always 1/e ≈ 36.8%, whatever the value of S.
When do I forget half of what I learned?
Retention reaches 50% at t = S·ln2, roughly 0.693 × S. With S = 2 days that is about 1.39 days — a useful signal for when to schedule the next review.
Why is retention 36.8% at t = S?
When t equals S the exponent −t ÷ S is exactly −1, so R = e^(−1) = 1/e ≈ 0.368. That is why the S mark is a natural reference point on the curve.
Is this an exact prediction of my memory?
No. It is a simplified model. Real retention varies with the person, how meaningful and well-understood the material is, and how well you sleep, so treat the percentage as a guide.