The basic idea
First described through memory experiments in the 19th century, the forgetting curve shows that recall of newly learned material falls off sharply within hours and days if it is never revisited, then continues to decline at a slower rate. The exact numbers vary by person and material, but the shape—fast early drop, slower long-term decline—shows up consistently.
The practical implication is blunt: reading something once and moving on is close to the worst way to make it stick, because the first review window closes quickly.
Why review timing changes the curve
Each time you successfully retrieve a memory, the curve for that memory flattens—it takes longer for the next review to become necessary. This is why spaced repetition reviews get further apart over time for material you keep answering correctly, and stay close together for material you keep getting wrong.
This is different from simply reviewing everything on a fixed daily schedule. A flat schedule ignores that some memories are already stable and others are not, wasting time on the former while under-serving the latter.
What this means for how you study
If the forgetting curve is steepest right after first learning something, your first review should happen soon—not weeks later. After that, the useful gap between reviews grows as the memory proves itself stable, rather than staying fixed.
This is the whole logic behind adaptive scheduling. Cueable uses an FSRS-based algorithm that estimates where each card sits on its own forgetting curve from your history, and times the next review to arrive close to the point where you would otherwise start to forget.
The forgetting curve is not an excuse
It is common to hear that the forgetting curve means a memory will fade anyway, used as a reason to skip review. The point of the curve is the opposite: it explains exactly when review is most valuable, so that a small, well-timed effort can hold onto material that would otherwise have been wasted.