Hermann Ebbinghaus (1885) showed that memory decays roughly exponentially: a steep drop within the first day, then a slow flattening.

Ebbinghaus, a German psychologist, ran the first rigorous memory experiments on a single subject — himself. He memorized long lists of "nonsense syllables" (meaningless three-letter combinations like WID or ZOF, chosen so prior knowledge couldn't help) and then tested how much he could relearn after various delays. Plotting retention against time produced a curve that falls fast at first and then levels off. The practical lesson: most of what you forget, you forget quickly — so the timing of review matters enormously.

Very well — a 2015 direct replication reproduced his data almost exactly, with his equations explaining ~98.8% of the variance.

Because Ebbinghaus tested only himself with one method, his curve could have been a quirk. Murre and Dros (2015, published in the journal PLOS ONE) painstakingly repeated his procedure and found their new data closely matched the 1885 original — an unusually strong replication for a 130-year-old finding. One wrinkle: they observed a small "jump up" in retention around the 24-hour mark, consistent with sleep helping to consolidate memory. This is well-replicated science, not a preliminary claim.