Forgetting is fastest right after learning
💎CORE ARGUMENT
Most of what I am going to forget, I forget in the first day or two. After that the loss slows to a crawl. So the cheapest moment to look at something again is soon, while a little effort still brings it back.
That is the whole seed. The curve and its formula are in the concept note. I do not need to re-derive them here.
What I can add is the one time I tested it on myself. Until April my lecture notes were written on Tuesday night and next opened in the week of the exam, by which time they read like someone else's. The midterm was a 68. For the six weeks before the final I looked at each lecture again within two days and then at widening gaps, and the final was a 74. Six marks, one student, a calmer spring: I would not publish it. But it matched what the curve says should happen.
The part I have not worked out is why this feels wrong while I am doing it. Reviewing on Thursday what I heard on Tuesday feels wasteful, because I still half know it. That half-knowing is the whole point, and it is also why I skip the review.
It breaks down for anything I use daily, which needs no schedule, and for anything I never understood in the first place. Review keeps hold of what was learned. It does not learn it for me. It is the reason behind 💎 Notes should be written for the person who has forgotten everything, and a cousin of 💎 A habit takes months and not three weeks: both are about a curve that is steep early and flat late, and about my habit of judging by the steep part.
🧱BUILT FROM
🌍WHY IT MATTERS
One rule so far: look again within two days, briefly, before deciding whether something is worth keeping. Everything else about spacing I am taking on trust until the autumn module.
☑️NEXT STEPS
- Look again at each lecture's notes within two days
- Compare what I recall from reviewed and unreviewed lectures before the exam
- Revisit this note once the autumn module has covered memory properly
📖 Ebbinghaus on forgetting: one man, lists of nonsense syllables, and a careful measure of how much less effort the second learning took.
💎 Notes should be written for the person who has forgotten everything is what this implies for notes. 💎 A habit takes months and not three weeks has the same shape turned over: fast change at the start, then a long slow tail that I mistake for failure.
Loss is front-loaded. A review placed just after the steep drop costs little and resets the curve at a shallower slope. The same review placed weeks later amounts to learning the thing again.
His material was meaningless on purpose. Things that connect to what I already know fade far more slowly, so the steep curve is the worst case, and applying it to everything would have me reviewing my own address.
Tested once in 🧪 Spaced review of lecture notes. Stated properly, with the formula, in 🧩 Forgetting curve.