Reading the forgetting curve like a chart, not a verdict
A short essay on what the forgetting curve actually tells you — and the daily Lemnly routine that turns it from an enemy into an ally.
The first time most learners hear about the forgetting curve, it lands as bad news: you will forget most of what you learned in 24 hours, and almost all of it in a week. That is true. It is also not the whole story.
What the curve shows
Ebbinghaus’s original chart plots retention over time. After one day, you remember about a third of what you learned. After a week, less than a quarter. The slope is brutal — but it’s also the slope of passive retention. Every retrieval flattens the next one.
What the curve hides
Three things the curve does not tell you, that matter more than the curve itself:
- Stability grows. The second time you recall a card, the next forgetting curve is flatter. The fifth time, much flatter. By the tenth successful recall, the curve is nearly horizontal. The word is yours.
- Difficulty matters. Some cards are intrinsically harder for your brain — abstract words, words too similar to others you know, words in unfamiliar scripts. A good scheduler models this per card so the hard ones get more practice and the easy ones get less.
- Context anchors retention. A bare word recalled in isolation forgets faster than the same word recalled alongside the sentence you first met it in. This is why Lemnly attaches the source sentence — it’s not decoration, it’s the anchor that stops the curve from collapsing.
The practical move
Stop reading the curve as "I am going to forget everything." Read it as: "If I retrieve it right before I would have forgotten it, I will remember it longer next time." That is the whole job. Find a tool that schedules the next retrieval well, and do the retrieval.
Your daily five minutes, in detail
Here’s the routine. It really is five minutes.
- Open the app. Today’s queue is the cards FSRS has decided are at the edge of your recall. There is no decision to make — just hit start.
- Answer in your head before you flip. Even a wrong answer is a retrieval attempt, and retrieval is what writes the card deeper.
- Be honest with the buttons. Again means you didn’t know it. Good means you knew it without strain. The scheduler is only as smart as your feedback.
- Stop when the queue is empty. Don’t binge new cards. Don’t hunt for "easy" wins. Close the app and go back to reading.
Why "just one more session" backfires
FSRS schedules cards to land near the edge of your forgetting curve. If you do a second session three hours later, almost nothing is at the edge yet — you’re reviewing cards you’d have remembered anyway, teaching the scheduler nothing new, and using up willpower you’ll need tomorrow.
The forgetting curve rewards patience. One five-minute session a day beats two ten-minute sessions every other day, every time.
What to do when you fall off the curve
You will miss a week. It happens. When you come back:
- Don’t bulk-review the backlog in one sitting. Spread it across three or four days.
- Expect more "Again" presses than usual. That’s fine. The scheduler will recover the stability over the next week.
- Don’t add new cards until the backlog is clear. New cards on top of a backlog is how Anki users burn out — Lemnly will warn you, but the warning is only useful if you listen.
The takeaway
The forgetting curve isn’t a verdict on your memory. It’s a graph of when to come back. Lemnly’s whole job is to plot that graph for every card you own and ping you exactly when the next retrieval is worth your time. Your whole job is to show up.
