'Bookmark It and Forget It' Isn't Your Fault: A Plain-English Guide to Spaced Repetition

2026-08-21 · 映灵浏览器 Inling · 中文版

Open your bookmarks and scroll back a few pages.

How many of those articles can you still name? How many would make you stop mid-scroll and think, "wait — I've read this already"?

It's not your fault, and it's not laziness. This is simply how brains work: anything you never have to retrieve gets treated as noise and quietly thrown out. Psychologists have studied this for well over a century, and the finding hasn't budged: reading something and remembering it are two different things.

The forgetting curve you've probably heard about

In 1885, the German psychologist Hermann Ebbinghaus ran a famously masochistic experiment on himself. He invented thousands of nonsense syllables — things like "DAX" and "QEW" — memorized them, then tested himself at various intervals to see how much had survived.

The nonsense was the point. Meaningful material sticks better, but that's common sense; Ebbinghaus wanted to measure memory itself, stripped of understanding. What he got was the famous forgetting curve: roughly 58% retained after twenty minutes, a third after a day, less than a quarter after a week.

And those were syllables he'd deliberately set out to memorize. The casual skim of an article — the way most of us "read" online — fades even faster.

Bookmarked articles fare worse still. Most never even got a full first read; you just filed them under "later." And "later," more often than not, never comes.

Spaced repetition: catch the memory just before it goes

The fix is simpler than you'd expect: don't review right after learning something; review right before you're about to forget it.

More precisely, you revisit the material at the moment it's about to slip away but still within reach. Every successful recall slows the decay a little more. The first review might come a day later, the second after three days, then a week, two weeks, a month — the intervals keep stretching.

The math is almost absurd: five or six reviews, a few minutes total, and a fact can settle into long-term memory for years. That's the scheduling logic inside flashcard apps like Anki, and it's why medical students, bar exam candidates, and language learners swear by it. The sheer volume they need to retain can't be crammed in by brute force.

FSRS: a smarter answer to "when should I review?"

The older algorithms were blunt instruments. Take SM-2 — the 1980s SuperMemo algorithm Anki shipped with for years. You grade yourself after each card, and the system stretches or shortens the next interval by fixed multipliers. It works, but it assumes everyone's memory behaves the same way and every card is equally hard. Neither is true.

FSRS (Free Spaced Repetition Scheduler), built by the open-source community in recent years, takes a different approach. Instead of one curve for everyone, it models each card individually, with three parameters:

• Stability — how much longer this memory is likely to hold
• Difficulty — how hard this particular item is for you
• Retrievability — how likely you are to recall it right now

Every review you finish and every rating you give updates the model and recomputes the ideal next interval. In benchmarks on real review histories, FSRS holds the same retention with roughly 20–30% fewer reviews than SM-2. Every review you get to skip is time back in your day.

Let's be honest: what kills the habit isn't the algorithm — it's making cards

Spaced repetition is easy to grasp, and the tools are free. So why do so few people stick with it?

Because the first half of the classic flashcard workflow is a grind. Finish a 5,000-word article and you're expected to decide which points are worth keeping, rewrite each one as a clean question-and-answer pair, keep every card small and unambiguous — a vague "what is X?" barely counts as a card — and then type the whole batch into Anki. That's twenty minutes of card-making per article, minimum.

Most people don't quit reviewing. They quit the prep work. Two weeks in, patience runs out, the deck has a dozen cards, and the app joins the bookmarks in gathering dust.

The direction Inling is exploring: from reading to practice

Inling aims to shorten the distance between reading something useful and starting to practice it. Knowledge organization, learning paths, lectures, quizzes, flashcards and review around saved sources are in development. These AI learning features are not available in the current beta.

For now, use Inling to browse, read and bookmark pages. To begin spaced practice today, select a few points worth retaining, write questions in your notes or a tool that supports review scheduling, and give honest feedback about what you recall.

Tools can help organize material, but understanding, answering and sustained practice still take your effort. Saving a link preserves a source; it does not mean you have learned it.

Inling’s beta is open on HarmonyOS, iOS and Android. Choose your device to install the browser beta. AI learning remains in development.

Further reading: FSRS is open source — you'll find it under the open-spaced-repetition org on GitHub. For a systematic comparison of study techniques, Dunlosky et al.'s 2013 paper, "Improving Students' Learning With Effective Learning Techniques," remains the standard reference.