Why Cramming Fails and Spacing Works: The Science of SM-2
If you have ever aced a quiz the day after cramming, then drawn a total blank on the same material two weeks later, you have experienced the difference between short-term recognition and long-term memory. Cramming feels productive because it produces an immediate result. But the psychologist Hermann Ebbinghaus documented this exact failure mode back in 1885 with his famous "forgetting curve" — showing that without reinforcement, we lose the majority of newly learned information within days, regardless of how well we felt we knew it at the time.
The fix Ebbinghaus discovered, and that modern cognitive science has since refined considerably, is deceptively simple: review the material again right before you would otherwise forget it. Do that repeatedly, at gradually increasing intervals, and the forgetting curve flattens dramatically — each successful review makes the next forgetting point arrive later. This is the entire premise behind the tool on this page.
What SM-2 Actually Does
The reviewer above runs on the SM-2 algorithm, developed by Piotr Woźniak in 1987 for the SuperMemo software — the same core scheduling logic that underlies most modern spaced repetition tools. Every card you add is tracked with three numbers: an ease factor (how easy the card has been for you historically), a repetition count (how many times in a row you've recalled it successfully), and an interval (how many days until its next scheduled review).
Each time you review a card, you rate your recall on a scale from "Again" (you blanked completely) to "Easy" (instant, effortless recall). That rating feeds back into the algorithm: a low rating shrinks the interval and lowers the ease factor, bringing the card back sooner and more often. A high rating stretches the interval further out and slightly raises the ease factor, so a card you consistently know well shows up less and less — freeing your attention for the material that still needs work.
Why This Beats a Fixed Review Schedule
A common but far less effective approach is reviewing an entire deck on a fixed schedule — every day, or every Sunday, regardless of how well each individual card is already known. This wastes enormous amounts of time re-reviewing material you have already mastered, while under-reviewing the handful of genuinely difficult cards that actually need the attention.
SM-2 solves this by treating every card as its own independent scheduling problem. A vocabulary word you nailed five times in a row might not need reviewing again for two months. A tricky formula you keep forgetting might come back tomorrow, then the day after, until it finally sticks. Over time, your daily review queue becomes almost entirely composed of the material genuinely at risk of being forgotten — which is exactly where your study time should go.
How to Rate Yourself Honestly
The algorithm only works as well as your ratings are honest. It's tempting to rate everything "Good" to move through cards quickly, but this defeats the entire mechanism. A useful mental test: did you produce the answer confidently and immediately, with no hesitation? That's "Easy." Did you get there, but it took a genuine effortful pause? That's "Good" or "Hard" depending on how much of a struggle it was. Did you guess, blank, or get it wrong entirely? That's "Again" — no exceptions, even if the answer feels obvious in hindsight once you see it.
Rating "Again" is not a failure. In fact, the moment of genuinely failed retrieval, followed immediately by seeing the correct answer, is one of the most effective learning events in the entire system — psychologists call this effect "desirable difficulty." The struggle is where the memory gets rebuilt stronger, not a sign you should have studied more beforehand.
Writing Cards That Actually Work
Keep each card to one fact. A card asking "What are the causes and effects of the French Revolution?" invites vague, unratable recall. Split it into several narrow cards instead — one cause per card, one effect per card.
Phrase the question so there's exactly one correct answer. Ambiguous cards make honest self-rating nearly impossible, since you can't tell if you actually recalled the intended answer or just something plausible.
Add context sparingly, not answers. A little context in the question helps trigger the right memory ("In cell biology, what organelle...") without accidentally giving away the answer itself.
A Realistic Daily Habit
The single biggest reason spaced repetition systems fail for students isn't the algorithm — it's inconsistency. A queue that's ignored for two weeks arrives back all at once, overwhelming and demotivating. The fix is small and daily: open this tool for five to ten minutes each day, every day, treating it the same way you'd treat brushing your teeth. Consistency at low volume beats intensity at irregular intervals every time.
Already have deadlines loaded into your semester? Pair this tool with our Syllabus to Calendar Converter — once your exam dates are on your calendar, add the relevant material to your review deck as soon as each topic is covered in class, so the spacing has time to actually work before the exam arrives.
Want a more complete weekly system that combines spaced review with structured planning? Check out our Student Reset Planner and other premium resources.