Blog

A four-week exam plan with spaced repetition

· Jason Lang

This is a constructed worked example, not a customer story, we don't have public users yet to draw a real one from. The numbers are realistic for a semester-long course, and the mechanics hold regardless of the exact subject.

The scenario

You have an exam in four weeks covering a semester's worth of material. You convert your notes into roughly 500 flashcards, a mix of definitions, cause-effect relationships, and a handful of cloze deletions for anything you tend to half-remember. You set a daily new-card limit of 20 and a daily review cap of 100, Mneme's defaults, and a 25-minute Pomodoro block for each study session.

Week by week

Week 1. You're mostly introducing new cards, 20 a day, which at this pace takes about three and a half weeks to see the whole deck once. That's fine, you don't need every card on day one, you need the earliest ones spaced out before you forget them. Reviews are light because almost nothing has come due yet.

Week 2. The first cards from week 1 start coming back for their second review, spaced a few days apart based on how well you graded them the first time. Review volume climbs but stays well under the 100-card cap. New cards keep arriving at the same 20-a-day pace.

Week 3. This is typically the heaviest week. Cards from weeks 1 and 2 are due for their second and third reviews, layered on top of new cards still arriving. You're likely to bump against the 100-review cap on your busiest days, which is what the cap is for, it caps the session length, it doesn't drop cards, they carry over.

Week 4. You stop adding new cards a few days before the exam and shift entirely to review. Cards you graded well in earlier weeks have stretched out to week-plus intervals and need less daily attention. Cards you kept fumbling are still coming back often, which is useful information: those are exactly the topics worth a focused re-read, not more blind repetition.

What the daily load actually looks like

The pattern across almost any deck built this way is the same shape: review load rises for the first two to three weeks as earlier cards start coming back, then plateaus or drops as well-learned cards earn longer intervals and only the genuinely hard ones keep showing up often. A study day in Mneme resets at 4am rather than midnight (the same convention Anki uses), so a late-night session doesn't accidentally spill your new cards into the wrong day's count.

Two Mneme settings shape this curve directly. Your desired retention (90% by default) trades review frequency against how many cards you're likely to forget, push it higher and you review more often but forget less; pull it lower and the opposite. Your daily review cap is a session-length safety valve, not a scheduling input, lowering it doesn't change when cards are due, only how many you see in one sitting before the rest wait for tomorrow.

What to do when you fall behind

You will, at some point, skip a day or three. Two things happen. First, overdue cards don't vanish or reset, they just wait, a card due Tuesday that you review on Thursday is graded and rescheduled from Thursday, the same as if Thursday had been its original due date. Second, a stack of overdue cards competes with new ones for your daily cap, so the practical move is to pause new cards for a day or two and let the review pile drain before adding more on top of it.

If a specific card keeps coming back wrong no matter how often you review it, that's what leech detection is for. Mneme flags it after repeated lapses (eight, by default) and can suspend it automatically, so it stops eating review time. The fix is almost always to rewrite the card, it's usually asking two things at once, or split it into two cards, not to keep grinding on it as-is.

What this doesn't fix

Spaced repetition schedules review, it doesn't create understanding from nothing. If a topic doesn't make sense, no amount of well-timed flashcard review will fix that, you need to actually work through it first, then make cards for what you now understand. And none of this rescues a plan started three days before the exam, the whole mechanism depends on gaps between reviews that a three-day runway simply doesn't have room for. Spacing needs weeks, not days, to pay off.

Get 33% off Mneme Pro at launch

Keep reading