Design Spaced Repetition Deck

---
name: design-spaced-repetition-deck
description: Use when a learner needs to build an optimized flashcard deck for long-term retention of vocabulary, grammar patterns, or phrases
source: Hermann Ebbinghaus "Über das Gedächtnis" (1885); Sebastian Leitner spaced repetition system (1972); Anki algorithm (SM-2, Wozniak 1987)
tags: [spaced-repetition, flashcards, retention, anki]
verified: true
---

# Design Spaced-Repetition Deck

Construct a flashcard deck with optimal card design, scheduling intervals, and content structure to maximize long-term retention with minimum review time.

## Why This Is Best Practice

**Adopted by:** Medical schools (Anki dominates USMLE prep), language programs at Duolingo, Babbel, and Pimsleur; military language training at DLI
**Impact:** Ebbinghaus forgetting curve research shows spaced review reduces required repetitions by 50–70% vs. massed practice; SM-2 algorithm users retain 90%+ of material at 6-month intervals with 10–20 minutes daily review
**Why best:** Spaced repetition exploits the spacing effect — memories consolidated during retrieval practice are stronger than those from re-reading; the algorithm surfaces cards just before forgetting occurs

Sources: Ebbinghaus (1885) forgetting curve; Leitner (1972) box system; Wozniak (1987) SM-2 algorithm implemented in Anki

## Steps

1. **Choose a platform** — use Anki (open-source, most algorithm control) for serious learners; Anki-compatible mobile apps (AnkiDroid, AnkiMobile) for on-the-go review.
2. **Define card scope** — decide what one card tests: one word, one phrase, or one grammar pattern; never put multiple facts on one card.
3. **Use cloze deletion for phrases** — "She {{c1::ran}} to the store" tests production in context better than translation prompts.
4. **Add sentence context to every card** — include the source sentence on the back of vocabulary cards; isolates the word but anchors it to authentic use.
5. **Add audio for listening-side recognition** — record or source native-speaker audio for each card's target word; trains phonological form alongside written form.
6. **Add images where possible** — image-word associations bypass the native language; reduces translation dependency.
7. **Set new cards per day limit** — cap at 15–20 new cards daily; more exceeds cognitive load and inflates future review burden.
8. **Calibrate ease ratings honestly** — rate cards "Again" if recall required any effort; overrating easy inflates intervals and causes future forgetting.
9. **Suspend low-frequency cards** — if a card is for a word you never encounter in reading or listening, suspend it; frequency of use drives real retention.
10. **Review at the same time daily** — consistent daily review (even 10 minutes) outperforms irregular long sessions; habit formation sustains the system.

## Rules

- One concept per card — compound cards (word + grammar rule + example) collapse under the "minimum information principle."
- Never skip review days — missing 3+ days causes exponential review debt that discourages return.
- Import pre-made decks only as a starting point — always edit cards to add personal context sentences for stronger encoding.
- Delete cards that have been failed 20+ times with no improvement — some items require a different learning strategy (imagery, story, association) before entering SRS.

## Common Mistakes

- **Making cards too complex** — a card asking for a definition, example sentence, and conjugation simultaneously fails the single-fact principle.
- **Rating "Good" when uncertain** — inflated ratings accelerate intervals prematurely; the card returns when forgotten, creating future avalanche.
- **Importing thousands of cards at once** — overwhelming new-card queues produce review backlogs that kill motivation within two weeks.
- **Never adding personal sentences** — generic example sentences encode less deeply than sentences derived from content you have read or heard.

## When NOT to Use

- Learner needs to pass an exam in under 48 hours — massed practice outperforms SRS for immediate cramming (not retention)
- Content is procedural (conversation skills, writing style) — SRS tests declarative knowledge; procedural skills require production practice
- Learner is at C2 level — vocabulary size is sufficient; focus shifts to collocations and pragmatics beyond flashcard scope