The Anki Mistake 80% of Japanese Learners Make (And the 5-Minute Fix)

You downloaded a 6,000-card Japanese deck on day one. By week two you’re reviewing 180 cards a day. By week six you’re at 320, retention has dropped to 60%, and Anki has become the thing you avoid before bed.

This story plays out in 80% of Japanese Anki users. The cause is almost never “you’re bad at memorization.” It’s one specific setting and one specific habit.

The Mistake: 20 New Cards Per Day

The default in most popular Japanese decks is 20 new cards/day. That sounds reasonable. It is not.

Each new card produces, on average, 10-12 reviews over its first 30 days (assuming standard ease and intervals). 20 new cards/day means you’re committing to ~220 daily reviews at steady state - on top of the 20 new ones. That’s ~45 minutes of Anki, every single day, forever.

Most people can’t sustain that. So they skip a day. The next day shows 380 reviews. They skip again. By week 4 the deck is in maintenance mode - meaning they’re drowning, not learning.

The Fix: 10 New Cards Per Day, Forever

10 new cards/day produces ~110 daily reviews at steady state. About 22 minutes of work. Sustainable for years. Over 12 months that’s 3,650 new items learned - enough for full N3 vocabulary plus 1,000+ kanji.

Why 10 Beats 20 Long-Term

It’s not a linear difference. Cutting new cards in half does not just halve your progress. It improves retention so much that long-term you actually remember more.

New cards/day	Daily reviews	Retention rate	Cards retained at 1yr
20	~220	62%	~4,500
10	~110	87%	~3,200 (with no burnout)
5	~55	92%	~1,700

The 20-card user technically has more cards in the deck. But they’ve forgotten 38% of them, hate Anki, and quit within 8 months. The 10-card user is at 87% retention and still going.

Three Other Settings Worth Fixing

1. Maximum reviews/day - set to 9999

The default cap of 200 hides cards from you. Better to see all of them and make conscious decisions.

2. Lapse new interval - 70%

When you fail a card, Anki defaults to resetting it almost from scratch. 70% means a forgotten 21-day card becomes a 14-day card after you re-learn it - not a 1-day card. Massive review reduction.

3. Easy bonus - 130%

Default 130% is fine. Don’t change it. But also don’t hit “Easy” on every card - you’ll inflate intervals, then start failing en masse 6 weeks later.

One Card, One Fact

If your card front says “食べる” and the back says “to eat (Group 2 verb, transitive, polite: 食べます, te-form: 食べて, past: 食べた)” - that’s six facts on one card. Your brain can’t learn six things at once. Split it: meaning card, conjugation card, example sentence card.

Yes that’s more cards. No, it doesn’t mean more time - you’ll fail the 6-fact card 4 times before it sticks, costing more total reviews than 3 single-fact cards.

The 5-Minute Reset

1. Open Anki → deck options
2. New cards/day: 10
3. Maximum reviews/day: 9999
4. Lapse new interval: 70%
5. Suspend any deck with a backlog > 500. Restart fresh tomorrow.

That’s the fix. The deck stops being a punishment. The retention number stops going down. And six months from now you’re still using it.

Frequently Asked Questions

How many new Anki cards per day for Japanese?

For sustainable long-term study: 10 new cards per day. This produces ~80-100 daily reviews at steady state. Most learners start at 20-30 new cards and quit within 2 months when daily reviews exceed 250 and quality drops below 80%.

Why is my Anki retention so low?

Three causes, in order: (1) too many new cards per day, (2) cards with too much info per side (one fact per card), (3) ease factor stuck at default 250%. Lowering new cards to 10 and using Type-in-the-Answer fixes most cases within 3 weeks.

Should I delete my overdue Anki cards?

Don't delete - suspend. Suspending overdue cards stops them from clogging the queue without losing your progress. Reduce new cards to 5/day for 2 weeks while you clear the backlog, then resume at 10/day.