Skip to main content

Overview

Anglio uses FSRS (Free Spaced Repetition Scheduler), a modern spaced repetition algorithm that optimizes your vocabulary learning by scheduling reviews at the ideal time.

How It Works

When you review a vocabulary card, you rate how well you remembered it. Based on your rating, FSRS calculates when you should see the card again:

  • Again (red): You forgot the word. The card will appear again in 1 minute.
  • Hard (orange): You remembered with difficulty. Short interval before next review.
  • Good (green): You remembered normally. Card scheduled according to your learning step setting.
  • Easy (blue): You remembered effortlessly. Card graduates immediately with a longer interval.

Card States

Cards progress through different states as you learn them:

┌─────────────────────────────────────────────────────────┐
│ │
│ ┌─────┐ Good/Easy ┌──────────┐ │
│ │ New │ ──────────────────────▶│ Learning │ │
│ └─────┘ └──────────┘ │
│ │ │
│ │ Graduate │
│ ▼ │
│ ┌────────────┐ Again ┌─────────┐ │
│ │ Relearning │◀───────────────│ Review │ │
│ └────────────┘ └─────────┘ │
│ │ ▲ │
│ │ Graduate │ │
│ └─────────────────────────────┘ │
│ │
└────────────────────────────────────────────────────────┘
  1. New: Cards you haven't studied yet
  2. Learning: Cards in the initial learning phase (short intervals)
  3. Review: Graduated cards with longer intervals (days/weeks/months)
  4. Relearning: Cards you forgot during review, back to short intervals

Learning Phase

When you first learn a new word, it goes through a short learning phase before "graduating" to longer review intervals.

How Graduation Works

With the default settings (10 minutes):

  1. First review: Press Good → card scheduled for 10 minutes
  2. Second review: Press Good → card graduates to Review state

That's it! 2 Good presses to graduate a new card.

What Each Button Does

ButtonIntervalEffect
Again1 minuteResets progress, try again soon
Hard~5 minutesDelays but doesn't advance
Good10 minutes*Advances toward graduation
Easy1+ daysGraduates immediately

*The Good interval is configurable in your profile settings.

Settings

You can customize your spaced repetition settings in Profile → Spaced Repetition:

Learning Step

This controls how many minutes before you see a card again after pressing Good during the learning phase.

  • Default: 10 minutes
  • Range: 1 to 1440 minutes (24 hours)
  • Tip: Lower values (5-10 min) for quick sessions, higher values (30-60 min) for spaced-out learning

The Again button always uses 1 minute, regardless of this setting.

Tips for Effective Learning

Use All Four Ratings

Don't just press Good every time:

  • Press Again when you genuinely forgot
  • Press Hard when you struggled but eventually remembered
  • Press Good for normal recall
  • Press Easy when the word is already very familiar

Honest ratings help FSRS optimize your review schedule.

Consistent Daily Practice

Short daily sessions are more effective than long occasional sessions. FSRS works best when you review cards around their due time.

Don't Worry About Forgetting

Pressing Again isn't failure—it's data. FSRS uses this information to schedule the card more appropriately. The algorithm adapts to your actual memory patterns.

Understanding Your Progress

Retention Rate

Your retention rate shows what percentage of review cards you remembered (pressed Hard, Good, or Easy vs Again).

  • 85-95%: Ideal range
  • Below 85%: Consider shorter learning steps or more frequent reviews
  • Above 95%: You might be reviewing too frequently

Card Statistics

Each card tracks:

  • Stability: How long the memory is expected to last
  • Difficulty: How hard the card is for you (1-10)
  • Reps: Total successful reviews
  • Lapses: Times you forgot (pressed Again)

About FSRS

FSRS (Free Spaced Repetition Scheduler) was developed by Jarrett Ye and is based on the DSR (Difficulty, Stability, Retrievability) memory model. It uses the forgetting curve to predict when you're about to forget a word, scheduling reviews just in time.

FSRS is more accurate than older algorithms like SM-2 because it:

  • Models memory decay mathematically
  • Adapts to individual learning patterns
  • Uses 4 ratings instead of 2 for finer granularity

Learn more at the FSRS GitHub repository.