Closed sineptic closed 4 months ago
The user may not review the card in the scheduled date.
but new stability calculated when user complete card. algorithm use predicted R at this moment?
Yes. FSRS is based on DSR model. You can read this for details: https://supermemo.guru/wiki/SuperMemo_Algorithm:_30-year-long_labor#Increase_in_memory_stability_with_rehearsal
Research
Question I don't use anki. Why it need retrievability to calculate new stability (in 4 version)? and why it needed second time to calculate next_review if it already used?