Could like your help on several noob questions that pop up after reading on the algo on github, which presumably could be easily answered by experts within seconds.
All kinds of advice (including simple Y/N, or even I don’t know, hey most ppl don’t fuss) are appreciated.
Conside the following scenario, describing a single card in a deck with default learning steps “1m 10m”:
- Card in “New” queue
- (Moment 1)
- (Moment 2) Choose “Good”.
- Card in “Learning” queue
- (Moment 3)
- Ten minutes later.
- (Moment 4) Choose “Again”.
- One minute later.
- (Moment 5) Choose “Good”.
- (Moment 6)
- Ten minutes later.
- (Moment 7) Choose “Good”.
- (Moment 8) Card graduated. Card assigned “Interval”, ie a due date.
- Some “Interval” later, card in “Due” queue.
- Review that card.
- (Moment 9) Choose “Good”.
(Moment 10) Card in “Due” queue.- (Moment 10) Card assigned a new “Interval”, ie a due date.
Q1. Is there any mistake on the above simulation?
Using SM2, “Interval” at Moment 8 is derived based on decision at Moment 7 according to Studying - Anki Manual . It means there are only two valid ease: “Good” (shown again the next day) and “Easy” (shown again 4 days later).
Using FSRS 4.5 Anki 24.06.3
Q2. What is the “first rating” moment? Which decision the “Interval” at Moment 8 is derived from? Moment 2, 5 or 7? I think the answer’d be Moment 7, bc all decisions in “Learning” queue should not affect “Interval” right?
The algo doc says … initial stability when the first rating is again
…
Q3. What is a “Again” first rating? How can a card graduate upon “Again”?
Given the default FSRS 4.5 parameters, if I answered “Again” at Moment 2 (the very first decision) and then “Easy” at Moment 7 (the gradating decision). What is the Stability value if any,
Q4. At Moment 1? and
Q5. At Moment 8? (the essential question)
Lastly, I am unsure but I recall reading about addition of new short-term memory parameters in the most recent algo,
Q6. Does FSRS v 5 have feature that could change the answer to Q2 ? Is it implemented in Anki 24.10 beta?