Spaced repetition experts: Assume an experiment where students revise study questions twice a week. For each session we inject a set number of new cards (from that session). They can also study outside of class, but are unlikely to.
Conversation
Would love an easy way to simulate/visualize different SRS algorithms, given [number of new questions at each interval, time between intervals, average retention], how many cards end up being available for repetition (falling behind), etc.
1
2
Replying to
I've written some simulations along these lines; for SM0 and SM2 they can be made in a few lines, e.g. gist.github.com/andymatuschak/
If they're only reviewing once a week for 12 weeks, I'd expect algorithm differences not to matter that much.
