4 ms·
How would you expect to interact with such a system? Part of the way Spaced Repetition works is to adapt its timings based on whether you got things right or w
by ColinWright 7y ago
How would you expect to interact with such a system? Part of the way Spaced Repetition works is to adapt its timings based on whether you got things right or wrong. Times to recall prompts can be less than a day, down to minutes, are you expecting to get an email every minute until you respond?
I don't see how the system can implement the fundamental underlying algorithm and still work via email. I've written Spaced Repetition systems, one that works on the command line, another that works via a web page, and I'd be interested o now how you envisage something like this working via email.
- Tomte 7y agoI would it expect to work a bit like Anki. Why would I get many emails? I would get an email in the morning with today's questions (or one email per question), and I'd reply "easy", "normal", "hard". That's all that's needed. Timing is unnecessary.
- ColinWright 7y agoI think we have a difference in understanding of how Spaced Repetition works. I don't know the specifics of how Anki works, but what you're describing doesn't match the specifics of the system I use, so I'm guessing here. The way you see it (as I deduce - I could be wrong) is that you get an email (or multiple emails, one per challenge) with a challenge, for which you are expected to know the response. If for simplicity we work on one challenge per email, then you get an email saying something like: What's element 66? You're then on your own to decide whether you know this instantly ("easy"), or with some effort ("normal") or not at all ("hard"). You respond, possibly by clicking a link in the email, possibly by replying with the answer in the email body. Is that what you imagine? (This is related to, but significantly different from the way my system works, so it's pure guess-work based on the clues in your response). If that's what you're thinking, I'm not completely disinterested in creating such a system as an experiment. Do you have a list of things you'd like to memorise so we can experiment? Contact details are in my profile.
- Tomte 7y agoYes, Anki works like this: You get the challenge, you think about it, you tap on it, get the solution, and then you provide feedback whether you were right and how easy it was to remember. This feedback then determines when the card reappers.
- ColinWright 7y agoThat's how my system works, but when I'm using it I get up to 30 or 40 challenges in a session. I imagine that getting that many emails in a short time would be annoying, so the system I use would not immediately convert to an email version. In particular, time gaps for a given "atom" can be as short as a minute or two when first working on it, and again, I can imagine getting emails that close together would be annoying. But my system could be adapted and throttled. Is anyone interested?