Your current setup breaches one rule: no sequential hint.
If you answer “chair” to “chaise”, you should not get “chaise” to “chair” on the same day. Too easy! Cheating!
You need to dissociate Recto > Verso and Verso > Recto. (by the way, their memorisation speed might be utterly different, requiring a very different number of repetitions)
Having used the flashcards for a while now, it looks like that recto/verso management is even worse than I thought: the 2 directions are not even managed separately as success/failure on the day.
Foreign language students learn much faster foreign > native than native > foreign. Those are 2 different neuronal imprints. They require separate routines.
As things stand, I need to move the learning to Anki. Not a great outcome.
You might want to go to Flashcard Settings and set this to “Mainly Active Recall.” You may get a mix of both directions early on, and yes, on the same day, but it’ll eventually shift to recall only (excluding cards that you’ve selected to be recognition only). This is based on what I’ve observed; perhaps someone else with knowledge can speak to the default setup options and confirm.
You are correct that Readlang’s spaced-repetition-system doesn’t store progress separately for foreign > native, which I call passive recall, vs native > foreign, which I call active recall.
My reasoning for this is:
the two directions aren’t completely distinct knowledge items. Knowing one direction will help you learn the other direction to a large degree. So storing them completely separately doesn’t feel like an ideal solution either.
active recall is harder than passive recall, and for any given word, I think it’s very very likely that if you can manage to do active recall, then you can also do passive recall. So for people who care about learning both directions, I think it’s a reasonable approach to switch to only practicing active recall during the practice sessions once you have reached a certain level with a word/phrase. This is what Mastery does, and what the Classic Flashcards does when in Mainly Active Recall mode. (And of course if you continue reading with Readlang which is the primary activity for most people, then you will also get more passive recall practice as the word comes up again in the texts you read, which I think makes it more justifiable to keep your practice sessions as mainly active recall, since that’s what you miss out on while reading)
You are correct that Readlang will often give you an active recall after a passive one within the same session. When this happens, if the active recall is successful, it will have a negligible (practically zero) effect on the scheduled future practice date precisely because your last exposure to this word was so recent. The algorithm takes the duration between practices of a word into account when deciding what the next interval should be. The reason that Readlang sometimes gives you active recall shortly after passive of the same word within the same session is that it:
gives you a bit of practice for a more genuine active recall in future
feels satisfying to accomplish an “active recall” (even if the spaced repetition algorithm won’t treat a successful recall that seriously)
Another point worth clarifying is that the spaced-repetition-system will schedule a word further into the future if you get an active recall correct vs if you get a passive recall correct, assuming the time since the previous practice is the same.
I’m not trying to say that this is all completely perfect, there is surely room for improvement, but I think it’s a reasonable approach. I think one area where it’s likely not to work that well is during the period after switching from doing all passive recall (Blitz mode and Classic Flashcards in passive only mode) to mainly active recall. In that case the system will have an overly optimistic view of your abilities and you’ll probably fail a lot of challenges which will reset progress on lots of words. If you then continue on mainly active practice it will gradually self correct.
Question: would anyone be interested if I wrote up an AI-assisted detailed explanation of the spaced repetition algorithm and how all the different settings and practice modes affect it?
Question: would anyone be interested if I wrote up an AI-assisted detailed explanation of the spaced repetition algorithm and how all the different settings and practice modes affect it?
I’d definitely be interested in some nerd stuff about the algorithm
But as I mentioned before: I’m quite happy that the user facing side of flashcards is simple and minimal. I still enjoy using it very much! (now exclusively with the Mastery mode)