Shun
状元
Philosophical asides like this one are worthwhile, so thanks for making it. 
Caveat: Please read the following only if you're interested in the difficulties of bringing technology into harmony with the way humans use it. If you're happy with the way you're doing things now, perhaps it's better to skip what I've written here.
Certainly, I agree with you that technical knowledge and optimization can contribute to a successful studying session. Of course, I also have the desire to optimize that. But I think a stack of cards where you're just varying the intervals at which cards are repeated (longer if you knew them well, shorter if you didn't know them), is all the brain needs and can't really be improved on. The brain wants to repeat cards at increasingly long intervals until you don't need to anymore—like Anki does. The only other thing one should optimize is more long-term, if you're studying a large deck of 10'000 cards, say, how these cards are apportioned in digestible chunks until you know them all. But I think this is already too much configuration, i.e. it's removed from the reality of studying, since the source material you're studying on the side may not match the order in which Pleco is configured to introduce new cards from the stack. So the only thing you can do is to select cards manually, in stacks of 50, based on the needs of what you were going to read, and then study them with the stack of cards method described above.
I'd say a learner who has optimized their memorization through practice doesn't really care about many settings, they just want to see the cards at the right intervals, and that's all they need to learn them well. Everything else about the SRS seems more like a distraction to me. For example, depending on the difficulty of learning a particular set of cards, the "Learned" threshold should be higher if the cards were hard and lower if they were easy. But there is only one "Learned" threshold that you can configure for all cards. Some cards will then be gone too quickly, before you've learnt them well, and others will bore you by showing up too often. (There is also the Easiness of a card, but his can again be too rigid because Easiness can also change over time.) So already, the technology doesn't suit the needs of the user precisely anymore, or there's a big risk that it doesn't. My point being, any more complexity here is quite hard to bring into line with your subconscious needs.
Of course, there can still be individual variations in the optimal way of studying and other motivations involved. I could also imagine someone improving their settings just to be able to think their Pleco is "better configured" than someone else's—not everyone, just sometimes. (It's a temptation.) But if you think about it, when you remove all the other motivations besides studying well, you will probably end up with a pretty simple system.
About your original suggestion, to make the multiple choice harder by offering very similar, but incorrect words as possible solutions, I think it's pretty smart. But how, for example, can you keep the user from getting the impression that when they have four choices, two of which are very similar, that it had to be one of these two cards? So that wouldn't really work, would it? You could give four very similar choices, perhaps. But that would only teach detail knowledge, then. You could perhaps vary this with today's random way of presenting multiple choices and give two similar cards in a misleading way, by making them both wrong choices. My point being, also here, it can be quite tricky to bring technical gimmicks into line with what the user really needs.
I only advocate that one stay conscious of these needs and that it often isn't easy to figure them out and put them into a setting. It’s fine with me if Pleco offers “too many” settings. Those who want them get them, and those who don’t need most of them can ignore them.
Cheers,
Shun
Caveat: Please read the following only if you're interested in the difficulties of bringing technology into harmony with the way humans use it. If you're happy with the way you're doing things now, perhaps it's better to skip what I've written here.
Certainly, I agree with you that technical knowledge and optimization can contribute to a successful studying session. Of course, I also have the desire to optimize that. But I think a stack of cards where you're just varying the intervals at which cards are repeated (longer if you knew them well, shorter if you didn't know them), is all the brain needs and can't really be improved on. The brain wants to repeat cards at increasingly long intervals until you don't need to anymore—like Anki does. The only other thing one should optimize is more long-term, if you're studying a large deck of 10'000 cards, say, how these cards are apportioned in digestible chunks until you know them all. But I think this is already too much configuration, i.e. it's removed from the reality of studying, since the source material you're studying on the side may not match the order in which Pleco is configured to introduce new cards from the stack. So the only thing you can do is to select cards manually, in stacks of 50, based on the needs of what you were going to read, and then study them with the stack of cards method described above.
I'd say a learner who has optimized their memorization through practice doesn't really care about many settings, they just want to see the cards at the right intervals, and that's all they need to learn them well. Everything else about the SRS seems more like a distraction to me. For example, depending on the difficulty of learning a particular set of cards, the "Learned" threshold should be higher if the cards were hard and lower if they were easy. But there is only one "Learned" threshold that you can configure for all cards. Some cards will then be gone too quickly, before you've learnt them well, and others will bore you by showing up too often. (There is also the Easiness of a card, but his can again be too rigid because Easiness can also change over time.) So already, the technology doesn't suit the needs of the user precisely anymore, or there's a big risk that it doesn't. My point being, any more complexity here is quite hard to bring into line with your subconscious needs.
Of course, there can still be individual variations in the optimal way of studying and other motivations involved. I could also imagine someone improving their settings just to be able to think their Pleco is "better configured" than someone else's—not everyone, just sometimes. (It's a temptation.) But if you think about it, when you remove all the other motivations besides studying well, you will probably end up with a pretty simple system.
About your original suggestion, to make the multiple choice harder by offering very similar, but incorrect words as possible solutions, I think it's pretty smart. But how, for example, can you keep the user from getting the impression that when they have four choices, two of which are very similar, that it had to be one of these two cards? So that wouldn't really work, would it? You could give four very similar choices, perhaps. But that would only teach detail knowledge, then. You could perhaps vary this with today's random way of presenting multiple choices and give two similar cards in a misleading way, by making them both wrong choices. My point being, also here, it can be quite tricky to bring technical gimmicks into line with what the user really needs.
I only advocate that one stay conscious of these needs and that it often isn't easy to figure them out and put them into a setting. It’s fine with me if Pleco offers “too many” settings. Those who want them get them, and those who don’t need most of them can ignore them.
Cheers,
Shun
Last edited:
