I have had the strategy to have a pretty low leech count and set Anki to suspend on leech. Now that I’ve almost finished the deck, I’d like to reintroduce some of those cards back to learn them as well. Would it be a good idea to make a filtered deck of the ‘leech’ tag and then do that slowly? My learning steps are 1d 6d so I don’t know how the ordering would graduate the cards to the main deck if I limit the amount of reviews per day (can I set that setting separately in a filtered deck?). Does anything happen to the ease? I’m thinking if them having a really low ease would be annoying…
Or maybe I should just reset all the leech cards and study them as new? Anyway, I understand I’d probably have to manually unsuspend the cards first, since filtered decks don’t bring in suspended cards?
There are several decisions that you could take when dealing with leeches:
You may consider that the difficulty was due to interference (see rule 11). In this case, nothing else has to be done: you can simply reintroduce the leech, if you think the other interfering card(s) will no longer interfere (because you learnt it), or delay the reintroduction because you haven’t learnt fully the other interfering material.
You may redesign the card. If it became a leech, it may be a sign that it’s too information-heavy, it’s poorly phrased, it may need to be separated in several cards or that it didn’t contain enough mnemonic techniques (see the other rules I posted earlier for a more exhaustive list).
You may have troubles memorizing these cards because you haven’t learnt their content yet, or you may not have a “global view” yet. Quickly reviewing the material outside of Anki for a short amount of time (ie. learning the lesson, reading a few times the poem, …) will quickly ease your memorization with Anki.
Otherwise, the card may simply be not worth it: it’s content is just too hard for you to learn, no matter how much you try redesigning, and it would take a lot of time to learn. It doesn’t mean it’s impossible, just that, when you think about it, you may consider that spending the time for learning a single card that you would usually spend in learning 50 other cards is not worth it.
As for a technical consideration, in all these cases, you should not make a special case of leeches while reviewing. This means that it’s ok to spend more time ahead of the reviewing to prepare these cards if they are harder to learn, but you should review them alongside other cards in their usual deck. This is because interleaving reviews (that is, asking small questions about unrelated cards rather that block-review a bunch of cards that have a common point) produces better results for memorization.
Also, you can’t review a card that is suspended (by definition). Once you want to start reviewing a leech again, you can untag its leech card and unsuspend it. This will allow Anki to mark it as leech once again if it’s still too hard to learn.
I’m hoping most of leeches fall in to the interference category . I’m using a pre-made deck, where all the cards are pretty similar and well designed, they just differ in difficulty (for context, this is a Kanji writing deck for a test for natives), so I’m not too keen on modifying it.
Yeah, this is the philosophy I’ve taken with most of my vocabulary cards. In this deck it’s a bit different, since most of the material ‘has to be known’ for the specific test. As I’m almost finished with it, there’s no new cards left anyway.
The only problem what happens here is as I go through the cards and unsuspend them, and when I then answer correctly on the next review, Anki thinks I haven’t seen the card in months. It gets given an absurdly long interval. I guess I have no choice but to reset the cards, but then they get lumped in with the ‘new’ cards, which are separately at the end of the review session for me…
It’s quite easy to spot these: it’s the cards for which you often give the same answer, but not the good one; or the one when you often say to yourself something like “oh, which one is it again, this one or that one?”. In general, interference is also somehow symmetrical: if you struggle to learn one card because it interferes with an other, you should also struggle learning the other one. If you noticed you started learning a card much better after the leech has been tagged as such, it probably means that leech was interfering.
Well designed does not mean pretty or information-dense. Often, a well designed card is a very simple card with very efficient mnemonics for you. Deck creators (usually) try to design cards that will be easy to remember for everyone, so it’s not necessary in general to change them; but for the cards that are especially difficult for you, it can be a huge difference if you modify them to make them “personal”.
Furthermore, the fact that the cards are similar doesn’t help (with this problem). Similar cards reduces the workload required to identify important information, so in general it’s a good thing, but it also makes it harder to identify each card individually.
Overall, this is something that is easy to try, and not very dangerous: make a backup, and try modifying a single leech card in order to improve it. Then you can see if that helps you or not. Anyways, this is a good practice to learn how to use Anki in a more “advanced” way (which is a useful skill in itself), so it’s not wasted time even if you don’t keep the modified version in the end.
. I’m using a pre-made deck, where all the cards are pretty similar and well designed, they just differ in difficulty (for context, this is a Kanji writing deck for a test for natives), so I’m not too keen on modifying it.