I recently started dealing with leeches by automatically tagging and suspending them (before I tried to avoid the problem by having Anki not do anything about them, thinking I’ll learn them after enough repetition… eventually studying the leeches started taking up to a third of my study time so I finally decided to deal with them separately. …when I have more time).
I have set the tag & suspend card lapse limit at 16 (the default is 8 but I set it a bit higher because I sometimes fail the card due to distraction, being tired etc., just to give some leeway). Cards which reach the 16th lapse during reviewing get tagged and suspended automatically. This is working as intended.
Cards which already have higher lapse count than 16, after failing them again they get tagged and suspended, but ONLY if they have no sibling cards (e.g. one card per note). The problem is that cards with siblings which have less than 16 lapses get through and remain untagged & not suspended, even though they should be tagged & suspended automatically.
One solution to circumvent the problem would be just to tag & suspend all cards with lapse count 16 or higher manually at once, but I would like to keep the cards in review unless they get failed one more time. Right now I’m dealing with this by checking the ‘Again Today’ filter each day after reviewing and manually tagging & suspending cards which I failed and have lapse count higher than 16 (if they were not tagged & suspended automatically), but I think this should probably be automated as it is done with cards with no siblings. Or am I missing something here?
Tags are for the whole note, not per card, which is why you can’t tag a single card with leech, and thus a note is probably considered as a leech when all its cards reach a certain leech threshold. IMO this is why your cards with 16 or more lapses do not get tagged as leech.
Ah yes you’re right. What would be optimal is that even though both cards would get tagged as leech, only the card with lapses over the threshold would get suspended (which I currently do manually as they are not automatically suspended (nor tagged… I think)).
It’s not a major problem as the amount of cards I need to suspend naturally gets smaller and smaller each day, but something that could be corrected nonetheless (unless it’s too much work) for other users.
Of course best approach would have been just to set the leech threshold at 8 or whatever number the user feels appropriate from the beginning (I wanted to disable the feature at first and just set the threshold at 99…).
Leech detection would happen on your configured 16th lapse, and then twice as frequently after that, so every 8 steps in your case. Cards that are between triggers will not be marked as a leech, and whether a card has siblings or not should not alter the behaviour.
16 is still quite high, and if you have not already done so, I’d recommend you think about ways you could reduce the frequency cards are lapsing - perhaps by investing a bit more time creating mnemonics when initially studying them, tweaking the learning steps, etc.
Now that you mention it I remember there being a feature like that, thank you! Indeed during today’s session a card with a 24th lapse got tagged & suspended but cards with 30th, 19th and 28th didn’t. So everything is working as intended.
At present I’m just suspending the excessive leech cards, and when I have more time (busy with other things these past few months) I’ll try relearn the suspended ones by making them more memorable (they tend to be a certain type of cards or cards with similarities to other cards).
Btw many thanks for creating this app, studying a language I would have never gotten to the level I’m now with out it! Learning how to write Chinese characters with AnkiMobile and a stylus was especially effective not to mention environmental-friendly with all those sheets of paper saved