TL;DR
When Burying → Bury review siblings is enabled and one sibling type is harder than the others, Anki keeps introducing extra siblings of that harder type to fill the Daily Limits → New cards/day quota while other siblings are buried. This results in being fed more new information in a day than intended (e.g. with New cards/day = 40, a 30/10 split introduces 30 new notes in a day, while a 20/20 split introduces only 20 new notes). A New notes/day option, used together with New cards/day, would cap how many notes can be introduced per day and prevent overload.
Context
Burying
Bury new siblings = ON
Bury review siblings = ON
Bury interday learning siblings = (optional)
Daily Limits
New cards/day = 40
Maximum reviews/day = 9999
New cards ignore review limit = ON
Example
Japanese vocab deck with 2 card types: one harder (recognition) and one easier (recall).
Day 2 → 20 more notes (20 recognition) + recall from Day 1 (delayed by Bury new siblings).
Day 3 → 20 more notes (20 recognition) + 0–20 recall from Day 2 (some may be buried due to Bury review siblings).
Day N → Each day adds 20 new notes, with siblings flowing in naturally when they’re not buried.
Crucially: if siblings are buried by reviews, they just wait. Anki should not introduce extra new notes to fill the New cards/day quota.
What happens now
The harder card type produces more reviews.
Bury review siblings hides its siblings during the session.
Anki then introduces more new notes to hit the New cards/day limit.
This means more than the intended number of new notes are introduced in a day (e.g. 30 instead of 20).
Why this is a problem
Having siblings delayed because they’re buried is fine and expected.
The problem is Anki compensating by introducing too many new notes.
This overloads study sessions with new material, skews balance, and disrupts pacing across sibling types.
Proposal
Add a New notes/day option alongside New cards/day.
Example: New notes/day = 20 and New cards/day = 40.
Each day introduces at most 20 notes (their first sibling).
Other siblings follow naturally on later days (via burying).
If siblings are buried by reviews, they simply wait. No extra notes are added to reach the card quota.
Why this helps
Workload is tied to notes, not just raw cards.
Prevents overload when one sibling type is harder and Bury review siblings is enabled.
Lets learners keep burying ON (to avoid seeing siblings back-to-back) without breaking pacing or balance.
Current workarounds (and their downsides)
Manually stop after N notes
You can keep an eye on how many new notes you’ve introduced and stop yourself at 20. Con: easy to lose track mid-session; error-prone.
Filtered decks per card type
Build a filtered deck daily with queries like card:TypeA is:new (limit 20) and card:TypeB is:new (limit 20). Con: clunky to rebuild every day; limited to two queries in stock Anki.
Disable “Bury review siblings”
With it off, both sides of a note can appear on the same day, so no siblings are skipped and no over-compensation happens. Con: you may see both siblings back-to-back, which many learners want to avoid.
Why a core feature is better
All of the above require vigilance, manual work, or accepting trade-offs. A simple New notes/day option would solve the problem cleanly and predictably:
No manual counting.
No filtered-deck micromanagement.
No need to choose between back-to-back siblings vs. overload.
I think you’ve got a couple mis-assumptions, and you don’t need a new feature to get where you’re trying to go.
Your premise is somewhat incorrect. Anki doesn’t emphasize one card type/sibling over the others because it is harder – it simply introduces the siblings in order. If you want the harder sibling to be introduced later, you can re-order your card types.
[And while it may be the case for you, it is not a universal truth that Recognition cards are always harder than Production cards – or that card1 is always harder than card2, since Anki can be used to study other things besides vocabulary.]
In your example, I don’t think you’re connecting with how cards are gathered – Deck Options - Anki Manual . New cards can’t cause Review cards to be buried, because the Review cards are gathered into the study queue first. So, when you have all of the burying options enabled, you should expect that all of your New cards will come from separate notes – and that those will be different notes from the notes you are studying Review cards from.
You should build that expectation into your New card limit. If you set your daily New card limit to 40, those will be cards from 40 separate notes. If you can’t handle that, and you intend to limit yourself to introducing cards from no more than 20 notes per day – you need to set your limit to 20 cards per day.
Your simplified example also assumes that after graduating from Learn, Review cards will always be due the next day. While that may have been more common with the SM-2 algorithm (with defaults and graduation with Good) or with multi-day learning steps – it’s frequently not the case, considering the impacts of FSRS parameters, Desired Retention, load balancing, smart fuzz, graduation with Easy, etc.
For every newly graduated Review card1 that is not scheduled for the next day, its still-New sibling card2 will be at the front of the queue, and a priority to introduce before introducing another note’s card1. It’s still a New card, even if it comes from a note where a sibling has already been introduced.
You missed the main way to limit your card1’s to a certain maximum, and fill in the rest of your daily New cards with card2’s – split your card types into separate subdecks with different limits. You could have your overall New card limit be 30, and for the card1 subdeck set a “This deck” limit of 20. Then you’ll get no more than 20 New card1’s each day, with up to 10 card2’s if they are available.
Another option is to only unsuspend a certain number of notes at a time, so you’ll never get too many card1’s in a day. Another is to “stagger” your card1’s and card2’s – start introducing the card1’s first, and don’t release the card2’s until a few days later (you can do this either with suspending or with subdecks/limits).
Thanks for taking the time to respond! I really appreciate you clarifying some of the mechanics.
Whoops, this was a mistake when writing the TL;DR, I didn’t mean to suggest Anki “prioritizes” based on difficulty. What I meant is that more notes end up being introduced through the harder type, since the easier type gets buried by review siblings and has to wait. That desync is the core issue.
Absolutely, fair point. My example assumes recognition is harder, but I’m pretty sure the same imbalance could happen in reverse.
Exactly! this is the crux of what I was trying to point out. I never said that new cards bury reviews, quite the opposite, I said that reviews bury new cards. That’s why I singled out “Bury review siblings.”
You mentioned that I should expect “new cards/day = new notes/day,” but that isn’t always the case. Once bury review siblings is enabled, new cards of seen notes will start to lag behind. This is the limitation I’m trying to highlight.
If I keep struggling with, say, card1 of Note A, then its sibling (card2 of Note A) will remain buried day after day. That’s fine and expected. But when card2 finally becomes available, it gets gathered before any fresh card1’s from other notes. At that point, the daily quota of “new cards” is partly filled by old siblings (like card2 of Note A), so the number of new cards introduced can be greater than the number of new notes introduced.
That behavior is exactly what causes the mismatch. It’s not that reviews are being buried by new cards: it’s that buried siblings accumulate, then “jump the line” in the new card queue when they’re released. The result is days where new cards > new notes, and that uneven pacing is what I think a “New notes/day” cap could smooth out.
And this ties directly to workload management. Right now I already push my daily limit up to 40 new cards as a workaround. It’s not that things would literally take twice as long otherwise, it’s that I can’t optimize the daily load across multiple sibling types under the current system.
Ideally, I should be able to catch up on notes I’ve already seen. I’m fine with going over the card limit for those while also keeping a predictable cap on how many new notes get introduced each day. Without that, lowering the limit just means I see fewer new notes per day and pacing across sibling types gets thrown off, so the whole deck takes longer to stabilize.
A notes-based cap would fix this: I’d keep the higher new card limit (to let siblings flow in when they’re unburied), while also keeping a ceiling on how many fresh notes are introduced in a single day. That way I don’t overshoot on workload, but I also don’t stall progress unnecessarily.
I tried splitting card types into separate subdecks with their own limits, but that breaks burying/the sibling relationship as they stop existing in the same note context. On top of that, it feels pretty tedious to manage subdecks for every type long term.
EDIT:
I was wrong. This is was a goof on my end as I didn’t correctly configure the decks’ options. This will definitely achieve what I want. Still, it I hold it to be quite tedious to manage subdecks for every type long term. I do wonder how other options play out with subdecks.
EDIT 2:
I was wrong again. This doesn’t solve the issue of wanting fewer note types introduced per day. In fact, it makes it worse. With subdecks, new notes = new cards is always guaranteed, which means some notes end up being introduced through type2 instead of type1.
What I actually want is: introduce 20 new notes via 20 new type1 cards, and then add anywhere from 0 to 20 type2 cards without introducing additional notes (catch-up only).
The “unsuspend a few notes at a time” approach is closer to what I’m after, and I’ve seen med students do exactly this with uncapped daily limits. But again, the drawback is that it’s manual, error-prone work. You’re constantly babysitting the queue, suspending/unsuspending, or micro-managing card types, when the whole thing could be handled much more cleanly by a built-in New notes/day option.
That’s why I keep circling back to it: the existing workarounds either (1) break the sibling model, (2) require manual micromanagement, or (3) make daily pacing unpredictable. A notes-based cap would solve all of these in a predictable, low-maintenance way.
Don’t get me wrong, I love Anki. It literally changed my life but that’s a topic for another day. I even attempted for 2 days straight to implement this on my own as a fork, but it appears I don’t understand the codebase that well to test if my implementation for the limit reduction in gathering.rs is sound. (If it’d be useful, I’d love to share what I tried so someone more familiar with the codebase could sanity-check it. It at least builds.)
One other angle where I think this would help a lot is in the user experience, especially for people just starting with Anki.
New users often struggle with the difference between notes and cards. The current system exposes them immediately to “cards/day” limits, which makes sense technically, but not conceptually. A New notes/day option would map more directly to what most people intuitively think they’re studying: ideas or facts, not card templates.
It could even open up useful feedback features: if Anki knows you’ve “mastered” all cards from a given note, it could highlight that note as complete. That would make progress feel more tangible: you wouldn’t just be clearing a pile of individual cards, but actually working through and finishing units of knowledge.
In other words: it wouldn’t just solve a scheduling/pacing issue for power users, it would also make the system more intuitive for beginners and more motivating for everyone.
I think the tradeoff boils down to what you want consistency in:
New cards/day (current system): consistency in reps. You always know how many cards you’ll face, but burying means you might get more new notes than expected.
New notes/day (proposed): consistency in concepts. You always know how many distinct ideas you’ll be exposed to, but the number of cards per day will fluctuate as siblings get released.
It doesn’t have to be an “advanced toggle”. Both limits can be set together, with the stricter one applying. That way:
If you only care about reps, you set new cards/day.
If you care about concept pacing, you set new notes/day.
If you care about both, you set both, and Anki respects whichever cap is hit first.
That gives flexibility without breaking the simple model, while also giving power users the pacing control they want.
I love this last suggestion for its simplicity. I have some note types where the number of generated cards depends on the fields. For these note types, the last gathered note is almost always cut off, because not all its cards can be included in the new-card limit. It would thus be great to be able to set a new-note limit.
But a toggle like ‘Always gather siblings together’ could also fix this. (This requires ‘Bury new siblings’ to be turned off.) Then, with a new-card limit of 50, you might sometimes get 48 cards, because the next note in the gathering queue has more than 2 cards.