Background
Some background first. Currently, filtered decks allow a preview mode (turn reschedule off) where the 4 Anki buttons show a delay based on user-defined times. So, you can select Good on a card and it’ll come back again in say, 40s or say, 20s. As we can see, the delays for each button is based on time elapsed which is how buttons work across Anki.
Suggestion
I suggest an alternative to time-based delays that instead work on the basis of review/card counts. Here is an illustration for “count-based delays” (but it uses only two-buttons):
One advantage of this is that I never run out of cards to do. So, for example if I’m trying to run an infinite loop with 30 multiplication cards, I can set my Good button to 30 (delays can be capped at total cards in deck). It’s much more neat this way than setting time based delays + learn ahead.
Another advantage is that if I’m rushing through a section of my cards, count based delays can compensate for the speed by showing me the cards earlier than I would have with time-based delays so that I shouldn’t be forced to hold 40 cards in my short-term memory.
Another advantage might be this:
Suggestion 2
To add some randomness, Anki can also allow setting a range. So, just like “show me this card after 20 more cards” we can also do “show me this after 15-30 cards”.
Well, this one can be done even for the current delays.
Note: this is a different suggestion so hope the mods won’t try to merge it with expertium’s card type (note type actually) suggestion.