Purely a thought experiment here not something I’m currently doing, but I’m curious if anyone has stress-tested this or can explain exactly how it would fail (or maybe work?).
The hypothetical setup:
-
Set FSRS desired retention to something absurdly high, say 0.95 - 0.99.
-
Set Anki’s built-in “Maximum reviews/day” to a hard cap of 250.
-
At the end of every single day, use the FSRS Helper add-on’s “Flatten future due cards” feature to spread the remaining overdue cards so that tomorrows (and every other days) load is also exactly 250.
Heres basically the logic I’m working w/:
-
The absurdly high DR tells FSRS I cannot forget anything making it to assign very short intervals and generate an enormous true daily due count, unfeasible normally.
-
The 250 review cap means you physically only do 250 cards and the rest are left overdue.
-
The daily flattening redistributes the overdue cards, preferentially pushing out high-stability cards to protect the most vulnerable ones.
-
In theory, the combination could create a predictable, daily workload of exactly 250 reviews, with the high retention setting acting as a “buffer” against the inevitable forgetting caused by the forced delays.