Write a question on one side and the answer on the other. Answer before you turn it over. A card you know moves one compartment back, a card you miss returns to the front, and the gap is set by how long you want to keep the thing. No streak, no score, and no total of cards as a measure of anything.
Size the gap to the goal
Decide first how long you want to keep the thing. Each number below is the wait, in days, before a card in that compartment comes back.
From the season: for a test a week away, come back after a day or two. For a year from now, the best gap in a 2008 study was about two and a half to five weeks, and a gap that is too long costs much less than one that is too short. Growing, even and shrinking ladders showed no reliable difference, so any sensible shape does the work.
Today
Answer it first, aloud or in your head
Add a card
One big stack works: in a 2009 study one large stack beat four small ones for nine people in ten. Making cards costs time that nobody has measured, so weigh it against what the cards are for.
Where the cards are
Every card
Cards hold facts. How many are in the box says nothing about understanding.
The record
What the record says. From the last episode of Study 1: The Cards: “This is what the record supports for the cards. The spacing effect is as good as the trials say, which is very good. What got sold on top of it was the schedule, and the schedule is the part with the least behind it. What is left is small and old: a box, a gap and a test, which a schoolboy in 1885 already knew to do with his vocabulary in the morning.”