I've done some thinking and experimenting with review intervals, as I developed and use
Flash! Pro vocabulary memorization software.
In theory, the most effective review schedule is one in which you review a word just before you are about to forget it. But how do you estimate when that will be?
In practice, the most effective review schedule I've found is to review those flash cards that I have not got right for a period of time longer than the time between the last time right and the last time wrong. So to answer your question, that would be the 1, 2, 4, 8, 16, 32 pattern, assuming you get it right each time. Get it wrong, and you return back to 1.
In Flash! Pro, this drill type is called "Elapsed Time," and its search expression looks like this:
((Now-DateLastRight)>(DateLastRight-DateLastWrong))
Again, this is the search expression that selects words for which the time since you last got it right
exceeds the time between when you last got it wrong and the time you last
got it right. To illustrate:
Say I drill a bunch of cards including "logos" for the first time at 12:00.
I get it wrong.
After the drilling session is over, and I get returned to the select tab,
"logos" immediately gets selected again.
I drill the bunch again (it is now 12:05). I get it right this time. It has
been 5 minutes between wrong and right.
Back at the select tab at 12:08, "logos" does not get selected. It has only
been 3 minutes since right; less than the 5 between wrong and right. I drill
some other cards.
Back to the select tab at 12:11. "logos" gets selected because it has now
been 6 minutes since I got it right; more than the 5 between wrong and
right. I drill it and some other cards; I get it right again. Now it is 11
minutes between wrong and right.
Back to the select tab at 12:14. It is not selected. 12:17: not. 12:20: not.
12:23: yes; it has been 12 minutes since right; more than the 11 between
right and wrong. I get it right again. I stop drilling for the day.
The next day, "logos" get selected. I get it right. It won't appear again
until tomorrow.
If I get it wrong even once, the process starts over as if I never knew it.
As it should.
I find this formula brings up the card just as I am on the verge of
forgetting it, for the most efficient use of memorization time.