NeuraCache offers two static algorithms and two adaptive algorithms at the moment. You can change between them any time, either for a single note or for a group of notes.
Static algorithms resurface your notes on a fixed schedule. The retention score does not have any effect on the next review date.
- 1, 2, 3, 4, 5 days
The review will happen on the next day, 2nd, 3rd, 4th and 5th from the start date (5 consecutive days) This pattern is best used for Cramming (Education)
Usage example: You have a known deadline (like exam next week). You would like to review your set of notes every day for the next five days, no matter how well you remember each note.
- 1, 5, 15, 30, 60 days
This pattern is a simple approximation of an Ebinghaus forgetting curve. It has larger gaps between reviews, even one month in the end. Therefore it is not suitable for cramming type of learning, but it works better for not urgent knowledge retention that does not have any deadline.
Usage example: Reading an article online which has a fascinating insight that you would like to retain for much longer. You could capture/start spaced repetition using Evernote WebClipper as in this example
Adaptive algorithms resurface your notes based on the latest retention score.
- Adaptive (SM2 based)
This algorithm is an adaptation of a popular SuperMemo2 algorithm. Detailed information and formulas of original SM2 can be found here SuperMemo.com.
In most of the cases, this is the best choice for all types of notes. The algorithm chooses a new review date based not only your last review score but also takes the history of previous reviews into consideration. It is optimized for storing learnings in your memory forever.
Usage example: Things you have learned or understand enough to express as a set of atomic Flashcards.
- Adaptive Simple (exponential)
This is a very basic algorithm that takes under consideration only the latest review score. It does not include the history of the reviews to calculate the next review date. Based on the score 0% - 100% it calculates the next review date from an exponential function in a range of "today - 6 months". So if your review score is 100% the note will be resurfaced in 6 months, if your review is ~80% the next review will be ~3months, 0% - same day, etc.
Usage example: Book insights, lectures, article insights, thoughts, quotes, etc.
Whatever you would like to put into your long term memory with lowered effectiveness (when compared to SM2) but without feeling overwhelmed with the number of reviews.
Which algorithm to select:
Although SM2 is the best default choice oriented for the results, it is worth experimenting with other algorithms to see how they work for you.
A typical situation is where you feel like there are too many reviews in Today's Queue. If this happens, try adjusting the algorithm to "Adaptive Simple" or "1, 5, 15, 30, 60" for some notes so that they come back for review much later in the future.
You can also use Today's Queue Limit feature.