Why testing beats rereading
The single most robust finding in the study of learning is the testing effect: retrieving information from memory strengthens it far more than reviewing the same material. Rereading notes feels productive because the material becomes familiar, but familiarity is not recall, and it produces confidence without capability.
Flashcards work because they force retrieval. The effort of trying to remember — including when you fail — is what consolidates the memory, which is why struggling briefly before revealing the answer is more valuable than flipping quickly.
This is also why passive review before an exam is a poor use of time compared with self-testing. The discomfort of testing is part of the mechanism rather than a sign it is going badly.
Spacing and the forgetting curve
Hermann Ebbinghaus demonstrated in the 1880s that memory decays sharply at first and then more slowly, and that each successful review flattens the curve — so intervals between reviews can lengthen as material becomes established.
Spaced repetition systems automate this by scheduling each card individually based on how well you recalled it, showing difficult material frequently and well-known material rarely. The efficiency gain is substantial: reviewing a card the day before you would have forgotten it is far more effective than reviewing it while it is still fresh.
The practical requirement is consistency. Twenty minutes daily beats three hours weekly by a wide margin, because the scheduling depends on reviews happening near their due date. A long gap produces a backlog and material that has decayed past the point the schedule assumed.
Writing cards that work
The most important principle is minimum information: one fact per card. A card asking "list the seven causes of X" cannot be graded honestly — you will recall four and have no clear answer about whether that counts. Seven separate cards each test one thing and schedule independently.
Avoid cards answerable by recognition rather than recall, and beware interference: cards that are too similar get confused with each other, which is why learning two related items simultaneously often takes longer than learning them separately.
Cards should be in your own words, which is why making your own is more effective than downloading a deck — the act of formulating a card is itself a learning step. Add context and mnemonics, use images where the material is visual, and for language learning prefer whole sentences to isolated words, since vocabulary in context is retained better and teaches usage alongside meaning. Cards stored here stay in your browser's local storage, so they are private to this device and lost if you clear browsing data.