Blue cards

this is a high-level explanation (some details are omitted — not essential here).

• while user does practice with a story, she can click on unfamiliar words ⇒ that shows the user the translation of the word.
• any translated word automatically goes into the user's deck of flashcards ⇒ becomes a "RED" flashcard.
• when the user later practices with flashcards, if she knows some words, they become "GREEN".
• when the user finishes the story, the system takes the words that were NOT clicked —
  • that can be quite a large number, e.g., 100-200.
  • these are "BLUE" words.
• we want to ASSUME that the user knows all the blue words (because otherwise she would have clicked them)
• but we don't want to assume this without checking a little: so we TEST 5 of these 200 BLUE words — with blue flashcards.
  • (these 5 words are chosen in a particular way from the 200 — we choose the "hardest" ones).
  • if the user in fact KNOWS these 5 words, she gets "credit" for all blue 200 words —
    • they all become GREEN.
  • if the user does not know all the TEST words:
    • those she knows ⇒ become GREEN
    • those she does not know ⇒ become RED
    • she can get another 5 BLUE cards — to repeat the BLUE test.
• whenever she gets all 5 BLUE flashcards correct ⇒ all BLUE words in the story become GREEN