How is the 80% calculated?



  • One of our daily goals is to recall words at an 80% accuracy.

    What is this based on? After I reach the 80% goal I could then get the rest of the words wrong so it seems to me that it's not 80% of the 100 cards for that day. It seems like it's based on correctly recalling the first few words we see?

    Could someone clarify this please?



  • @silkandspice @Levi-0

    When I practice e.g 213 cards / day and I once reached 80% for the first 8/10 cards (which I better knew) I can go to the Progress / Today tab:

    The previously reached number 80% is corrected down to 50% "correct repeats".

    So the Lingvist algorithm seems to re-calculate the total percentage per day and daily card deck, which is good.

    When you learn a new language from scratch, how do you ever want to reach the 80% daily goal for all 100-200++ repetition and new cards, where NEW cards are also repeated multiple times in a learning session or a day?

    This will work better for easy words, but not for more difficult ones.


    My experience:

    When I do not know a NEW word well, it does not matter how many repeats I will encounter, either 1, 3, 6 or more.
    I will constantly do this more difficult word wrong.

    And Lingvist will show the same sentence over and over again, without showing me a different context / sentence for the same word.

    I have the same problem on Memrise for a few NEW word (learning session) or OLD words (review session).

    And it really get's on my nerves (both on Memrise and Lingvist) that I can not SKIP (exclude) those "harder words" for the moment / current session after a few tries and better focus on other words to review or focus on bulk learning other new words.

    I definitely want to see quick progress with the total of 3000-5000, but not getting exhaused with learning a few words, which I may first have to learn offline from Lingvist (e.g grammar book).

    Memrise in comparision shows me the statistics for a defined review session:

    • My review words are set to 100, then the review session automatically ends.
    • And all the statistic numbers are shown for the current (defined) session (100 words).
    • If I get exchausted reviewing 100 words, I can also terminate the session earlier (e.g 30, 70 words) and then the statistic numbers are calculated for those words reviewed.

    I am not sure if I can adopt to the "Lingvist concept" WITHOUT any dedicated / defined review sessions (incl. progress bars), and that the card repeat / practice deck is constantly filled on the fly (either repeats of NEW words and incorrect answers, or spaced repetition words learned the same day (e.g 4/12h) or previous day(s)).



  • Then the 80% is misleading. I wonder what the logic is of only asking for 80% accuracy on only 10 words. Shouldn't it say 10% of any given day's cards?



  • 8 out of the first 10 words for that day. From the 11th word on it stops tracking until the next day.


 

Recent topics: