Failure rate - rather than time not practised - shd determine word strength
If I always get the "Cat eats bread" sentence right, it should decay in strength much more slowly than sentences/words that I get right less often... Ie there should be a better algorithm for determining word strength, like in Memrise, for example. Would let us focus on practising words where practice can make the greatest effect. Ta.
The algorithm does take into account when you get something incorrect. Any sentences you get wrong and the words you hover over most often will be found in your practice sessions more frequently. You will not see these incorrect words or sentences immediately, but they do surface after a short period of time. We find that this helps in the learning.
I wouldn't mind a combination of the new features with the old hearts. I found that I would work much harder when completing a lesson if I didn't get a full 3 heart rating because I would redo it over and over until I got it right. Now I can just complete a lesson with 1 heart left and it has the same effect as if I had gotten 3 hearts. I like the idea of being reminded to look back and practice a section, but I would also like to be able to track individual progress per lesson. Perhaps we could still track the number of hearts we received per lesson, but then the new overall strength rating for the section?