Feature idea: Sort lessons by new word percentage

I’m currently starting out learning Japanese, and I find that the difficulty of the beginner lessons varies a lot: anywhere from 10% to 60% new words per lesson. Because of this, I have to do a lot of browsing to find lessons that count as comprehensible input (as opposed to incomprehensible input).

I’d love to be able to sort lessons by new word percentage so that I could find appropriate lessons more easily. If this is technically unfeasible (e.g. too heavy a database query for real time search), an occasionally updated list of low-percentage lessons could also work.


There used to be a way to filter lessons based on a set of ranges of percentages (e.g. 20% – 40%, or something like that), but they removed it a few years ago. I guess it just required too many database queries. To be honest, I would be quite curious to know how a search query of this sort can even be implemented.


I just noticed that this is still possible via the “Classic” section (next to the community link). Just go to Classic → Library and click on Show More Filters → New Words Percent.

That’s really interesting! Sounds like the query was indeed too heavy to handle. With tens of thousands of words and hundreds of lessons, that’s a lot of tabulation to do.

I’m kind of impressed LingQ can display those percentages when browsing - I wonder if they cache them for a few days or if they’re recalculated after each lesson.

1 Like

Also the way the new word % works now is different than it used to be a couple months ago. They currently calculate new words using unique words only. It used to be done with total word count. I can understand doing it either way but I actually prefer the old method because it was in general more consistent. The way it’s done now, proper nouns can easily skew the % quite a bit.

Frankly, I don’t really care about the % of unique words because it doesn’t neceassarily matter if an unknown word is surrounded by lots of common words (high usage known words), the meaning is still inferable from the sentence.