Original analyses of Japanese, computed from Kumi's dictionary and knowledge graph. Every number regenerates from live data, so a report cannot drift from what it describes.
Kanji readings
Why the easiest kanji have the most readings
The 500 most common kanji average more distinct readings than the rarer ones. The characters you meet first are the hardest to read, and only a handful of them have a single reading.
3 min read
Pitch accent
Nearly half of Japanese words have no pitch drop
Pitch accent measured across 81,000 words. Heiban, the pattern where the pitch never drops, accounts for close to half of them, and a small share of words have more than one accepted accent.
3 min read
Radicals
Twenty-five radicals appear in nine out of ten kanji
The radicals behind the 2,501 most common kanji are wildly unequal. A small group appears in the overwhelming majority of characters, while the median radical appears in far fewer than the average suggests.
3 min read
Learner behaviour
Multiple choice feels like progress because it is easy
Built on learner behaviour rather than dictionary structure. Picking a word from options is both more accurate and faster than producing it from memory, which is exactly why a study routine drifts toward it.
3 min read
Looking for how to actually study something? The guides cover getting started, how mastery works, and what makes Kumi different.