Demote the incorrect answered term as well

Suppose you’re a moron like me, and you can’t tell the difference between, 半、来、平. If prompted with next/come , and answering rice or flat, then shouldn’t rice or flat also be demoted?


How would the system even do that? Imagine you typed 評 as “judge” when it means evaluate, should it demote 判、審、裁、every other kanji that remotely means judge? How is it supposed to know what you mistook it for?

Suppose it uses the visually similar Kanji list. Which looks like for flat would be half.

It’s an interesting idea, but in some cases there could be more than one term matching your incorrect answer. How would WaniKani know which one to demote? Not to mention if you do this for reading where there will often be many kanji with the same reading.


So you want wanikani to mark kanji with matching radicals wrong when you get a different kanji wrong? I don’t get what purpose that is meant to serve. The next time that kanji comes up, if getting the first wrong doesn’t help you differentiate, you’ll get the other wrong as well anyway. If you’re having trouble distinguishing visually similar kanji, that’s what radicals and mnemonics are there for. No need to be thoroughly demoralised by getting 6 kanji marked wrong when you mess one up.

Edit: stupid autocorrect

Edit2: not to mention that what is “visually similar” or gets mistaken between two similar kanji will differ from person to person.

1 Like

Or you could answer the ‘rice’ wrong on purpose if you feel like it. I certainly did that when I mixed the two up


I like for WaniKani to take the choice away from me as that seems to be the accepted trend.

1 Like

Sure, but they have a visually similar list, why not utilize that as a minimum.

This makes no sense. Just because you confuse say 来 for 平 doesn’t mean you’d confuse 平 for 来。 I once wrote “fault” for 矢, confusing it for 失、but I wouldn’t confuse them the other way around because 失 is much more common. Getting the one wrong can make you reconsider the differences between both in your head, thus strengthen the differentiation.

You already get faulted for mixing them up by getting the kanji you confused demoted in the first place, having the system guess for you which one you mixed it up with so it can notch that down as well is counterproductive.


I have found myself putting in the same meaning for three different kanjis, only getting the last one right. I can see your point…

If that’s the case you need to actually take some time after the mistake to go over the mnemonics and the radicals in each, or create a better suited mnemonic that you can remember. SRS isn’t going to help if you get it wrong, look at the right answer, input that in 5 seconds time with no further thought, and then wait however many hours to get it wrong again, and getting multiple answers marked as wrong simultaneously won’t fix that.


“Visually similar” is hardly an exact science. In fact, it’s very subjective.


There’s “helpful difficulty” and then there’s “counterproductive punishment”. Demoting other characters like you suggest would make for a learning experience where you’re too apprehensive about making mistakes.

Say you mistake “rice” for “to come”. It’s the red bar of shame! But in that instant, your brain registers a minor “bad” emotion which helps your memory to mark the significance of the mistake and increases the chances of remembering of what happens next:

  • 米 doesn’t mean the thing you thought, which hopefully leads to you to remember what it does mean (rice) - bonus, you just remembered something and strengthened a connection in your brain!
  • ok so “to come” isn’t 米 in japanese, but it’s close… and if you can’t remember it there and then, you’ll probably make an unconscious effort to remember it better when you see it next so you don’t feel that bad feeling again

If WK started attaching disproportionate consequences to getting a single word wrong, well… I’d personally find that really discouraging. Just because you’re getting stuff wrong doesn’t mean you have to be punished. I find I get really steamed after messing up a word repeatedly motivates me to remember it all the more.


That’s good advice, thank you.

I think you would find that progress would slow immensely if WaniKani implemented something like this. Like, there are a LOT of visually-similar kanji and kanji with similar/identical meanings. It would be a lot of delayed level-ups (frustrating) and items stuck in Guru or Apprentice (overwhelming).

1 Like

I’ve often thought about a dream script like this. It would make “buddies” out of some of the kanji and vocab you’re mixing up and only let them climb the ranks together. So like, you’d link two of them as buddies, it would knock them both down to the same SRS level to start, and you could only promote them if you get all four questions correct. Get one wrong and they all demote. Just to force yourself to learn them better. And you’d only be able to link kanji with kanji, vocab with vocab. Alas, I can’t code, so it’s just a dream of mine as I mix up 例 and 列…again…for the millionth time…

I try to be honest and intentionally fail a review if I know I mixed something up with it earlier, because it means I don’t know this item or the other item well enough. However, there are times when I get the answers right on an item first, only for its confusion buddy to show up later…and me to mess the second one up anyway. Love it when that happens :expressionless:

Flat is the one that’s flat on top.


I like this idea

Well, there has been more than once that I mistake a couple of similar kanji and then when the second one arrives i guess it correctly because i failed the first one

So yeah, it’s nice, or you could fail it on purpose to help you remember

1 Like

I think this could work out with proper analysis as to which sets of Kanji people are actually confusing with each other, whether or not it would be visual because, as someone mentioned, it is rather subjective. This is pretty helpful when a kanji you’d have probably set up to a higher level or whatnot (and would probably not be coming up any time soon) would be followed up by a fairly similar Kanji (again, subjective) which would cause you to get confused of both even if one was already in a higher level. What could be done is the gathering of data and perhaps a pop up could ensue to ask with what you confused the kanji with (with the options of the higher tier Kanji you have)