Hi guys!
Although I’m not new to Japanese, I am fairly new on Wanikani. I’ve been lurking on this forum for a while now and I have noticed many users saying they prefer to take their time to complete a level. Do you think I am going to do fast based on these stats? How long is it supposed to take to complete a level?
It doesn’t really matter how long you take to complete a level, as long as you are feeling good about your progress and you’re not overwhelmed with all the reviews, then you’re good!
But if you start getting overwhelmed, reduce the number of lessons you do per day, you don’t wanna get burned.
Also, the first levels go faster because the time between reviews is shorter, now they’ll start getting longer, making it so that you can pass level in 7 days instead of 4, that is, if you do all your lessons quickly.
Welcome! The levels can be done at different speeds, have a read of jprspereira’s awesome guide to get your head round how they work.
For an example of what going fast looks like, check out the post from @neolaron who just hit level 60 today- congrats! They went fairly close to the theoretical maximum.
Main lesson though- there isn’t a universal ‘right speed’ to do wanikani- go with what lines up with your broader learning goals, keeps you motivated, keeps you happy and doesn’t burn you out. That might be different now than it is in 20 levels time, and that’s great too. Good luck!
I usually spend around 10 days on a level. After you get past the initial 8 levels or so, as KiwiSmint mentioned, you’ll be able to get through a level in about a week if you do your reviews right as they come and make no mistakes.
But of course, that’s not a reliable way to learn. You’ll make mistakes, you’ll miss out on a day of reviews occasionally. My advice to you is to not worry about timing your reviews and instead just do them whenever you have time.
As others have said, it all depends on how fast you’re personally able to go! Some people have a lot more free time, or have a major reason why they need to learn kanji fast, and other people have very busy schedules or are trying to pace themselves to avoid burnout. Personally, I’m going at about two weeks a level, roughly.
The important thing to realize is that the pace can really pick up very quickly once you get several levels in and have hundreds or thousands of reviews in circulation, so my recommendation would be to think about your future self when you choose to do lessons. After you do a lesson, you’ll see that item in your reviews 4 hours later, then 8 hours later (assuming you get the review correct), then a day later, then two days, then a week, then two weeks, then a month, then four months… etc. So if you binge 50 lessons at once, that same clump of items will all come back at the same time for each of those intervals. In the early levels, this doesn’t really matter, but if you already have a hundred or two hundred reviews to do that day, those extra 50 reviews on top of that can add a lot of stress!
jprspereira’s guide goes into a lot of detail about tips for pacing yourself, and I second his recommendation of doing a set number of lessons a day (instead of binging all of the lessons available to you at once). It’s a good way to establish a consistent level up pace and a very even, predictable workload, which makes it easier to plan the rest of your life and studies around WK.
Generally, you want to be able to complete all of your reviews every single day and don’t let them pile up. It’s best if you can find time to do your reviews three times a day to take advantage of the SRS intervals, which will move items along faster, plus increase your accuracy on them (higher accuracy = less reviews = less work!).
If things start to pile up, don’t panic! Just stop doing lessons and focus on getting the review numbers down. Some people recommend keeping your number of apprentice stage items 100 or less (so, stop doing lessons until the number drops below 100). Personally, even that is too much for me, so mine often hover in the 60’s and 70’s. Setting limits like that can help set a good pace for you and keep you from burning out.
The other important thing to consider is that you need to leave time for studying other aspects of Japanese besides kanji. WK can’t be the only Japanese studying you’re doing, basically. Going full speed is pretty time-intensive, so I would not recommend it for the majority of people (unless you have a very specific reason to go fast), because most people find it hard to also fit in grammar study and immersion if they’re spending a few hours just on WK every day.
I did it on average in 10-14 days. But this level is already taking me 51 days.
Depession hit hard, and I only manage 2 new kanji and some vocab. I drastically lowered my apprentice items, and my new lessons.
It has to be unde 100 apprentice items, before I do vocab.
I do hope the next level can be a bit faster
The average will depend on a lot. What your accuracy is (will fall over time), how quickly you do reviews (do you stay up late or wake up in the night? I don’t recommend it), how many reviews and lessons you do each time and so on.
It really doesn’t matter, three things matter: 1. Make sure you enjoy the pace, you don’t want to burn out of Japanese by the time you can read it 2. Don’t cheat, it will lower your average, but literally noone would care and you give up on actually knowing things 3. Finish, doesn’t matter when, it can be in less than a year or more than 5, that’s ok, because you did
I don’t know. I don’t think there are stats published for average, only fast ones. I suspect most users might be around 2-3 years finished 60 levels, translated into 10 or more days per level. But, again, I am not sure about that number.
In short - yes you’re doing marvellously - keep it up. But don’t beat yourself up if you ever need to slow down a bit in future. Because that’s still progress, and better than falling by the wayside. Most people fall by the wayside. My advice is not to worry about your stats and comparing yourself to others, but to go at a pace that you find motivating but not stressful.
I’ve slowed down a lot lately (levels about 15-20 days - this is not ‘preference’ but necessity) because lots of old stuff is coming back round again after a long time - and I don’t remember half of it, so it drops down to ‘apprentice’ and takes up brain room that I’d like to learn new things with. I expect my brain is older and slower and more full of junk than yours so that might be in your favour! Also as you already know some Japanese I’m sure that will help, as you will recognise some words and get them right quicker (I knew zero on arrival).
here endeth the lesson good luck
Welcome aboard!
One thing not mentioned above:
Those 99% accuracy stats tell me you probably already knew most or all of the characters in the levels thus far. It’s normal to proceed very fast in the early levels, but unless you already have a really large vocabulary and know a LOT of characters, this will change eventually.
I also flew through the early levels, but started taking more time once I started seeing more characters that I didn’t already know (right around level 17 or 18):
Speed-runners never slow down on lessons, but that gets extremely difficult quite quickly.
Most of us mere mortals are more focused on long-term retention, and find it painful to have more than five to ten absolutely unknown characters in the Apprentice 1 or Apprentice 2 stages.
In my case, I almost never add more than 5 new characters in the early days of a level, no matter how large my lesson queue grows. This is because I’m now at the point that the majority of the characters introduced are brand new to me.
I always try to get my review queue down to zero every single day, though, and I always do my reviews before doing my lessons (if my reviews were too difficult, I might not do any lessons at all).
It feels pretty effortless at this pace, which is nice. I actually look forward to doing my reviews every morning with a cup of coffee. My accuracy stats are right around 90% overall, which feels about right to me (still easy, but difficult enough to be interesting and make quick progress).
To each their own, but I can’t relate to speed-runners. They need to maintain very high accuracy rates while still doing every kanji lesson as soon as they become available. It doesn’t seem like fun at all to me, as it requires a great deal of mental effort (and I suspect various “cheats”) but it can’t possibly be optimal for long-term retention to go so fast.
I think most of the non-speed-oriented learners like me feel most comfortable maintaining around a 90% accuracy rate. If it gets much lower than 85% it starts to feel like a grind and you start looking for excuses to avoid your reviews. If your accuracy is much higher than 90% it usually means you can safely do more lessons faster.
“Burning” most of the 6,000+ items in two or three years seems like it would be an amazing accomplishment to me (and quick!). It also seems like a more practical, useful goal than merely “guru-ing” 2,000 or so kanji (getting to level 60) as fast as possible. I’ve already burned around 3,000 items in a bit more than a year and a half. I’m quite pleased with my progress, and happy with this pace.
Good advice Rrwrex!! (I dream of 85% though. That’s a rare occurrence! But I grind on and finish my reviews every day. It’s okay. It’s become a way of life )
I started and forgot of wanikani on the second day when in the first level (my memory is that good)
I usually have trouble when learning kanji (I started wanikani only knowing around 20 kanji) and when learning vocabulary that the convination of the kanjis do not make sense to me, or the reading is an exception (for example 相手), if not, is only just remembering if the reading is kun or on.
To be honest, I see accuracy stats in three different places and they never seem to quite agree with each other.
The most important to me is the percentage in the upper right of each review page, next to the thumbs-up icon:
I usually start to get slightly irritated if this drops much below 85%. Below 80% I start to get pretty frustrated. Most days I hover between 85% to 92%.
Then there is the “review summary” after I finish the session of however many reviews. This number is often quite different for reasons I don’t quite understand:
I think the difference might be that the running total in the upper right just considers each answer individually, ignoring that two questions (reading/meaning) might be for the same item, while the review summary counts only complete pairs. Usually, the summary number is worse than the running total, but the very short session I did above showed the opposite (likely because I quit before I finished all the pairs).
Lastly, there is what wkstats reports. I think this is basically a running/cumulative average of what the “review summary” reports. I know it changes quite slowly, and I’m currently hovering right around 90%:
If I understand correctly, this means that since I’ve started I’ve got 89.89% of review items correct for both reading and meaning. This can be broken down further:
-
91.27% of all reading questions were correct (89.52% of the kanji readings and 91.87% of the vocabulary readings).
-
88.63% of all meaning questions were answered correctly (89.76% of radicals, 89.27% of kanji, and 88.32% of vocabulary)
I think the running “thumbs-up” percentage on each review page correlates best to how I’m feeling during an individual review session.
Above 90%: I’m a golden god! This kanji stuff is easy!
85% to 90%: 微妙. Ho hum. Another morning coffee.
80% to 85%: Man, today’s kinda tough.
Below 80%: This is brutal! I’m never gonna learn this stuff. Japanese is just ridiculously difficult!
This breakdown tends to correlate to how long I’ve been on a level. Once I start “seeing pink” on a level my accuracy stats start to suffer. After I’ve been on the level for a week or so, it picks up.
These are my stats. I try to be at 0/0 by the end of each day and I’m aiming for level 60 at a bit over a year. I have a lot of free time though so I don’t think this is anywhere near the average
Levels 1-3 are accelerated so you have to exclude them to get a better idea of what your pace is for normal levels.
I think an average speed for people who finish WK is probably around 11-15 days per level.
Well I’m impressed. I’m only on level 4 but I’m looking at 10 days per level. But I only can get at it in the early morning and the evening, so it kind of stacks up, and I’m fine with doing 70-80% correct on the reviews. I’m only competing with my brain from yesterday, not anyone else.
I wanted to make quick progress so it was between 6-8 days for me. I also did the so called fast levels.
The more levels you speed through, the heavier the workload will get. Just be prepared to get anything from +300 to +500 reviews later on should you keep that pace.
During review the accuracy is number of correct answers / number of shown questions, so if you have 3 items, you mess up one twice, and the rest you know immediately, that’s 3 / 5 = 60%
At the end it’s number of items you knew immediately / number of items in the session, so the previous example would be 2 / 3 = 66%
Thanks for the reply, but I must be slow. I’m still a little confused.
I think you’re saying that the latter doesn’t count repeats for the same question (meaning or reading), while the former does.
My brief session above though, I missed at least one question when I took the screenshot (89% correct answers, with two correct for both reading and meaning of the same kanji is how I interpret that first screenshot).
The next screenshot with the review summary says I got 100% correct, though, with 2/2 kanji and 4/4 vocabulary. It doesn’t show any mistakes whatsoever? Maybe because I only saw a meaning question and not the reading before quitting (or vice-versa)? [Yes: because I quit the session without completing all “items.”]
I think the running tally is correct-answers/#-of-questions (either meaning or reading) and the session review number is both-correct/#-of-complete-questions where “complete question” is both the reading and meaning for an item. [This is correct, but the session review only counts it as a correct answer if both are answered correctly on the first try. The running-tally counts all answers, even it if takes multiple attempts to get it correct.]
I’m not completely sure, but I also think the “check” number in the running tally only counts kanji, too, not vocabulary (or radicals?). [Wrong.]
I’m sure this is all documented somewhere, but I don’t know where.
The code itself counts it like this:
Whenever you answer a question, doesn’t matter, wether incorrectly or correctly, it counts that as “questionsShown”, and whenever you give an incorrect answer, it counts that as “incorrectAnswers”, and then it displays 1 - (incorrectAnswers / questionsShown)
Where it showed you 100%,that’s probably because that session was actually 100% maybe? Maybe you were lucky and quit it when you didn’t have any incorrect answers saved up?