I’m not sure, I just now realized there is a race page where I can see average days/level, I could take the overall average of that
I tried to see if I could get a decent average or median from the level data I had from making the level distribution graphs, but due to the high number of people either dropping out, resetting or reaching level 60 in the 1 year period between the two data points it’s hard to get any real concrete data. There are surprisingly few people who even level up once every 2 weeks it seems though, so I’d estimate once every 3 or 4 weeks might be closer to the average.
You know what this means Time to gather data more frequently
I could take a snapshot right now and see what the average speed has been in the past 2 months. It takes about half an hour to gather the data though, so it takes a while to fetch everything
I think most people, if we are only looking at users who are consistent with their lessons, land somewhere between 10 days and 20 days. Some people go a lot faster, others go slower.
I’d say aim for the halfway point of about 2 weeks but give yourself the upper limit of 3 weeks when scheduling out a general idea of time.
Lots of people take in levels based on their apprentice amounts, others level based purely on a set amount of lessons per day. See what you find is most sustainable for your learning style.
Thanks everyone for the replies, I think I might make my own data collection when I get home? Or just something as I would like to try the API and have not yet
In theory, you can get roughly 8 levels for 2 months of subscription if you go at one week per level. Most people tend to go at roughly half to a third that pace, and others that have busier schedules may go 1 level per month.
A good way to figure it out is to gauge how many lessons you could do per day. Each level is going to have 120~150 lessons in total. Unless you’re using the reorder script to bypass vocabulary (NOT RECOMMENDED) then you’ll basically have to do all of those before getting to the kanji of the next level.
You’ll be able to gleam what you’ll get per month based on how many lessons per day you can sustain.
So far I am at about 75 reviews a day, and I very impatiently await my later reviews, refreshing my app every hour, I could probably do like 200 easy at once and then start to feel burnt out. I can’t give a number until I try it, thanks for the replay.
Careful, you eat all the reviews and lessons up at the earlier levels but it’s gonna be another story once you reach even the few higher levels after that.
You’ll only know what your real pace is once you complete level 5-6 I’d say, and even then be careful because those enlighten and burn reviews won’t care that you have already 150 apprentices on-going
I ran the script using the new data and the data from halfway through January as a comparison. Taking only people who leveled up at least once into account in order to filter out inactive users, I get an average of 15.5 days per level, and a median of 17 days per level. So once every 2-3 weeks.
A question, can your bot retrieve info about users who don’t have an active subscription anymore?
(just a thought about speed runners perhaps being more likely to have used a yearly sub, and then drop off WK after reaching lv 60)
To be clear, B is gathering data from the forums not the WK API. The WK api only provides information about our own progress
Oh interesting, I didn’t think of that, I thought everyone just had their own gate into the API you just needed a key for it (everyone accesses the same data but you need an account).
The API only gives you access to your own data. For data such as level distributions we tend to look at the forums as an indicator instead. For example, I get most of my data by looking at the users with the Basic badge which you get if you spend 10 minutes on the forum, which should be most forum users. Discourse has an undocumented internal JSON API, so I tend to use that for statistics.
For example, level distribution data and the estimate I made earlier are based on data from this endpoint: https://community.wanikani.com/user_badges.json?badge_id=1. Badge 1 is the basic badge for example.
While this is a reasonable question that many ask, here are a few few thoughts I’ve not seen mentioned above:
Many people progress faster through earlier levels. This is especially true if you already spoke Japanese or could read a few characters before you started with Wanikani.
Wanikani levels aren’t really for measuring progress, they exist to pace out the introduction of new lessons. Every person is different: some can absorb information much more quickly than others.
For the first several levels, up through about level 7, you won’t go longer than two weeks between reviews. Then you will start reviewing items you haven’t seen for a month. The real fun starts after around the six month mark when you start reviewing items you haven’t seen for four months. Most of us find we’ve forgotten many items that we haven’t reviewed for months.
Learning a new language is a multi-year effort (I’d say a lifelong effort). Even if you get to level 60 on Wanikani in a year you’ll still be at the beginning of a much longer journey. Learning to read kanji is an extremely helpful first step, though. If you can do it in a year that’s fantastic, but many of us take two, three, or even more years. It’s reasonable to just commit to a monthly subscription to see if it’s for you, but you may want to consider an annual or lifetime subscription if you get serious (they’ve offered a substantial discount on lifetime subscriptions every year around Christmas time, for what it’s worth, and they pro-rate the price if you’ve already got an existing subscription).
To give you more concrete data, here’s a single datapoint (my own experience thus far). I’m definitely not a speed runner (several people do successfully get to level 60 in a year or so). By my own estimation, I’d say I’m at the mid to slow end of the scale.
In my case, I’d been speaking conversational Japanese for many years prior to joining Wanikani, and already sorta knew perhaps a couple of hundred common characters (“sorta” meaning I may have known the meaning and maybe just one reading). The first 7 or 8 levels were a breeze. As you can see my time per level increased dramatically around level 17, however (see below).
I started in January of 2020. I reached level 4 at the end of January. Two months later, I was at level 10.
At level 17 I started reviewing enlightened items I hadn’t seen for four months. I also started seeing a much higher percentage of completely unfamiliar characters in my lessons. But I made the mistake of continuing to do all my lessons as soon as they were available. The combination of these two things made my daily WK reviews much more difficult (almost overwhelming) and my time per level started to shoot up.
Since about level 30, I became smarter about how many and when I do new lessons. Accordingly, the number of reviews I do each day and the time I spend on each level has been much more constant.
Good luck! I hope you join us.
Wow, that is a very long, detailed post, while I cannot send another post as long to talk about your post, I can add a few pointers of my experience. First of all, thank you for the charts, that helps me out, how did you get the charts? Is there an extension that I am missing?
I started learning Japanese about a year ago and as a result, I know many words in Hiragana but not in Kanji, so my ability to speak them is much better than my read/write ability. I heard about WaniKani and I immediately wanted to hop on the WaniKani train as it looked both fun and informative, and being fun is what learning a language is about, at least for me.
I noticed the sale earlier when WaniKani had emailed me for the Christmas one, but the title of the e-mail was quite literally “WaniKani is On Sale (But You Shouldn’t Buy It),” so as a result, I succumbed to the Crabigator overlords and, well, didn’t buy it.
I am now facing the issue of my app that I use to learn Japanese grammar and Vocab is a subscription, and WaniKani is going to be a subscription in about a week maximum, leading to two of my resources being paid, while my app only has a few more months before I run out of content and eventually turn to something in Anki, WaniKani seems to have at least 2 years to its name, and I am not a fan of paying $20 to learn something that I picked up for fun and maybe something later on like visiting Japan but nothing major such as living there.
I have a few follow-up questions to your post if you are still with me after my long post:
You mentioned that you used to do all reviews as soon as available and then learned that it was not the efficient way of using the app. What do you do now, what is seen as the efficient way? I thought that since I was doing my reviews as soon as I got them and the SRS spreads them to be where they should be, that this was the optimal way.
More of a general question, not related to anything that was said in your post, but what do you use now for learning Japanese? Or have you gotten to the point where it is mostly just Kanji and words you do not know you simply throw in a search engine, and read the Japanese word, with a Japanese definition?
Thanks for the post once again, and if you read this entire thing, thanks for that. I tend to rant on about things so sorry if I did that here, I tried to be concise. I now realize that against my first comment, I made a post as long as yours to respond…
I’ll add another point for caution here. I was doing 20 lessons per day and keeping my apprentice count under 100 and was getting 150-200 reviews per day with spikes up to 300 as I hit the fast levels at 45+.
Think about it like this, once you’ve hit around the 6 month mark, your reviews on a given day are based on:
- lessons you did yesterday
- lessons you did a week ago
- lessons you did 2 weeks ago
- lessons you did a month ago
- lessons you did 4 months ago
It adds up.
I now realize after this post, I used the completely wrong word here, I meant reviews, because on average I do around 75 reviews a day, 45 at one point, and then the rest are scattered, it isn’t due to my lack of work ethic, but the lack of reviews due as I am level 3.
Yea, in the early levels reviews will be fewer due to the SRS timings. However, as pointed out by others, it takes roughly 6 months for a schedule to stabilize.
I did Wanikani at basically full tilt, clearing out all lessons on the day they became available. By the time my reviews stabilized, 6 months later, my weekly review count was somewhere around 1200. If spread out properly that’s fairly manageable, free time providing, at roughly 171 reviews a day.
However, there is what’s known as the “fast levels” that start at level 43. These are where burnout can happen if you attempt to do something insane like clear out all lessons as they become available. My schedule by 60 had stabilized at roughly 2200 reviews per week. This makes the fast levels a bit of a trap for those that try and clear out lessons immediately; those that do a set amount of lessons per day won’t really run into any problems in the fast levels as the rate of new apprentice items won’t change for them.
That’s my middle name. The abbreviated version, anyway.
You may also be interested in one or two of the scripts I wrote if you continue:
As you noted later:
You should do your reviews every day (multiple times per day if possible). You should try to get the number of available reviews down to zero at least once per day.
You do not need to do lessons every day. Lessons determine how many reviews you’ll have in coming days. If you do 20 lessons, you’ll have 20 more reviews in 4 hours. Answer those correctly and you’ll have the same 20 added again in 8 hours, then 1 day, etc. It’s wise once you get to later levels to space out your lessons and not do too many at once.
I’ve immediate family in Japan so I tend to get plenty of conversation practice. I’ve recently subscribed to Satori reader for more reading practice (it’s excellent, but a paid service I’m afraid). Bunpro is great for practicing grammar (also paid, and I’ve not been using it as much as I should) but the Cure Dolly series on Youtube is also excellent (once you get used to the creepy uncanny valley thing — subtitles and speed settings help).