When you get an answer wrong, you shouldn’t have to wait months to try again. let me explain:
I have been using Wanikani for 5 years now on and off, and I have to say that the reason I stopped for 2 years was because getting a review answer wrong is extremely punishing.
I would learn a word, go on a streak of getting it correct over and over, then 2 months later get flooded with 600 reviews on one day and inevitably get a couple wrong. the words I got wrong I did not get better at, and even though I am at level 15 and I still have level 5 and 6 words floating around in my review pool.
It’s a vicious cycle: I learn a word, I frequently get it correct, then months later forget it and it gets de-levelled back to apprentice, and the cycle starts again. I understand this means that I need more practice with the word, but that also means the current practice is insufficient. Now, when a word I dont remember appears in my review, I would rather cheat and study it again after rather than get it wrong and wait a long time to try again.
Here’s my proposal: Wait times for reviews should only occur once.
If you get a word wrong, you can get it de-leveled, that’s fine. but you shouldn’t have to WAIT AGAIN to level it up. a word is in master tier and I get it wrong? give me rapid lessons to re-learn it before I can level it up.
Wanikani is meant to be a self-study tool, and in a self-study tool you should not be incentivized to cheat. I believe that changing this mechanic would be better for users, returning the onus of practice to us rather than when the program deems you need it.
anyone else have thoughts on this?
PS: I understand that there is a “lets study the ones you got wrong” section, but it doesn’t stick when you forget the word and don’t have to try again for another 2 months.
Well this is an unsual request… more often than not people want less not more reviews
But I know what you mean, Master and Enlightened items give me the most trouble too.
If you are open to using scripts, Self-study Quiz can help: start with the “leech training” template and adjust the formula as you see fit (I currently use a ratio of 2).
There was a partial solution back then – Summary Page. I think this can be remade with a script for now. You can see there what you have got wrong, and self-study by yourself.
Self Study Quiz is another solution for now, though not sure of exact configuration. I know leech formula is more complicated than that.
imo, quizzing on wrong items until you get right, once or twice, is a good idea, rather than waiting for longer than a month. But baking into the default reviewer as a solution is kind of opinionated.
This is the script for the summary page.
But just for this particular purpose, it’s not really needed as WK now has its own “Recent mistakes” section so one could note down the failed items from there to practice them separately.
Yes yes, they explain how the formula works in the script’s settings screen
I think we all feel this frustration to some extent. It’s always painful to miss a burn review … again. But WK is unlikely to ever make the change you are requesting, simply because it undermines the premise of the SRS. “Burned” should mean that you can go four months between reviews and still remember the word. Maybe that’s because you’ve learned it so well that you just remember it four months after you last saw it, or maybe that’s because you’ve seen the word in the wild frequently enough that it stays fresh. But if you find yourself needing to artificially introduce extra reviews during that 4-month window, then maybe you don’t really know the word well enough to call it “burned”. Once you burn it, you’ll stop doing the extra reviews, at which point you’re likely to forget it within four months. At least, that’s the basic idea of SRS.
How you respond to that premise depends a lot on your personal philosophy and how well you want to learn the words. If you want to truly master every word, then I’d encourage you to address leeches with some approach other than more frequent repetition. Maybe try to come up with a better mnemonic to anchor it more solidly in your memory. Or practice writing the word by hand, which has been thoroughly proven to improve retention.
On the other hand, maybe you don’t need to master every word. If you’re reading Japanese regularly, but you’re not ever coming across a certain word, then maybe you just decide you don’t care very much if you can remember that word four months later - you never see it in the real world anyway. In that case, there are plenty of scripts that let you “cheat” - i.e., deviate from the basic SRS timeline by doing extra targeted self-study (e.g., on leeches or by SRS stage), or by just using an undo button to change a wrong answer. No judgment here. Find an approach that gets you where you want to be with the Japanese language, and don’t lose any sleep over it.
Honestly, once I complete level 60, I’m not sure if I’d try to burn everything. At that point I think it’s better to do more immersion like reading books and watching shows or practicing conversational skills with a tutor.
This is ultimately why I quit wanikani. The format didn’t work for me after about level 20 or so, and even with scripts I found it wasn’t right. I think creating a one size fits all program like wanikani is itself an act of hubris, although I also admit that it has its merits.
In any case, having an unchangeable default setting is itself opinionated, although it works for some.
I totally get this. That’s why I add my WK words to JPDB for some extra practice (though I’m thinking of moving my extra drills to MM since I need the grammar review anyway).
The option for studying wrong items doesn’t help much after the time passes but you still don’t see them for a while.
That being said, I don’t care too much if I get it wrong and have to do it again. It happens. But I’m also a person that doesn’t mind seeing 200-400 reviews pile up. I’ll get them all to 0 before I level up, little by little.
The problem is… a fixed review schedule will never fit everyone.
Sure, you might get to a point where you can go four months (or more!) without seeing an item and still remember it, but for some that might happen only once they’ve seen it 100 times before that 4-month period, not the 11 times (or however many there are) WK showed it up to that point.
Hence the need for additional practice in-between. The leech training option of Self-study Quiz is quite cool, because it will show you items until you reach a successful streak in WK’s review schedule, then it’s gone until you fail it again.
So eventually you’ll still get to not see it for four months before you can “burn” it.
Of course, WK or any other SRS system for that matter, cannot account for your own reading. You might get to see an item a lot of times in content before the final burn review and you ace the review when it comes up. Does that really mean the SRS concept succeeded, if you actually saw it in a book a week or three before the review?
I have 2800+ burned right now in WK, but I am certain that some (who knows, maybe even most?) of those are not truly seared into memory
Yes, I completely agree. Extra practice to reinforce words is very helpful. I just don’t want to see people fall into the trap of cheating themselves by performing targeted reviews shortly before the words come up for review. That keeps them from realizing the benefit of SRS. And my original point was that SRS is the structure WK is based on, so WK won’t simply do away with the spaced repetition intervals, per this thread’s original suggestion. The fixed timing that WK uses isn’t a one-size-fits-all solution for everyone, but it’s a good framework for those who want to use a long-term SRS option, especially with the personal adaptation available through user scripts. Just be sure you understand the system, and be mindful not to use scripts in a way that will undermine it. It’s a little sad to see threads here in which people say they’re quitting WK because the system isn’t working for them, when they’ve also made many comments about all the ways they’ve sidestepped the system with user scripts because they’re in a hurry to burn.
One problem that seems to lead to a lot of user frustration is the misconception that we should be burning words as fast as possible, after the initial 9 reviews, and that if we get a burn review wrong, that means we’ve failed, or the system has failed to work for us. That’s silly. Of course you’ll need to see some words more than 9 times. Missing a burn review doesn’t mean something is wrong - it just means you need more exposure to that word. Internalizing a large vocabulary takes a long time. Spreading out your study over many months will probably help you learn it more permanently than cramming a million reviews into a shorter window. There may be stubborn words that you’ll need to recycle repeatedly over the course of a couple of years. That doesn’t mean there’s a problem; that’s how the process is supposed to work. The miracle of SRS is that there are so many words that do get burned after only 9 exposures.
This long-term SRS approach may not be the one every Japanese learner wants to use, but it is a perfectly valid approach if you relax and trust the system.
The system itself is a giant misconception. There is no such thing as permanently burning something. Your brain is constantly in the process of forgetting things. If you get an item right 9 times in a row, it doesn’t mean you magically won’t forget it on the 10th. If you spend 20 years to burn an item on the 100th try, it doesn’t mean the next time you see it you will remember it.
It’s basically trusting something that will not work the way you believe it will and chasing something that you will never get.
SRS is something you can do in tandem with immersion and application to increase your efficiency in consuming native material. This consumption is what will build the strongest connections to the numerous items.
There is no point waiting for predetermined intervals to review or see something again when you could either be seeing it or using it all the time through immersion, or never needing it ever because your chances of seeing it in practical application is slim to none.
So since the system itself is flawed and it’s not leading towards some mythical end goal of having everything truly “burned”, the notion of undermining it, or cheating it/yourself, or sticking to its highly rigid foundation is kind of pointless. You should be able to customize it and do with it whatever you feel fit to get the best gains for yourself.
To be frank, at some point I feel like the goal should be to drop SRS all together and let immersion sort out the rest. Spending a life time on it believing it will make magic happen is the real trap itself. This is kind of the point of “finishing” WaniKani quickly. You got what you feel like you needed out of it and you moved on to bigger and better things.
When I got to level 60, I burned everything, but it was kind of to complete everything and not really for some perfect kanji recital benefit. However, the part I did benefit most from was continuing with Wanikani for at least 6 more months, which gives you the opportunity to burn everything nominally. Anything that falls between the cracks at that point is something that you can look through by yourself, doesn’t really benefit you to wait 2 months for 1 review at some point
I’m now going through Wanikani again, mostly because the knowledge didn’t really stay when I didn’t read anything significant after completing it, but also because I want to go through it again with the knowledge that I have now. I am noticing much more benefit from that compared to burning every single little thing if I’m quite honest
Well, yes and no. I think it’s definitely good to try to quickly get to using the language and having that be a big part of your study (it’s a lot more fun than SRS reps, for a start). But I also think it has its limits: I spent quite a long time on “just reading” without doing much srs work, and I found when I restarted doing vocab work that way that actually there were quite a lot of words which I was encountering in my reading but just kind of skating over without really learning them, because I was getting the gist of what was going on with the sentence and the plot without them. So I’ve found a little SRS work on the side to be pretty handy for moving up from a 10-15K-ish vocab plateau I was on to something more like 20-25K.
MaruMori. JPDB is nice (and free) but it is Anki-style and I wanted to be quizzed more strictly.
Personally I don’t think extra reviews in between ‘hurts’ the SRS process - at least not for me. Having multiple SRS started at different times means I create natural staggering (I try not to do them back to back) and I find it very reinforcing. Then I go read books and I have found quite a lot of the vocab there - which I would normally have glossed over or looked up (and immediately forgotten) if it weren’t for the SRS drills.
Getting things wrong is normal. Failing an item about to be burned is normal. But wanting to do extra reviews to drill things more is also normal and if a person determines, after trying it out for some time, that it legitimately helps them, then it’s ok to do that.
I see what you mean and I definitely agree that there is value revisiting srs when you feel the need for it. I kind of just meant stepping away from it for at least awhile, not necessarily never coming back to it again.
Although when I say immersion, like lots of reading for example, I don’t just mean doing it in a vacuum. Or in other words, reading what you can and just moving on. I mean still translating or looking up the words and grammar that you didn’t understand.
In some ways, reading can basically replace srs because every time you see a word in a sentence, it’s kind of like doing a review in and of itself. So if you see an item in your reading, and you don’t know or don’t remember it, you look it up like a lesson/review, and then when you see it again in a different sentence that is your “srs” review.
The advantage to getting a lot of srs out of the way is that it increases your efficiency and rate of consumption of immersion. So you are not stopping at every single word having to look it up. This gives you more opportunity to essentially do more reviews of everything because you are understanding most of it and stopping for the things you need more practice or review on.
So at some point you hit this threshold where after some srs work like Wanikani, you have enough floating in your head to where you can read or listen efficiently enough to for example get through 5 or 10 pages or an episode of something in a reasonable amount of time while still stopping and looking things up.
At that point, I just feel like it’s worth slowly stepping away from srs because the immersion itself takes over as your “new srs”.
I definitely feel the frustration of having an item go back several stages (particularly when those enlightened items go back to guru), but we should also think about how big this “punishment” really is. If you really did know the item, happened to get it wrong but it was close to memorized, then you’ll see it once again to guru it, once again to master, enlighten, then burn. If you really know it, that takes 2 seconds each time. Sure it sets you back a long time, but it’s one item out of five hundred. The effort itself is very little.
The only items that are a real pain, that actually take up a significant amount of your time, are the ones that you continue to get wrong over and over again, that go from master to apprentice to guru to master to apprentice again. The ones you don’t actually know, and need to keep reviewing.
SRS measures how solidly you know an item based on how long from when you last saw it will you still remember. So it doesn’t matter if you got it right it once in April or 10 times in April, it wants to know if you’ll still remember in September. If you get it wrong in September, rapidly practicing it a few more times that day isn’t gonna help you in March- but periodically practicing it until you’re more solid on it should.
So yeah, it’s annoying to see an item you feel like you know go back in review stages, but for the ones that you do basically know, then how much more effort is WK actually asking of you? But I mean, the more important question is just if SRS is effective for you. If you feel like all of this is only slowing you down, then fair enough. But I think WK’s logic is sound, personally.
Hmmm… yes and no.
If you’re at a stage where you can read 100+ pages a day, probably. At that speed it’s likely you already have a 15-20k base and whatever new shows up will be “easy”.
But early on, when barely manage 10-20 pages until the brain melts, there is nowhere near enough repetition of words to ensure you see them regularly enough for them to stick. Unless, dunno, you read something that has zero enjoyment because it keeps repeating a relatively small vocab sub-set, but why would you put yourself through that?
I just finished a 400p book with (according to JPDB) 3684 unique words used once and it took me 3 weeks and a half. Out of those, I don’t expect many to show up in the next book and even if they do, if I only saw a word once 2 weeks ago and in the mean time I had to deal with a few other thousand, the chances of my remembering it next time are… well… nil.
I think I agree with some of what you are saying, but I kind of feel like it’s being a bit over exaggerated.
Right after I wrote that part I mentioned it being a threshold where after a certain amount of SRS you have enough floating in your head to become more efficient with immersion and it’s worth slowly stepping away from SRS.
I’m not suggesting doing no SRS, I’m just saying that it can be replaced/removed over time, and probably sooner than people feel comfortable with stopping it.
I’m trying to understand your example with the book you read, but I’m not really sure I get it. By the 3684 words used once, do you mean that was the total unique words in the book or there were others used more than once? There was nothing repeated mutliple times in the book to where you will remember anything again? If those words were so specific and unique and you don’t expect to see them in the next book, why is it important to memorize them with SRS?
Also, if you’re saying that after 2 weeks you might forget a word while studying others, Wanikani has you waiting months to see a word again.
