For me it makes sense that some people feel like they have too many reviews to get through, whilst others struggle with too few to feel even mildly challenged or activated - its because all learners will have their own different abilities, time-constraints, and prior levels of proficiency outside of WK. Its a really basic concept that different learners have different requirements and why so many SRS systems often feature codes which are adaptive to learners pace.
My original hope was that I could reach L20 in time, which I have quickly worked out isn’t possible due to WKs rigid progression. It means I can’t really use it for reaching my current goals, but not that it’s entirely unsuitable as a long term side revision tool
Was more wondering how those who do find it frustrating initially push through to get to the later levels where it’s no longer so. It seems to be a common issue that people struggle with in the earlier levels, regardless of whether or not it’s an issue in later levels. So was hoping to turn to the community of learners who’d had this experience and ask their advice on staying motivated through that unchallenging period.
That’s a really nice chart!
Any chance there’s some floating around for what happens when someone doesn’t keep up with reviews? Missing time on this site is the worst…
well if you want an example of someone who kept missing reviews, there’s a thread up right now featuring someone who is currently sitting on a massive 1500 reviews
my understanding is that the chart is a theoretical 100% accuracy
so anything lower than that will increase the slope everywhere
delays would cause pits and spikes, but assuming it’s 100% accuracy the average would remain the same. I do believe your accuracy will drop if you disrespect the SRS interval
Coming back after a multi-year hiatus, I’m glad to see other people have come to a more nuanced, less religious conclusion besides the “you’ll understand later” that’s in the FAQ. The fact that this topic keeps coming up is a real problem with the WaniKani approach that needs to be properly understood. The fundamental issue is that language acquisition is complicated, and can’t be solved with 20 lines of PHP.
Basically, WaniKani levels 1-5 are really easy, regardless of your experience level, and leave you feeling impatient, again regardless of your experience level. Levels 10-20 are kind of the sweet spot for the schedule – just hard enough that you’re likely to get them wrong on the first few tries, and you start to feel like the backlog might just start to get overwhelming soon. After that point, the kanji are hard, but the pace remains the same. The component radicals are more complex, there are more of them, and the mnemonics get progressively more arbitrary. Not only are you dealing with the combined weight of all the previous words in the backlog, but you’re also more likely to get the current ones wrong. Each one requires more effort in order to get to a state of automaticity.
The long-haulers here give good advice on how to manage these distinct phases, recommending patience in the early stages, and mitigation strategies for the inevitable flood in the later stages, but the fundamental issue is that the SRS algorithm is flawed. Anyone using it needs to be aware of the limitations, and not blindly trust it like it’s your sensei.
you mention arbitrary mnemonics, but aren’t the mnemonics already ridiculous to start with? You’re telling me that they get more absurd, not less?
Anyways, I get the impression that yes, the SRS algorithm is flawed, and there are a lot of dropouts, but I think overall Wanikani is a solid package. It could almost certainly come with some better instructions, though.
I wouldn’t say flawed. It has advantages and drawbacks just like any other method.
I think it’s just that the way WK is built doesn’t lend itself to item skipping.
I think there’s an argument to be made for something like having all levels be fast, but I also think there’s a very low likelihood of WK implement it.
