Oh, Nath. You beat Gsai in every way.
1 Like
Made it to 99.00%? It sounds weird, but my accuracy is also falling in that direction. English meanings are getting more obscure, and burn reviews have finally arrived. Why are meanings so much harder than readings? 
I mean made it to 99% because I started around 97%.
- About half of my meaning and reading errors came from the first five levels:
At first, due to me not distinguishing between radical/kanji/vocab and meaning/reading. Now I finally got used to the visual cues (blue is radical, white on black is reading, …) but it took a while. - Then typos became predominant. I had to learn to slow down and proofread myself. It still feels pretty hard and frustrating, but the red feedback I get from typing “scyckle” instead of “sickle” is worse.
- Now, all my mistakes are pretty much due to me writing a meaning that I think is valid. When I get it red, my first reaction is “BUT IT MEANS THE SAME THING!”, then I try to make a sentence with that “synonym” and I’m like “wait, that doesn’t actually mean the same thing
”. So I am slowing down even more, and I try to make sentences (in English) in my head, so that I don’t type “other” instead of “separate” for 別 or something.
@gojarappe Sadly, I will never have a hawt body like he does, so I guess not.
2 Likes
I knew absolutely zero Japanese other than kana when I started on WK so I feel like I’ve done pretty well so far.
2 Likes
Here as consolation to ungodlike people such as I am:
I’m 48 years old, I have a job, and I’ve never studied Japanese before. Also too old to feel competitive, which is nice. Join me!
6 Likes
I think your stats are still excellent for your situation. Fantastic job!
Whoops, my bad. Still, that means you did way better than 99% to reach that average. Congrats!
Thanks!
It takes unreasonable amount of time and focus while doing my reviews to achieve it, though, so now that I have bragging rights, I may let myself slip back to something a bit lower 
1 Like
i’ll take it! Thanks.
My accuracy is like 85%. You must have a way easier time with it then I do. (And way less reviews!)
Well, since making a mistake knocks an item back 2 ranks, it means you need to get it correct an extra three times (that’s an overestimation, though, since you may get a mistake on both reading AND meaning in the same session, or the item might have been apprentice 1 or 2).
So, 85% gives you an approximate 145% workload in terms of reviews, while 99% means 103%. So you are indeed doing a lot more reviews than I am
Other than that, it’s my fourth time around learning Joyo kanjis
Look at them pretending to be all burned and stuff:
But no matter what, I still have leeches and stuff I cannot remember. Hopefully this time around things will stick in my head.
(Still, I do have a lot of a head start)
2 Likes
Still okay numbers, I think, though lately reviews with <90% accuracy seem to have become very common for me.
This thread was interesting to scroll through lol
My accuracy:
I want to make these numbers even higher though. I tend to make stupid mistakes as I often rush through my reviews
It used to be higher, but I went on vacation twice this summer, and a bunch of apprentice and guru readings flew right on out of my head each time Still not too shabby I think.
Nice thread you guys have here.
If I get to lvl 60, I’ll post my accuracy chart. Any one is free to remind me to do so.
I want the full panorama of it, after all.
y’all doing great, keep it up
1 Like
For you Accuracy aficionados, I would be curious to see your accuracy by SRS level.
If you have Open Framework script installed, paste this in the Javascript console:
Click to open code
wkof.include('Apiv2');
wkof.ready('Apiv2').then(fetch_data).then(process_data);
function fetch_data() {
return wkof.Apiv2.fetch_endpoint('/reviews');
}
function process_data(json) {
let stages = ['Lessons', 'Apprentice 1', 'Apprentice 2', 'Apprentice 3', 'Apprentice 4', 'Guru 1', 'Guru 2', 'Master', 'Enlightened'];
let reviews = json.data;
let srs_total = new Array(9).fill(0);
let srs_correct = new Array(9).fill(0);
reviews.forEach(review => {
srs_total[review.data.starting_srs_stage]++;
if (review.data.ending_srs_stage > review.data.starting_srs_stage) {
srs_correct[review.data.starting_srs_stage]++;
}
});
let output = ['Accuracy:'];
for (let i=1; i<9; i++) {
if (srs_total[i] > 0) {
output.push(' '+stages[i]+': '+(srs_correct[i] / srs_total[i] * 100).toFixed(2)+'%');
} else {
output.push(' '+stages[i]+': ---');
}
}
console.log(output.join('\n'));
}
And after it finishes loading the data, you’ll get something like this:
Accuracy:
Apprentice 1: 85.78%
Apprentice 2: 96.92%
Apprentice 3: 96.34%
Apprentice 4: 94.23%
Guru 1: 86.94%
Guru 2: 80.85%
Master: 79.42%
Enlightened: 73.54%
(Note: This only covers reviews since about Fall of last year, since that’s when Wanikani added review results to the API)
(Also, if you don’t have Open Framework installed, you can also run it from the Javascript console while on the new in-development version of wkstats.com [here])
7 Likes
-googles how to open javascript console-
Accuracy:
Apprentice 1: 97.89%
Apprentice 2: 98.38%
Apprentice 3: 98.39%
Apprentice 4: 97.39%
Guru 1: 94.82%
Guru 2: 92.97%
Master: 92.50%
Enlightened: 91.36%
/programmerstatusachieved
6 Likes