80% is somewhat low, but it's still massive progress. Think about it, you answered 4 out of 5 cards right. I agree with snispilbor, what's most important is, how many times have you failed those 20% before? If most of them are the first time you've failed them, who cares? Just make the story slightly better and spend more time thinking/visualising it. I mean, it's 1 card out of 5, your failed pile can't be big or anything.

It depends on who you are.  I've heard some people say their sweet spot is 70-80%.  It reinforces the un-learned ones while bringing new stuff at a fast enough pace to keep them interested.

In the end, it's all about you, and not about anyone else.  If you can't stand failure, you're better off studying fewer and staying in the 90's.  If you crave progress and can handle the load, you're better off taking on more at once and dealing with the 1/4 failure rate.

I read somewhere (actually, I think it was in Anki itself) that 80-95% is an ideal range. I don't know who came up with those numbers, nor from whence they were pulled, but let's not dwell on that. I personally aim to maintain a 90%+ retention rate for sentences and ideally 98%+ for kanji, but I think the latter may be overly masochistic.

Hey nicko that is cheating! erm.. a nice idea!

I hadn't reviewed in a couple days..   There was a hefty 132 kanji in my stack.

I passed with 92% ..  That's pretty good, right? ^-^


It's still nerve wracking when you can't come up with the particular images.. especially when words like distinguish and distinction and other similar pairs exist..

I just started (140 Kanji) and I'm getting 50-60% but I think my methods are improving so I'm not getting frustrated or anything yet.
But I was wondering about the new kanji that I just studied (I try to do 20-25 at a time). Should I add the cards and review them right away or do I benefit more from doing it the day after or something?

I think 80% is a good goal to look for in the long run. DO NOT expect it on your initial review though. Of the 36 I added two days, I failed 18 the following day on initial reviews. On normal reviews I can range from 70% to 92%. That's with an overall average of 83% on 26000 reviews.

As you get further in, you'll get a happy medium of adding cards, reviewing cards, studying failed cards, reviewing added cards. Don't let the results of one day sway you one way or another too far.

Last edited by Nukemarine (2008 October 07, 11:13 pm)

It depends how rapidly you are adding new new characters.  If you are adding 10-20 a day, you may get by with 70%, but if you are going faster or if you miss a few days, the number of characters to review can get painful very quickly.

For example, I completed RTK1 last weekend after zerking through the last 150 or so.  Having burnt out, I didn't review for a day and a half, and was rewarded with a 660+ character review.  Now, you would think a passing percentage of 92%+ on that many characters is pretty good, but I was still left with a fail pile of ~50 cards.  OUCH!

It's a matter of efficiency.  If you fail at a high percentage or if you keep failing certain characters, you will pay with more repetition. 

Is it better to repeat or to spend more time to learn it right, first?  It depends on your personal biases.  If repetition turns you off, then you should spend more time to learn it right.  If taking too much time to learn it right makes you fall asleep, then you should fall on repetition or reduce the number of characters per day. 

Know yourself, and you will know the correct percentage range.

The WORST thing to do is, to stop reviewing or not adding any characters at all.

My reviews were always bad, sometimes as low as 30%, never as high as 80%. And I still managed to do it in 3 months.

There is just 1 rule: Never cheat your reviews.

Hmm I don't think I have ever had a review as low as 80%. I very rarely drop below 95%. I don't like to get even 1 wrong.

How about the other way around: having too high a pass rate?  I'm doing about 15-20 new kanji per day and maintaining a pass ratio of > 95% on all reviews (with the exception of just one day where I did the review in the morning before coffee -- I could feel my thinking processes go all sludgy).  I've been wondering if I could be more efficient if I went faster and accepted a lower pass rate on my first couple reviews.  It might be that 95% pass rate is suboptimal in terms of getting things done quickly ...

mentat_kgs wrote:

My reviews were always bad, sometimes as low as 30%, never as high as 80%. And I still managed to do it in 3 months.

There is just 1 rule: Never cheat your reviews.

Amen, brother!