100 a day to 'learn' Kanji is nigh impossible with any method, as far as learning readings and everything else go. For assigning keywords, Heisig is probably the quickest way to do it in terms of amount of hours it takes, though everyones mind works differently.

raseru wrote:

I didn't see the post but the thing is, lots of people say they can learn a ton but it doesn't mean jack if you can't retain it the next day.

Also, he's probably studying a similar way. Learning radicals to learn kanji makes a lot easier than school methods. Personally I think heisig has an amazing idea he just does it poorly because it book is filled with mistakes, his stories are truly garbage, and he chooses stupid names for the stuff he makes up making it harder for us to learn it.

You really think Heisig's stories are garbage? Personally, I think he is unusually talented at making stories. For me, so many of them stick instantly and are amusing as well.

Not at all. I'm adding 300-400 kanji a week, and I'm not struggling.

Well, I had one stint of 3 days where my review dropped into the 60's, and I didn't clear out my fail pile until it had grown to epic kanji binge (completed) porportions. However, now my reviews are always over 90%, and my fail pile is less than 5% of my total kanji learned.

How fast you study new kanji is not so strictly related to your retention rates, imho.

It took me 10 months to do the first ~1400 characters, but then for various reasons last summer I decided I wanted to speed things up, and did the last 600 in 10 days. Well, I haven't noticed the slightest difference in retention between the two groups of kanji. I can remember those last 600 just as well as the rest.
I think that doing reps diligently even after you've finished the book is what really matters.

For those reasons, looking back, I'd suggest doing as many kanji a day as you can, without worrying too much about making your stories perfect. Amazingly, a good balance is the best thing.

This thread has a lot to do with the 'Thinks you're crazy thread' IMO.  If someone says they learnt 200 kanji in a day we should be slapping them on the back and high fiving them, not hassling them and calling them liars.  If someone is trying hard then I salute them.  Doing them fast is admirable, not deplorable.  Wan Zafran did Heisig in twenty days, didn't he?  A hundred a day.  Awesome.  Good job.

Who knows what some people can do?  If you skip over to the 'how to learn any languages' forum you can meet a guy who knows dozens of languages.  As in more than thirty.  He can speak olde Nordic, hit it up on youtube some time, it sounds incredible.  He knows Chinese, and Korean, and Japanese, and Hindi, and Arabic, and Persian, and Sanskrit, and he can write in all of them.

NightSky wrote:

100 a day to 'learn' Kanji is nigh impossible with any method, as far as learning readings and everything else go. For assigning keywords, Heisig is probably the quickest way to do it in terms of amount of hours it takes, though everyones mind works differently.

Mmmm.... I was learning about 100-200 kanji(just meanings) a day using that cheesy game Slime Forest Adventure(the Kanji Dream version). I didn't really play the game so much as walk back and forth to make kanji pop up. I had good retention rates too, except that after I finished the kanji, I stopped reviewing. I think it only took a couple hours each day too, so it was really fast and I could do each day in one sitting. The game has plenty of flaws, but at the time it was totally worth it... anyway, this is why I can't say I actually finished RTK, because I switched to SFA part way through.

The problem with doing too many kanji a day is that the number of daily revisions can quickly skyrocket to something unmanageable. It's not that hard to study 100 kanji a day for about a week but then you'll have to review the kanji from the week before and after that those you reviewed two weeks before and so on.

My estimation for the SRS system of this site, based on my experience is that you should expect daily revisions to peak at about 5 times the number of kanji you add daily so plan accordingly.

Also while it's easy to conclude that there is little impact on retention rate after a revision or two, what really matters is the retention rate once those kanji are moved to the long term stacks. And then your results could be different.

I'm not saying that you shouldn't go fast, I think going fast is important with Heisig, once you have seen a kanji even once it's much easier to reinforce your knowledge through reading. But be careful, don't bite more than you can chew.

The amount of revisions will decrease but not that fast and after finishing the book it's harder to stay motivated. Personally, I just stopped reviewing after a while and let expired cards accumulate up to ~1500. Now I went through all my expired cards quickly and I have a huge failed pile that I'll need to slowly reintroduce into the SRS. But that's OK as I don't have the pressure to finish the book. I have already seen all those kanji, if I don't remember them all perfectly it doesn't matter I'll slowly revise everything.

Codexus wrote:

My estimation for the SRS system of this site, based on my experience is that you should expect daily revisions to peak at about 5 times the number of kanji you add daily so plan accordingly.

When does this happen?

I've been going at about 40-50 a day for 2 and a half weeks, and the reviews have never gone over about 105.

After 54 days in theory as that should be the time it takes for the first kanji you studied to expire on the 5th pile. At that point you have the blue pile of the first review of the new kanji (which I used to review the kanji I added the day before), you have to review the second pile of 3 days ago and the other piles after 7, 14 and 30 days so that's 5 revisions in total.

And that's if you never forget a kanji. In reality you're going to see some kanji more often than that so the number of revisions is going to increase more quickly and you'll get to that number without even counting the blue pile

NightSky wrote:

I'm not sure what you are trying to say here, that Heisig isn't the best way to learn english keywords for Japanese Kanji? That your game does it better?

I'm saying RTK isn't the only method one can use to learn kanji quickly. The game was faster than RTK, but like I said it had problems in many areas.

NightSky wrote:

Forgive me if I'm wrong, but weren't you posting on TJP a while back, singing Heisig's praises and getting a bit of a hard time for it? The only reason I bring it up is that it was quite a long time ago, and if Heisig is as effective as you were saying you should have had it all finished quite some time ago The fact that you are still trying to get the keywords down (I say keywords, since it seems you haven't actually started learning to read Kanji yet) is quite interesting :]

I don't understand how attaching words to kanji after learning their form is in any way inferior to attaching words to kanji while learning their form.  In both cases, you are learning kanji and learning words, so what's the big deal?  I happen to feel that splitting things up and spending a month or two upfront learning how to write the general use kanji is a worthwhile endeavor, since it makes it so much easier for me to acquire vocabulary later, but as long as you're learning Japanese and spending time with the language, I say more power to you.

I really don't get all this Heisig/Anti-Heisig infighting. People have reached fluency with Heisig's methods as a start, and people have also reached fluency without them.  Arguing about how one way is better or worse than the other seems like a terrible waste of time.