Hey guys, here's the problem:

I start learned RTK1 about 9 months ago and knocked out about 700 kanji. Sadly, I stopped learning and reviewing since that time, but have since decided to come back and make the commitment. After finally logging in after all this time, my stacks have all expired plus there are still some failed kanji to review. So the question is, would it be better to just start over from scratch or try and continue? Any thoughts/ideas would be helpful. Thanks.

http://forum.koohii.com/showthread.php?tid=202
http://forum.koohii.com/showthread.php?tid=318

All these kind of threads that pop up every month should be made a sticky or maybe a FAQ thread with all these links in it.

Bite the bullet and start over.

If you have good recollection they will get blasted through

If you don't then they needed to be started again anyways.

SRS should adjust either way but it's easier to get back into it in the original order I think, I am on my second time through and blasted through 500 in like 6 days without trying very hard with 95%+ retention usually

My advice would be to just start from scratch. You should be able to get back to where you were quite quickly since you already know most of them.

Something else you could do is, just pass everything and continue like normal, everything will even itself out eventually.

Where else but Kanji.koohii would we be taking advise from users named Hotkiller and Bombperson on a serious subject? Hey! Where's Nukemarine?

Start back over. You need the early kanji embeded in your mind to ensure the creating stories process for the rest of the book to be as fluid as possible.

Start over, but try and do it a bit quicker. Heisig recommends to try and finish the book as quickly as possible. Might help you, rather than taking it slowly.

That's odd, I'd say keep going, after a quick review of the old ones (especially the primitives). You will have plenty of time to restore your memory after you've finished.

I made it to over 500 and then took a 4 month hiatus. Had over 400 cards expired. I whipped through the reviews in ten days and have never looked back. I failed a lot of those reviews, but not all. The ones that I failed came back quickly. No need to start from zero. Anything you remember after 9 months is already in long-term memory, no need to move it through 4 piles. Why waste time starting from scratch?

It really doesn't matter which route you take.

If you are honest, than the system takes care of the rest! stuff that needs review gets it, either way

I'm of the "keep going" school. If you have indeed forgotten the kanji, you'll fail it, and it will start all over in the failed kanji stack, for you to focus your review on what has been forgotten. But if you remember it, then it deserves to move ahead to the next box. Trust the review to tell you what to spend your study time on!

I don't think there is any advantage to starting over from a learning point of view.

I stopped for more than a year and then picked up again. I don't think it made any difference in my retention rates than if I had started again from scratch.

[ファブリス]

I wouldn't start over. Pass the cards that you still remember, and "fail" the rest. The refactored site (test.koohii.com) includes an easier way to review the failed list, so it is more practical to pick up from long expired cards, then it was before. I'm still tweaking the test site, but it should go live in a couple weeks or so, until then, you may even experiment with your expired flashcard on the test.koohii.com site, since you account will exist there too.

I did the same thing as the OP. I ended up starting over from scratch but I found it rather easy to make up for the lost time. Unfortunately, I only became really enthusiastic a month ago and since then I have been pushing forward. I think that not starting over would encourage you to work harder until you get them back in your head. In this case, I would just keep what you have and work hard. I wish I hadn't started over, but at least I gained the enthusiasm eventually

I too suggest starting from scratch. You'll be able to quickly get through all the familiar material, while also refreshing your memory of it, and relearn what you've forgotten.

I suggest not starting over and letting the SRS do it's work. Just make sure you are honest with your answers.

It was easier for me to get back into the swing of things and redevelop my SRS habits by starting with the more familiar materials, plus having 2000 cards expired and failing a whole bunch of them and having so many faileds to re-learn, i personally found it much less intimidating to hit "reset" and go from scratch since the SRS is out of whack anyways. I was then able to easily review 100 in a day, quiz them in and keep up with reviews that way with much better retention thus far than when I was trying to get back into my SRS with the 2000+ review pile ><

Like I've said it doesn't really matter which you do, the SRS ends up adjusting to you as long as you are honest and keep up with your reviews. It's a personal choice and the IMPORTANT thing is that you keep with it, it doesn't take long to get back to where you were either way, just keep at it and move on with your studies, maintaining your SRS no matter what ^^

I did the same thing. I wasn't really feeling the Heisig method so I was basically had to force myself to learn it. I stopped at around 500 kanji and was stuck there for several months. Just don't get caught up with it. Even if you forgot a couple of hundred it's okay. Finish it and read something in Japanese.

Fabrice,

Will the new site have the option to take into account long pauses in review? Example, a card in box 3 (1 week review) sat there for 4 weeks, so if passed will count as if it were in box 5 (1 month review) and get put in box 6?

I like the box approach, but there shouldn't be too much wrong with taking paused reviewing into account. Seems a waste to remember a number of cards after a month delay only to have them go into a 1 week or 2 week review pile.

[ファブリス]

EDIT: Nevermind. I see what you mean but I don't know if that would work with Leitner actually. Currently the boxes represent directly the number of successive succesful repetitions of an item. If a card jumps two boxes because of longer interval of review than the one that was scheduled, it could get a greater next interval, but the number of repetitions wouldn't match if approximating the next box based on the interval.

Maybe a simple improvement would be to adapt the next interval with a "bonus" offset of up to 1 month or so, to account for the extra expired time?

So yeah, if the intervals can vary greatly while keeping the box as number of successive succesful repetitions; than the graph can still be representative of the level of knowledge, but not of the expected review load. I mean, you could have green cards in box 3 that stay there for much longer than others.