I'm just gonna say this now, if you're going to use flashcards, anki is the best way to go about it. 2200+ paper flashcards is stupidly unwieldy.

Ok, so quick update: I was unable to start studying seriously until early October. Since then, I have been doing about a percent a day, and I am now up to about 25%. I took this past week off to review what I'e learned so far, and I am going to try to continue at up to 10% a week until I get to a point were I feel comfortable taking another break to review. I'm sure both of you are likely past 50% at this point, but I'm going to try to catch up as best I can.

aldebrn Wrote:

john555 Wrote:Here's a suggeston from someone (me ) who finished RTK1: try doing it for a while without using this Anki thing.

Explain. Do you mean make paper flashcards as you pass through the book and review all flashcards periodically? Do you mean don't review (no Anki, no paper flashcards), at all or until you've learned another big chunk (~20?)? I feel like an idiot for entwining "RTK" with "Anki/Koohii/Memrise/online tool" for so long that I don't understand how to do RTK by itself, despite hating all the extra baggage of software.

It's really simple actually. After you process a bunch of characters (meaning study the character and make up a story about it) you practice with the book itself, covering up the keyword with a piece of paper and writing the kanji, then later test yourself randomly (by using paper flashcards OR sorting an excel list randomly and printing it out). I used to do around 10-20 kanji a day.

Every weekend do a marathon review of 300+ kanji. I used to go through the Heisig book in order, again using a card to cover the kanji but show the keyword, write the kanji, then slide the card down to the next one. But you could also print out a randomly sorted excel list with the kanji hidden and the keyword displayed. The you write the kanji on the printout beside each keyword. Then check your answers.

Every few weeks do a marathon-marathon review. So I would do #1 to #508 on Saturday, then do #509-1,009 on Sunday. By review I mean look at the keyword and write the kanji from memory.

Unwritten_Oracle Wrote:Ok, so quick update: I was unable to start studying seriously until early October. Since then, I have been doing about a percent a day, and I am now up to about 25%. I took this past week off to review what I'e learned so far, and I am going to try to continue at up to 10% a week until I get to a point were I feel comfortable taking another break to review. I'm sure both of you are likely past 50% at this point, but I'm going to try to catch up as best I can.

Nah, you're ahead of me. I wound up splitting up my full RtK deck into 5 separate decks (each which corresponds to a JLPT level), and started from scratch. I added all the cards in that I could already write easily, and then filled in the blanks in the N5 and N4 decks, so that I was 100% thru them. You can see my post on that here: http://forum.koohii.com/showthread.php?tid=12260.

This worked much better for me because my class has Kanji tests every other week, and the kanji closely mirrors the same kanji on the N5 and N4 exams. (We're finishing Genki 2 now, and and the kanji tests include those plus some others that my teacher wants us to know).

So overall, I am 21% thru the deck. 10% are mature, 11% are "learning".

The breakdown is:

N5 (108 cards): 80% mature, 20% learning
N4 (211 cards): 40% mature, 60% learning
N3 (406 cards): 8% mature, 14% learning 78% unknown
N2 (356 cards): 2% mature, 6% learning, 91% unknown
N1 (1,119 cards): 1% mature, 2% learning, 96% unknown

My focus now is on just moving as many of the learning ones to mature, and really only adding new ones as my teacher adds them. I already don't like the number of reviews Anki asks me to do each day. I'm also taking the N4 exam in December, so this focus works well with that too.

If I decide to continue with Japanese after that, I'll likely focus on adding the N3 kanji first, and taking the N3 exam next December.

Quote:Nah, you're ahead of me. I wound up splitting up my full RtK deck into 5 separate decks (each which corresponds to a JLPT level), and started from scratch. I added all the cards in that I could already write easily, and then filled in the blanks in the N5 and N4 decks, so that I was 100% thru them. You can see my post on that here: http://forum.koohii.com/showthread.php?tid=12260.

This worked much better for me because my class has Kanji tests every other week, and the kanji closely mirrors the same kanji on the N5 and N4 exams. (We're finishing Genki 2 now, and and the kanji tests include those plus some others that my teacher wants us to know).

So overall, I am 21% thru the deck. 10% are mature, 11% are "learning".

The breakdown is:

N5: 80% mature, 20% learning
N4: 40% mature, 60% learning
N3: 8% mature, 14% learning 78% unknown
N2: 2% mature, 6% learning, 91% unknown
N1: 1% mature, 2% learning, 96% unknown

My focus now is on just moving as many of the learning ones to mature, and really only adding new ones as my teacher adds them. I already don't like the number of reviews Anki asks me to do each day. I'm also taking the N4 exam in December, so this focus works well with that too.

If I decide to continue with Japanese after that, I'll likely focus on adding the N3 kanji first, and taking the N3 exam next December.

Ah, well then good to know. I'll remember this in my pacing, and I'll be sure not to push too hard and get burnt out. I'd still like to be done with RTK1 by February at the latest so I can get to learning "real" Japanese (Beyond basic grammar rules), so I'm going to try to continue a good, constant rate. I'm just going to focus on N4 and N5 for now, and move one to N3 after that. I'm not sure If I'll ever need to go beyond that.

I'll post my stats again in about a month - probably right about the JLPT time. Hopefully by then the mature columns on all the decks are much higher.

I think that if I keep up with Japanese next year, I probably won't take a class in Q1. I like the idea of learning the kanji before the class. Whatever text my teacher uses will probably introduce the N3 kanji, so I can just focus on those. And then take a class from Q2 or Q3 onwards.

After I reached 40% I decided to stop and not resume until I've made flashcards for each % (10 Keywords, Kanji, and Stories per card). This will take a while but it's the fastest way I can learn.

Cool. Murtada and unwritten_oracle, thanks for posting and staying in the game with me. It helps me to keep it up.

Quick update guys.

I figured out another way to help me make progress thru the deck. In addition to subdividing the deck from N1 ... N5, I also started looking at the "unseen" kanji, and changing the order to put ones that I am already familiar with towards the top. I often spend time at night doing this, and also just familiarizing myself with what's to come. I find that this greatly improves my correct rate after I wind up clicking the "add new cards" button.

Here are my new stats:

overall (2200 cards): 12% mature, 13% learning, 76% unknown. (Finally hit 25%!)

N5 (108 cards): 81% mature, 19% learning
N4 (211 cards): 44% mature, 56% learning
N3 (406 cards): 9% mature, 20% learning 71% unknown
N2 (356 cards): 3% mature, 11% learning, 85% unknown
N1 (1,119 cards): 2% mature, 2% learning, 96% unknown

So within a week the unknown % in N3 went down 7pp, and the unknown % in N2 went down 6pp. So basically, something so simple - reviewing them before "adding" them, and choosing the order with which I add them, enabled me to add them a lot quicker while still maintaining a higher % correct rate (today, 84%).

The last 10 days have been very productive for me. Here are my stats:

overall (2200 cards): 12% mature, 19% learning, 69% unknown.

N5 (108 cards): 82% mature, 18% learning
N4 (211 cards): 48% mature, 52% learning
N3 (406 cards): 11% mature, 33% learning 56% unknown
N2 (356 cards): 4% mature, 24% learning, 72% unknown
N1 (1,119 cards): 2% mature, 5% learning, 93% unknown

Comparing with just 10 days ago there's a huge decrease in the number of "unknown" cards. This has to do with a few things. Obviously I added a ton of cards - that's the only way the unknown goes down. But my motivation for doing so was my upcoming trip to Japan (I just arrived last night), and a desire to understand the signs around me, even if I can't read them. I think the N5, etc. breakdown is useful for building up a nice vocabulary as you build your grammar skills and practice the four language skills (speaking, listening, reading, writing). But I remember when I used to live here I hated that even after so much time there was so much written material I encountered that I was just clueless about.

Also, I probably mentioned this earlier, but I changed my study method a few months ago, nuked all my stats, and then started from scratch. Now I actually write out by hand all my reviews. That makes my reviews take longer, but it's a trade off I'm happy with. I also take more liberties with the order in which I introduce cards. That makes it more fun and has improved retention for me.

Also, I highly recommend the Kanji Grid plugin. It's a fun way to visualize your progress and identify easy, unseen cards that you want to work on next.

Aite guys, it's December 1, a new month, let's compare where we're at. Here are my stats:

Overall: 32% thru the deck. 14% mature, 18% learning.

The breakdown is:

N5 (108 cards): 82% mature, 18% learning
N4 (211 cards): 54% mature, 46% learning
N3 (406 cards): 14% mature, 36% learning 49% unknown
N2 (356 cards): 6% mature, 22% learning, 72% unknown
N1 (1,119 cards): 3% mature, 4% learning, 93% unknown

Basically, a huge month for me. The combination of divving up the deck by JLPT test, reviewing the cards before "adding" them to the deck, switching the order of new cards to suit my needs, and using the Kanji Grid plugin to explore the decks really helped to improve my retention rate.

Any goals for the next month? Mine is to get more of my cards from learning to mature, thereby reducing the number of reviews I have each day. I crammed a bunch of new cards on my (long) flight over here, and am now dealing with the consequences. I also crammed a bunch before my trip. So it's kinda kanji overload for me.

Well I haven't been keeping up as good a pace as I was hoping, but I'm still doing decent. I'm around 39% through, with 14% mature and 25% learning, so we're pretty close ariariari.

I'll be going to a developing country with most likely no internet connection for a month so i'll have a looot of time on my hands. I plan on getting to 70% by the end of that trip and i plan on winning. Punks

Edit: In case you're wondering how,you don't need internet for the Anki app.

Alhusseini Wrote:I'll be going to a developing country with most likely no internet connection for a month so i'll have a looot of time on my hands. I plan on getting to 70% by the end of that trip and i plan on winning. Punks

Edit: In case you're wondering how,you don't need internet for the Anki app.

OH SNAP! A NEW MEMBER JUST THREW THE GAUNTLET DOWN!

As for me, my attempts to slow down have failed. I just hit 33% (15% mature, 18% learning). I swear, once I get back home, I'll slow down.

I swear.

Relax ariariari, it's me, murtada.

ariariari Wrote:Any goals for the next month? Mine is to get more of my cards from learning to mature, thereby reducing the number of reviews I have each day. I crammed a bunch of new cards on my (long) flight over here, and am now dealing with the consequences. I also crammed a bunch before my trip. So it's kinda kanji overload for me.

I just thought I should mention that this is less than optimal. Anki's algorithm, while far from perfect, spaces things out at pretty good intervals. Attempts to speed things up by cramming or "get more of my cards...to mature" actually take more time then they save and slow down the overall process. The best thing you can do with anki is to keep your overall accuracy around 70-80% (by adjusting starting ease), and keep learning accuracy at decent levels (by adding extra learning steps).

Of course it's fine if you have other reasons for doing things ahead of schedule, but don't think you're speeding things along by cramming.

yogert909 Wrote:

ariariari Wrote:Any goals for the next month? Mine is to get more of my cards from learning to mature, thereby reducing the number of reviews I have each day. I crammed a bunch of new cards on my (long) flight over here, and am now dealing with the consequences. I also crammed a bunch before my trip. So it's kinda kanji overload for me.

I just thought I should mention that this is less than optimal. Anki's algorithm, while far from perfect, spaces things out at pretty good intervals. Attempts to speed things up by cramming or "get more of my cards...to mature" actually take more time then they save and slow down the overall process. The best thing you can do with anki is to keep your overall accuracy around 70-80% (by adjusting starting ease), and keep learning accuracy at decent levels (by adding extra learning steps).

Of course it's fine if you have other reasons for doing things ahead of schedule, but don't think you're speeding things along by cramming.

Thanks for the feedback yogert909! Can you explain more about adjusting starting ease and adding extra learning steps. I feel like there's a lot of important tweaks one can do with Anki but even after reading the docs a few times I don't understand everything.

Yeah, I was totally willing to cram new kanji so that, during this trip to Japan, I could overcome one of my biggest frustrations from when I used to live here: just not understanding the written materials all around me. It's been successful in that regard, but when I go back home in a few days I really don't want to be reviewing 100 cards a day. I guess that I'm sensitive to the number of reviews Anki asks me to do each day, and I want to do them all (so I feel like I'm using Anki "properly"). So I figured I'd stop adding new cards until that number got more manageable for me.

Thanks again - your comments on this board have always been very helpful to me!

@yogert909, just a quick follow up with some stats:

Kanji:
Studied 119 cards in 56 minutes today.
Again count: 25 (79.0% correct)

Vocab:
Studied 200 cards in 36 minutes today.
Again count: 38 (81.0% correct)

It looks like my % correct is in what you think is the optimal range. I just hate how much time it takes every day. So post-presentation and post-jlpt I just wanna stop adding new cards for a while so it takes much less time.

Sounds like a good plan ariariari.

ariariari Wrote:Can you explain more about adjusting starting ease and adding extra learning steps. I feel like there's a lot of important tweaks one can do with Anki but even after reading the docs a few times I don't understand everything.

Learning steps are for new "learning" cards. Anki's default is 1 10, meaning that if you pass a brand new card you'll see it 1 minute later and then again in 10 minutes. If you find that new cards are hard to learn initially, it's a big help to change the learning steps so that there are more of them. Personally, I use 0.5 3 15 60 360 1440.

Starting ease is the how much the interval is increased every time you press "pass" for a review card. Anki's default starting ease is 250%, so intervals will be 1, 3, 8, 20....(notice that intervals are rounded up). A good idea would be to set your starting ease to the average ease of your deck. You can find the average ease in the statistics page.

Both settings can be changed in the deck settings. Here are the relevant sections in the manual for deck options and learning steps.

Alright guys, it's the end of December. I just finished writing out my anki reviews for RtK for the day. Time to post where you're at!

Overall: 36% thru the deck. 21% mature, 13% learning, 2% suspended.

The breakdown is:

N5 (108 cards): 88% mature, 11% learning, 1 suspended card
N4 (211 cards): 66% mature, 29% learning, 5% suspended
N3 (406 cards): 26% mature, 30% learning, 40% unknown, 3% suspended
N2 (356 cards): 18% mature, 13% learning, 67% unknown, 3% suspended
N1 (1,119 cards): 5% mature, 3% learning, 92% unknown, 1 suspended card

This is a *huge* improvement from where I was at the beginning of the month, which I post here:

Quote:Dec 1:
Overall: 32% thru the deck. 14% mature, 18% learning.

The breakdown is:

N5 (108 cards): 82% mature, 18% learning
N4 (211 cards): 54% mature, 46% learning
N3 (406 cards): 14% mature, 36% learning 49% unknown
N2 (356 cards): 6% mature, 22% learning, 72% unknown
N1 (1,119 cards): 3% mature, 4% learning, 93% unknown

My "lifetime achievement" graph in anki looks like this:



Overall, I think if I continue with Japanese in 2015, which it looks like I will, my goals for this deck will just be to 1) keep my daily reviews low and 2) learn all the remaining N3 kanji. I just have 163 unknown N3s, so I don't think it should be a problem.

I haven't added many new cards since my last update, so I'm still at about 40% total, but I've been making some pretty good strides in getting my daily reviews down. I'm at 20% mature and 20% learning. At this point daily reviews are low enough that I plan to start adding cards again, although I hope I'll be able to dedicate as much time as I need to after school starts again in a week.

For the new year, I'd like to get learn as many kanji as possible mature, and make better strides in grammar and vocab!

Thanks for posting, unwritten_oracle. It definitely helps me to keep my motivation up. What amazes me is how little work I'll have to do to finish the N3 kanji. Just 1 a day would let me finish by the first half of the year. I think on the N4 exam I did best on kanji and vocab, and it looks like I'm on track to continue that trend for N3. I know that Murtada created this thread to help discuss RtK, but I want to take a minute to post my vocab stats too. Here were my main decks from last year:

Minna no Nihongo 1 vocab: 1,960 cards, 93% mature (woo hoo! broke 90%!)
Genki 2 vocab: 1,115 cards, 79% mature (will be nice when I break 80%...)

I'll never add anything more to them, which is nice. I also used this one, which I continue to add to:

Out of class vocab: 595 cards, 71% mature.

Last month I created a new vocab deck for my N3 studies. I expect to be adding a lot to it this year:

N3 vocab: 237 cards, 0% mature, 69% learning, 26% unseen

My N3 vocab book says it has about 1,200 words in it (2,400 cards), so there's a lot to add. I originally started adding a lot to it, but now I'm slowing down. One of my lessons from the N4 test is that my grammar is much weaker than my kanji and vocab. I had a lot easier time retaining grammar when I actually lived in Japan - it was reinforced all the time by the people around me.

I've just started downloading and creating my own grammar decks. I'm still experimenting, so I'm not ready to post stats yet. But I hope by the end of the month to have something worth sharing.

OK, so I finally figured out a good way to learn grammar, creating a workflow that ties together my classwork and Anki. (I just posted about it under the title "Learning Grammar"). So I decided that whenever I post my kanji here, I'll also post the number of mature grammar and vocab cards too. So to give me a baseline I have:

grammar: 199 mature cards
vocab: 3,199 mature cards

And my kanji rtk deck is: 469 mature cards

ariariari Wrote:OK, so I finally figured out a good way to learn grammar, creating a workflow that ties together my classwork and Anki. (I just posted about it under the title "Learning Grammar"). So I decided that whenever I post my kanji here, I'll also post the number of mature grammar and vocab cards too. So to give me a baseline I have:

grammar: 199 mature cards
vocab: 3,199 mature cards

And my kanji rtk deck is: 469 mature cards

Aite, we're done with the 中旬, thought I'd check in with you guys. Here are my stats:

grammar: 208 mature cards
vocab: 3,246 mature cards # flying thru this!
kanji: 478 mature (22% mature, 12% learning, 3% suspended)

Very happy with the vocab and grammar increase. RtK remains a slog for me, but I am mostly thru the N3 deck (just 33%, or 136 cards remaining).

@ariariari

I wish i had your motivation :/