Hello Koohii. I started learning Japanese again after a two year hiatus. I'm breezing through RTK1 and will be done soon enough. After that I'm not sure what to do. I have an old deck with mined sentences with over 700 reviews pending. Should I do those reviews, or should I start a new deck? I'm not sure the old sentences have much of an "impact" to me anymore for them to be remembered, and I'm not sure if I made a good decision in what words/grammar to remember in the first place, so I might have to suspend much of it anyways. On the other hand I know how significant the time it takes to make new sentences is.

What would you do in a situation like this?

Reviewed Wrote:Hello Koohii. I started learning Japanese again after a two year hiatus. I'm breezing through RTK1 and will be done soon enough. After that I'm not sure what to do. I have an old deck with mined sentences with over 700 reviews pending. Should I do those reviews, or should I start a new deck? I'm not sure the old sentences have much of an "impact" to me anymore for them to be remembered, and I'm not sure if I made a good decision in what words/grammar to remember in the first place, so I might have to suspend much of it anyways. On the other hand I know how significant the time it takes to make new sentences is.

What would you do in a situation like this?

Why not just get a basic Japanese reader for beginners and work through that, passage by passage? If that's too difficult maybe work through a basic Japanese grammar book.

john555 Wrote:Why not just get a basic Japanese reader for beginners and work through that, passage by passage? If that's too difficult maybe work through a basic Japanese grammar book.

Why not just not offhandedly act as though your way of doing things is vastly superior to what he wants to do?


Anyway, it depends. Do you care about those sentences? Are they easy to get through? What did you make them from?

If it's mined from something interesting you might as well keep them. If it's not, you might as well go find something that's interesting and mine from that instead, if that's how you plan to continue. Realistically - if you remember anything from a two year old deck you don't need to review it, and if you don't it's basically the same as starting again so unless it was a really good deck you might as well make a new one learning from your past mistakes (and if it was so good, why did you stop?).

I think worrying about which words and grammar are good to remember is a waste of time - you are in no position to judge what is good to remember or not, so you might as well learn everything that crosses your path (and is easy and/or enjoyable).
There are tools out there to automate sentence mining, so if it takes you a long time to make cards you should look into those.

If it were me I'd review the old sentences. You often end up remembering more than you thought, plus it refreshes where your learning was, and even if you decide to head in a different direction than before, you still have a solid starting point. And 700 isn't that much if you're review consistently.

Aikynaro Wrote:

john555 Wrote:Why not just get a basic Japanese reader for beginners and work through that, passage by passage? If that's too difficult maybe work through a basic Japanese grammar book.

Why not just not offhandedly act as though your way of doing things is vastly superior to what he wants to do?

It's not "my way of doing things". Working through a textbook systematically is a basic method followed by many.

Sure, but it's like me going into a thread asking which textbook to choose and saying 'why don't you just mine sentences and put them into Anki?'. It's rude and doesn't help.

edit: Now I think of it, I've done the same thing before, in one of your threads no less - and was rightly shot down by Tzadeck for it. So, y'know, sorry about that. But please don't do it anyway.

[RawToast]

Quote:If it were me I'd review the old sentences. You often end up remembering more than you thought, plus it refreshes where your learning was, and even if you decide to head in a different direction than before, you still have a solid starting point. And 700 isn't that much if you're review consistently.

+1. I reviewed ~500 of my old core sentences and found the process to be beneficial.

I'm with Aikynaro on this one: you'll remember basically nothing and so you might as well do whatever you feel like doing. Nothing kills my motivation quicker than reviewing old decks I no longer care about so I definitely would start again, but you probably will remember things slightly more easily when working with previously studied material so I wouldn't say there's a right answer, especially if you plan on returning to whatever material was the source of those sentences.

If you do decide to work through the old deck I'd recommend you suspend everything and treat unsuspending as the equivalent of adding new cards. You can mix them with actual new cards or work through the whole deck first but the 700 due cards figure is meaningless at this point.

P.s. I think people underestimate the importance of seeing words in multiple different contexts. Retreading old ground isn't always the best revision technique, let alone the best way of making progress.

EratiK Wrote:If it were me I'd review the old sentences. You often end up remembering more than you thought, plus it refreshes where your learning was, and even if you decide to head in a different direction than before, you still have a solid starting point. And 700 isn't that much if you're review consistently.

This. It can't hurt to see how it goes with the old sentences. I think you'll be surprised to see how much you've retained. I would start them off at a high interval though - sentences don't ever need to be on a 1 day interval.

Thank you all for your input. I'm going to have a go at that old deck, see how it goes and then decide whether to keep at it or not, can't knock it 'til you've tried it, right? I'll be back with updates for whoever might get into this same dilemma.

I should've added, I'm not a beginner, so beginner textbooks are out of the question. I might go for an intermediate/expert textbook later, but for now what I'm looking for is something interesting to review after RTK1 (again).

The sentence deck I'm talking about is mined from manga I read, most of which I didn't finish. I did like each sentence and liked reviewing them, but I'm not sure that'll still be the case. If I needed sentences whether they hold my interest or not I do have an old Core6000 deck with even more pending reviews.

If you do find that retreading the old mined material is killing your old material, though, then feel free to just suspend those cards or split them out into some kind of archive so you can start fresh. Anything important enough for you to have to remember from those old cards will come up again in future reading at some point, and the stuff that doesn't come up again probably wasn't very important.

That's my advice for mined sentences, though. If it was a more structured deck that was built to be highly efficient then I would advise going back through it to review. Mined sentences have a lot of duplicate material, and so jettisoning a lot of them isn't too big of a deal. You'll re-add the important material again in the future anyway.

A question for the original poster: Are you doing just flashcards with random sentences in random order? Or are you doing any reading of prose passages with connected sentences? I'm just curious.

Random sentences in IIRC random order. If two sentences are connected, it's a coincidence.

Well, I feel silly now, the 700s figure I got somehow counting new cards, that obviously I'm not going to review just yet. I'm not sure how ran into that number, but the actual ammount of reviews was 478, much more manageable.

After playing around with the settings, what I did was set Anki to review cards with the shortest intervals first, then sit down and review as much as possible with a suspend-happy mindset, for a total of six hours in two days. Back in the day I used to have the failed interval set to a .1 multiplier. I had doubts about it as the failed intervals seemed too high (> 10 days for low-mid interval cards) so I lowered it to .03, still having doubts about whether to set it above 0 or not. Well, the following weeks worth of reviews will answer that question.

* Retention rate for mature cards: 27% for day 1, 40% for day 2 (as expected, it was easier to remember less overdue cards). Global mature is now 63% (it was over 67% before I started)
* Despite those retention rates, I was surprised in seeing how much I retained, as some of you predicted (mostly readings, meaning not so much).
* Day 1 average time per card: 27s. Day 2: 22s (global time: 15.7s). It was difficult to get to understand the meaning of each sentence, due to grammar and having to look up meanings in dictionaries.
* I realized I'd read way more than I remembered, and the cards are as interesting as they were back then.
* Some of the cards that were n+1 suddenly turned n+4 or so, I suspended many of these.
* I thought not remembering context would be a problem, but the image at the back was enough to remember its context.
* I learned where I'm standing in Japanese (need to somehow improve grammar and word definitions, I'm still not in a good position to judge whether a word is frequent enough to be worth studying.)

Thank you all for your input. All in all, I think reviewing that old deck was a positive exercise, I'd recommend it to others in a similar situation.