When Does Anki End For You?
Anki is a program that takes a massive amount of info, breaks it down, and spreads out over time so you’ll never forget. Studying with it is fine for most people. You accept the daily reviews it spits out for you every day. But as with any study tool, you want to know when the need for it will end. Many people are worried that they will have to do it forever. They will be slave to Anki. And what would be the point of that?
I have used Anki for 7 years now. I stopped adding new cards to my own personal deck of like 14-15k cards (including RTK mod) about 3 years ago. And what’s my daily review count at?
20-30 cards a day. That’s it.
And still going down. Doing it is the time it takes to brush your teeth. And that’s what I do. I do Anki while brushing my teeth (seriously).
Why not just stop completely?
Because I have nothing else to do while brushing my teeth? Maybe. But for the ultra minor effort required, I am able to upkeep something powerful. Sure, if I dropped Anki completely, things wouldn’t change that much.
But many cards are in the intervals of 12-15+ years. Which means they are essentially gone. And in a few more years, my reviews will drop towards near 0. So why not? Anki has been there for me. The least I can give it is a few seconds a day?
What about new or unknown words I come across?
I look them up online, sometimes, just like I would in my own language. I find the urge to look up a word I don’t know while doing something comes only in one of two scenarios:
1. The word is vital to understanding whatever I’m reading
2. I’ve seen the word multiple times and it’s causing me to wonder what it means.
I don’t put it into Anki because if it falls in these scenarios I am very likely to remember it without even bothering with Anki.
And other unknown words? You can’t know everything. And that’s fine. In your own language, you often skip words you don’t know unless they fall into the above. And that’s fine for Japanese too.
So Anki winds down. Anki ends. Your way to approach new knowledge changes. Life moves on. And things are good.
Founder of Jalup. iOS Software Engineer. Former attorney, translator, and interpreter. Still watching 月曜から夜ふかし weekly since 2013.
Totally agree. The falloff rate is much faster than you’d ever believe (once the new cards stop). I’ve been at the JALUP RTK for 22 months, with the ease turned down to 130%, and I’m at 40 cards a day or less. I have less than one J-E card a day.
Of course, The One Deck, at 15,000 cards, is much heftier, but it too will get there.
And just like you said, why stop? As it falls off, Anki becomes even more bearable.
Yup, the message for everyone to keep with them is “it gets better.”
Title is kinda misleading. I thought Anki died or something.
This wasn’t intentional, but looking at it now, I realize that it sounds that way. Changed it!
I thought so too. I was like, “Oh no! Everyone will be so crushed!” Though thought that couldn’t really be it.
A while back just out of curiosity I wrote a little program to simulate Anki use over time. Here’s a plot showing how many reviews someone would have every day over ten years, assuming they add 10 cards a day, have a 90% retention rate, use the default interval modifier (250) and stop adding at 10,000 cards. This isn’t quite realistic, in particular I’d imagine after some point retention rate goes to pretty near 100% because everything is being constantly reinforced through daily exposure. Anyway, the most striking feature is what Gregory pointed out, the daily count drops precipitously once new cards stop.
(As a disclaimer I’ll note that I eventually decided that Anki wasn’t for me, but I hope folks who are using it find these results interesting!)
I can actually provide a sort of (partial) real version of that graph.
As you can tell my reviews dropped significantly starting about two months or so ago, when I finally reached 10000 J-J cards (at which point I severely reduced the number of new cards, though I haven’t quite stopped).
As for the “striking precipitous drop”, well… to a first (and fairly decent) approximation the decay when you stop adding cards just follows the shape of the decay for a randomized card, and it’s not hard to see that that is approximately an exponential tail decay, which is what Octonion’s picture shows… I suppose a perk of being a mathematician is that I knew this the day I started using anki without having to do any simulations….
GRAPHS!
I love ’em.
I’m currently using Anki JALUP RTK, a self-authored J-E, and The One Deck for Japanese and building out decks for Java, JavaScript, PHP, and generalized algorithms. This simulated graph verifies (semi-empirically) what I keep telling myself. It get easier.
Thanks for the simulated data. :D
How many reviews would you recommend doing at the height of your new vocabulary acquisition period? I’m going through The One Deck now and I’m not sure what a reasonable rate of exposure is.
Think I remember him saying something about 20-30 being a good number and then dedicating the rest of your time to immersion. Could be wrong though.
85% completed JALUP advanced 0.0!!! So excited
I agree with this. I think 25 a day is a solid pace. You would do 10,000 sentences in 400 days at that pace.
Wow, 25 seems like super power-leveling speed to me. 10 new J-J cards, plus ~30-40 reviews, takes me ~45 minutes on average. That’s just over 5 hours/week, plus another 5 for RTK (which I’m still trying to finish up), and another 5-10 for active immersion, as time allows. That works out to ~2-3 hours a day, which I feel like is already fairly intense.
Of course, everyone is different and it’s entirely possible that 25 is the right number for you. You should try it and see how it goes. If it feels like too much, scale it back gradually until you reach a comfortable pace :)
Heh, admittedly I may be an anki addict. I definitely prioritize it over my immersion, which has its cons (as Alexandre pointed out in a great article here: (http://japaneselevelup.com/listening-immersion-doesnt-work-for-you-or-does-it/).
Of course it’s different for everyone, but I don’t think 25 is too unreasonable. I just finished a stretch today where I pushed 50 a day for 20 days to crank a quick thousand out, and while that’s definitely an unsustainable pace for me (I had a few free weeks in summer to devote more time) I have heard of much crazier paces to the tune of 100+ per day (especially on RTK and J-E). Just look at Alexandre’s graph above: he went at a pretty brisk pace.
Just to add to this, the amount of new cards I used to add varied greatly. There were times when I would add 5-10 cards a day, and there were times when I would add 30-40+ a day. Sometimes I would just save up new additions for weekends.
Find what works for you right now, but feel free to adjust to a pace that matches your goals. You’ll find that your pace will vary a lot as you move forward.
I didn’t use Anki for nearly as long as Adshap, but while it had an enormous impact for a significant period of my initial use of it, eventually I just… grew out of it. I stopped using it. I can read most stuff, and when I don’t know a word I treat it exactly like described above.
I doubt I’m anywhere near Adshap’s level, but I feel like Anki just wasn’t actually improving me very much any more. At this point, what I feel I need to do is just keep on using Japanese.