Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Spaced repetition systems have gotten better

domenic.me

Post reply on HN
Print view View original

321–330 of 538 posts

Re: Spaced repetition systems have gotten better

#321

Post by ralferoo »

I'm not ready to share a link to my actual implementation yet, but I've been working on an SRS system for Chinese and been using it as my daily driver for about 3-4 years now, after previously using Anki and getting frustrated with how reviews pile up after a couple of days off.

I've done lots of tweaking to the algorithm over the years to make it feel like I'm less surprised by the scheduling, and less like a slave to it. One very stark difference between mine and Anki is that I have a large number of "overdue" cards, but the system still prioritises when to show me the overdue cards with quite a few different metrics based on how overdue it is, how new it is, how long the current interval is, etc. So, like Anki, I still just double the interval for correct cards, but for incorrect cards, the reviews are repeated same day until they're correct, and then the interval is reduced a lot more than Anki. So, the cards then become overdue sooner, but because the scheduling of overdue cards is better, they get pushed later if your overdue queue is too large, and sooner if you've not got anything more useful to review.

FWIW, my typical session is 40 minutes per day during my daily lunchtime walk, and I'll get through about 150 cards in that time. If I'm on a long train journey, I'll often clear out double that or more, but the disaster situation of being on holiday for a month might leave the queue with a couple of thousand extra cards, but they never seem unmanageable. Even after a 2 month break when I was travelling last year, and only doing reviews on flights and trains, I'd definitely forgotten some words from not reviewing at the appropriate time, but the percentage of totally forgotten cards felt better than I used to experience after just missing a few days with Anki.

One thing the article mentions that I don't massively concern myself with is desired retention. I'm not sure I'd want to express it as a target percentage, but I've definitely been thinking about how I want to change things to deprioritise stubborn words without just suspending them or deleting them. I definitely find that having them keep showing up, so I might see a pattern of them wrong twice each day before finally getting them right, after a few days of that they do usually suddenly stick for good. But sometimes I look at the word and think I don't really care if I remember it or not.

Re: Spaced repetition systems have gotten better

#322

Post by indolering »

kqr wrote: post #318
aaldrick wrote: post #232

I see a lot of discussion about SRS, and I think most can agree they have improved. What I would like to see covered is a more vague area, but almost more important: It’s the space in between reading/understanding something and the SRS. There are almost no standalone tools dedicated to creating flashcards easily from existing programs (web browser, PDF readers etc.) into popular SRS (Anki, Mochi etc.). They should wo…

> There are almost no standalone tools dedicated to creating flashcards easily from existing programs I think this is a common misunderstanding. Half the benefit of SRS comes from working out what the flashcards are. You have to circle around a concept, look for similarities, differences, examples, generalisations, properties, etc. Is it hard work? Yes. Does it help understanding? Massively. This is also a very diffi…

The technical term is "elaboration" and it is key to long term retention.

Re: Spaced repetition systems have gotten better

#323

Post by bspammer »

MostlyStable wrote: post #224
TechDebtDevin wrote: post #204

Pro Tip , if you're using LLMs to learn, create an MCP tool for them to insert Anki cards on topics you're discussing in a csv on google drive, then sync that with you anki decks on your phone. This was a game changer for me and working with LLMS, while I still think they make you dumb, and we essentially use them to offload critical thinking (almost only find myself using them when tired lazy, and just cant), if you…

I created a python script that checks my anki deck for the cards that I'm scheduled to review the next day and asks an LLM to generate new sentences for the cards, so that every time I see them, I see them in a new context. I did this because I realized I was hitting an issue where I theoretically "knew" a word (would get it always correct on the card), but wouldn't always recognize it in a novel context. I'm hoping…

That’s a really clever use for LLMs, I have the exact same problem with my Anki deck. Do you store all the historical sentences for each card or is it just a destructive overwrite each time?

Re: Spaced repetition systems have gotten better

#324

Post by _Algernon_ »

cjauvin wrote: post #54

What I find interesting about spaced repetition is the underlying thesis that raw memorization, in certain contexts, is playing a more important role for learning than what some modern education ideas would make you assume. In mathematics or programming, for instance, there is this idea that understanding a concept is better than memorizing algorithms or recipes (derivation methods for instance). But spaced repetitio…

It makes sense to memorize elementary operations that are reused frequently because that frees your mind to focus on the "higher level of abstraction" of learning. You probably learned the multiplication tables by heart before you were asked to do more complicated multiplication problems for example.

Re: Spaced repetition systems have gotten better

#325

Post by 0xDEAFBEAD »

Alex-Programs wrote: post #273
0xDEAFBEAD wrote: post #257

Earlier quoted context omitted.

Maybe an 80/20 approach where you only create flashcards for the 20% of knowledge that's most useful? E.g. for a language, you could create 1000 flashcards for the 1000 most common words which allow for basic real-world communication?

My personal approach with A levels was to strictly learn content through classes, then organise things (which teaches you a lot on its own!) and make my own flashcards. Then I used the flashcards to keep it fresh till exam season. It was crazy, during revision lessons everyone else in the class would be going "uhhh I have no clue, that was two years ago" while I'd just know it. I have tried different approaches, incl…

I appreciate you replying to my comments!

I confess I'm interested to hear your thoughts re: the usefulness of SRS from a more holistic perspective.

Gwern writes:

>...if, over your lifetime, you will spend more than 5 minutes looking something up or will lose more than 5 minutes as a result of not knowing something, then it’s worthwhile to memorize it with spaced repetition. 5 minutes is the line that divides trivia from useful data.

https://gwern.net/spaced-repetition

My sense is that there are very few facts I will spend more than 5 minutes of my life repeatedly looking up. And even then, many of those are facts that I will naturally end up memorizing regardless of SRS, since I'm using the info so often.

I understand the utility of SRS for test takers or language learners. When Google is impractical or unavailable, memorization makes sense. But for everything else -- why not just Google it?

Re: Spaced repetition systems have gotten better

#326

Post by pessimizer »

uselesswords wrote: post #130
amluto wrote: post #85

My personal peeve about Anki: I don’t like its data model. It seems to me that there ought to be collections of notes (which might be things one would download or generate with an LLM or make yourself or share with friends or students). On top of one or more collections of notes are the sets of cards to want to learn, and they can derive from the notes. This includes, roughly, templates plus some concept of which car…

It’s amazing how every single point you’ve made here is wrong as the other commenter already dove into. On top of that Anki is one of the best documented pieces of open-source software I’ve messed with. If you’re able to program, ChatGPT can basically handle any task you want it too, I data mine the sqllite database regularly for my own insights.

Actually, every point was right and the data model is terrible. I've been using it for years. The other commenter just mentioned a list of things that the software does do and basically said "isn't that good enough for you" a bunch of times. No, it's not. Anki's concepts of flashcards, and how it stores and manipulates them, are horrible.

It's hard to do many, many things in Anki that should be trivial, impossible to do many, many things that should be possible, and the things you can do involve the types of queries being run over your entire collection that causes the app to slow to a crawl after you add about a dozen decks. And in general: I can adjust far too many things that I don't even care to adjust and probably shouldn't be adjusting, and things that should be trivial to do are impossible.

It's bad. Ankidroid is a little better, but they're also stuck with the data model.

Re: Spaced repetition systems have gotten better

#327

Post by vjerancrnjak »

aaldrick wrote: post #232

I see a lot of discussion about SRS, and I think most can agree they have improved. What I would like to see covered is a more vague area, but almost more important: It’s the space in between reading/understanding something and the SRS. There are almost no standalone tools dedicated to creating flashcards easily from existing programs (web browser, PDF readers etc.) into popular SRS (Anki, Mochi etc.). They should wo…

I’d say a missing thing is figuring out the order in which cards have to be shown.

Showing schlafen before ausschlafen, or contextually simpler cards before more complex ones.

Optimizing blindly easy medium hard and next time to show a card is probably very far away from efficient learning.

If the process is inspected, most of the forever knowledge that I acquired I heard once and internalized.

Re: Spaced repetition systems have gotten better

#328

Post by edanm »

aaldrick wrote: post #232

I see a lot of discussion about SRS, and I think most can agree they have improved. What I would like to see covered is a more vague area, but almost more important: It’s the space in between reading/understanding something and the SRS. There are almost no standalone tools dedicated to creating flashcards easily from existing programs (web browser, PDF readers etc.) into popular SRS (Anki, Mochi etc.). They should wo…

I've never used it, but my understanding is that something that does this is considered a "killer feature" of Supermemo. It has a mode that lets you read text (via a PDF viewer) and lets you create flashcards as you read in some kind of semi-automated way. Or something like that. Like I said, I never used it.

Re: Spaced repetition systems have gotten better

#330

Post by pessimizer »

Alex-Programs wrote: post #160
maximus-decimus wrote: post #111

Earlier quoted context omitted.

I don't understand, wouldn't it be worse for motivation to take longer to achieve the same results?

Spending half an hour mind-numbingly learning words through flashcards will teach you about as much vocabulary as an hour watching educational videos, but it'll be far less fun and you'll feel like it actually took two hours. Keep that up every day and you'll burn out much faster with option 1 than option 2. Now, maybe you have enough motivation for that not to matter, or the self-discipline to keep going - as I did…

> Spending half an hour mind-numbingly learning words through flashcards will teach you about as much vocabulary as an hour watching educational videos, but it'll be far less fun and you'll feel like it actually took two hours.

The first part is definitely untrue, you won't learn any vocabulary spending an hour watching an educational video, you'll be lucky if you remember one new word tomorrow. That half hour on Anki will be spread out over six months, and will teach you 20 words.

As for the second part, doing Anki is like doing through any sort of timeline that spits out random rewards and failures. I get a rush whenever I remember stuff, and I get bummed out when I forget; it's basically facebook.

I understand why one wouldn't think that with single-word vocabulary flashcards, because they are horrible to do and unhelpful. You should be running sentences, not words. Words rarely translate well, change form when they are in sentences, and often show up as part of seemingly ungrammatical set phrases.

Post reply on HN