I was wondering how many items we should add to SuperMemo per day. What average worked well for you and your friends when you were in high school?
I was a heavy SuperMemo user for about 2.5 years. In that period, I added 6,000 items to my English collection. Therefore, my daily average was about 6 items per day. The typical scenario was that, every few days, I would sit down and add anywhere between 10 and 50 words and phrases to my English collection. In addition to that, I added an average of 4 items per day to my German collection.
Doesn’t sound impressive at all, does it? If there ever was a conference for users of spaced repetition software (SRS), I think 6,000 items would get me laughed out of the room. I personally know people who have memorized more than 20,000 English items. My friend and ex-partner at Antimoon, Michal Ryszard Wojcik, added twice as many items as me.
Even though my achievements in SRSing were modest (probably due to laziness), it worked out quite well for me. By third grade, my overall English skills were better than anyone else’s in my class, including Michal’s. While he was busy reading books, browsing his Webster’s dictionary and adding items to SuperMemo, I was watching American TV and trying to seize every opportunity to converse with native speakers. He had a clear advantage in advanced vocabulary, so he was much better prepared, for example, to read Lord of the Rings in English — most likely better prepared than the average native speaker! But in more everyday situations, advanced vocabulary didn’t matter. When it came to talking to people or watching a movie, I was much better. (He began to catch up with me only after he
began talking in English.) I believe a big part of my success was that I spent less time working with SuperMemo.
That doesn’t mean you should spend less time on SRSing; it just means that in my particular situation, more SRSing would have meant less listening, which would have been harmful to my overall English level. The “general” lesson would be that if you overdo SRSing at the expense of other activities like listening and speaking, your English skills will become “unbalanced” — that is, skewed towards written English.
How many items to add is a difficult question. I am certainly not saying that 6 items per day is some kind of golden number. For one thing, Michal and I were adding definition-word items. If you’re adding sentence items (and you should be), you can probably add twice as many items in the same amount of time.
On the other hand, our SuperMemo experiences happened before the Internet (gasp!). English-language input was not a click away as it is now, so it made sense to review what we had with SuperMemo. If I were learning English today, I would never add many of the words that I added to my collection: why add them if I’m likely to encounter them again on some Web page within the next month?
That doesn’t mean that I would add fewer items. Instead of adding vocabulary from books, I would add things like common mistakes (especially ones I’ve made myself), less frequent grammar structures (e.g. conditionals and the present perfect tense), rare-but-useful vocabulary, and phrasal verbs. I would focus on useful stuff that is hard to learn from input alone. All this knowledge could easily add up to thousands of items.
So instead of giving you some made-up figure, I will just say two things. First, you should not focus on SRS so much that you neglect other activities like listening, writing and speaking (at least not for long periods of time). Second, you should be selective. You should only add things that are (1) useful and (2) hard to remember.
The “hard to remember” part depends on the amount of input you’re getting. The more input you get in your target language, the less SRSing you will need. There’s no point in SRSing things that are kept in your memory anyway through regular exposure to the language.
The amount of input you can get depends on the target language. If you’re learning English, you will have no problem finding a wealth of interesting input on various difficulty levels, so you will get more “natural reviews” and will have less need for spaced repetition software. But if you’re learning a language like Japanese or Russian, it will be much harder to find input you actually like. (I mean, English learners can watch South Park and House MD — what can Russian learners watch that’s equally good?) So, with a less popular language, you will generally get less natural input and will have to make up for that using artificial input in the form of SRS items. In other words, more things will be “hard to remember” and worth adding to your SRS.
If the number is, for example, 20 … what happens if we encounter 20 new useful/important words or expressions before we read 10 full pages? Do we keep reading and ignore the new words?
Or for example, what happens if we learn 20 new important words/expressions while watching TV. Do we not read for that day, or do we read and not add the new words to SuperMemo?
As long as you add only (1) useful and (2) hard to remember things, and you don’t neglect other activities — especially listening, writing and speaking — I think you should use whatever system is most enjoyable.
What I did when I was in high school was write words down in a notebook while reading or listening. Then, whenever I felt like it, I would fire up SuperMemo and add some things from the notebook to my collection. I used a notebook because I didn’t want to add items while reading or listening, and I didn’t want to feel as if I had “missed” something. The disadvantage of this system was that it sometimes led to 20 pages of unadded items. It felt like a backlog of things to do and made me feel like a bit of a loser. Certainly if I could start over, I would be much more selective when writing down things to add to my SRS.
A more radical technique for dealing with an SRS backlog is to set a fixed limit, e.g. 2 items per page, or 10 items per hour, or something like that. If you think there are 4 interesting phrases on that page you just read, that’s too bad — you have to choose the two that you think are most useful and hardest to remember. I’ve never tried this system — in fact, I just came up with the idea — so take this suggestion with a grain of salt.