Maximum lifetime spaced repetition cards

Maximum lifetime spaced repetition cards

id:: 625b21c8-94a2-4bfa-a5e9-4637c4155833

I think it’s better to think about this in terms of how many cards can you add per year and still stay at roughly the same daily time commitment to spaced repetition. Also, I’m starting this consideration from a state where I already have about 4,600 cards in my spaced repetition system already (I use Anki)

For the last month, I have been Net matured spaced repetition cards cards at about 3.3 per day. and learning cards at 2.9 per day. I happened to average adding 1 card per day over the last month. While my review times were all over the place, the were trending slightly down below the 15-20 min max I want to have for a lifetime commitment (I think my ideal might be 10 minutes: How long should you study spaced repetition every day?).

See detailed stats in this pdf from 2021-12-25.

All this comes out to say that I think I could add ~2 cards a day and maintain a manageable daily review time.

Actully, I should just try using the Anki Simulator add on to see what the maximum new cards per day value is before increasing review time.

![Simulations](../files/_Pasted image 20211225221134.png)

My current average time per card is just under 8 seconds. This means that 75 repetitions can be expected to take me 10 minutes, which is the max time commitment I would like. I can stretch to 15 minutes for 112 repetitions.

We can see that 4 new cards a day rarely jumps above 60 repetitions (except for the near term bump I have because I missed some reviews). And 8 new cards a day rarely jumps above 100 repetitions. To be safe, I think the 4 new cards a day is a fair baseline, maybe stretching to 6. This makes it ok to miss days and still be able to keep most days under 10 minutes of work.

So taking my 4 cards per day baseline, that means that through spaced repetition I can retain about 1,460 atomic pieces of knowledge per year.

Of course, this isn’t the only knowledge that’s retained, but I can supplement my existing knowledge and learning with this more brute force knowledge retention about 1,460 times.

I’ll try adding cards more consistently moving forward to see if this hypothesis actually plays out.

I’m realizing I didn’t really address the question posed in the title, since it really needs to consider how the initial cards are added (most of my 4,600 were added to my system large groups as prebuilt decks). It would probably be a different story if I started with 0 cards and added 4 new cards a day. In that scenario, it might be possible to increase the load without increasing the daily review time (4,600 cards would take over 3 years at 4 cards per day).