Looks nicer than before.
4 Likes
I assume “Additional new cards” means “Cards that are yet to be created, on top of already existing ones?”. If so, I’m not sure what to call it to make it easy to understand.
Personally, I think the simulator should just use real deck sizes.
Do you mean I should remove “Additional new cards” (it means it will be equal to zero forever)? I think it’s useful and the simulator add-on also has it:
2 Likes
I’d say yes, remove it (I’ve said that real deck sizes should be used a whole lot of time, maybe too much), but I also wonder what other users think.
I would leave “Additional new cards”, but make it zero by default.
A lot of people, me included, are adding cards to decks regularly and it may affect the simulation.
5 Likes
Any ideas on what to call it? “Additional new cards that you will create in the future” is a bit of a mouthful
Why not leave “Additional new cards” as in the simulator add-on and explain further in the help tooltip?
I’m pretty sure most people never discover tooltips.
1 Like
I am also convinced that most users will not scroll options menu far enough to even reach simulator.
2 Likes
There are people who still don’t know what FSRS is.
How is deck size even affecting the simulation? If it’s a simulation of future reviews, whether there is 10,000 more cards in the deck or 20,000 cards shouldn’t matter, unless you expect yourself to stop doing new cards past a certain number of cards.
I feel this can be smoothened a little bit, by default or by using a slider (and with load scheduling, the real daily load will also become much less volatile than this).
Yeah, and Liad Balancer will make the simulation more complex. Poor LMSherlock, this just gets more and more complex.
Anyway, regarding deck dize. If deck size/new cards limit is <= days to simulate, that means that the virtual user will learn all cards by the end of the simulation. If deck size/new cards limit > days to simulate, the virtual user will still have cards that he has never reviewed by the end of the simulation.
For example, if deck size is 10 000 and new cards limit is 10, it would take a 1000 days to review each card at least once. So setting “Days to simulate” to 365 days vs 3650 days would likely make a big difference, as in the latter case the workload will eventually go down.
1 Like
I suggest adding Knowledge/Desired retention graph. Thoughts, everybody?
I have seen people querying what desired retention should they use for an exam next month or next week… are such graphs useful for these situations?
I personally always use what CMRR gives me, even for exams so I don’t know.
edit: I meant total “knowledge” at the end of simulated days, if it wasn’t clear already.
One more suggestion: This will be useful if you’re studying for a particular exam, a Just will be more clutter. Ignore.simulation cutoff date that works like days to simulate except you select a future date.
Any graph with desired retention on the X axis will be computationally expensive, and LMSherlock doesn’t seem to like that. If you want a crude estimate, I’d say try “Compute minimum recommended retention”, measure how long that takes, and multiply it by 5. Though it depends on deck size. Larger deck size = more card histories to simulate.
What’s their to not like about it? It can always be made optional as already suggested. Let us hope he realizes these graphs are useful for people and worth the trouble of waiting for sometime.
Thank you for this wonderful work.
Btw is there anyway to turn time on y axis to number of cards?
@L.M.Sherlock well, this is a bit too late since the PR has already been merged, but I think being able to switch between the number of minutes and the number of reviews in the simulator graph would be good
1 Like
I don’t know whether I have enough time to implement it before the next release.
1 Like