In order to give users a nudge in the right direction, we could display something like this (all in one place, not scattered across deck options):
Your learning steps should be shorter than 1d / Your learning steps are configured correctly
You are using default parameters, click “Optimize” / You haven’t optimized parameters in a while / You are using optimized parameters
(I’m unsure how to handle that for empty presets though)
You haven’t used “Compute Minimum Recommended Retention” / Your desired retention is below the recommended minimum / Your desired retention is not below the recommended minimum
(please don’t bikeshed the exact wording, at least not yet)
This should be displayed under “FSRS” in deck options.
This would tell users what they need to do to configure FSRS without them having to read the manual and this is much simpler than making an interactive tutorial.
Perhaps it could be checked and mentioned when opening up the statistics page, since that’s what people do when they’re interested in their progress and potentially seek to optimise it
Well, if someone never checks their deck options, I don’t think there is anything that can be done. Other than telling them “Go use Quizlet or Remnote or something”.
As a sidenote: If I install a new OS, app or similiar or get another device, I always check every setting that exists there. Apparently most people don’t, though.
So having something like that in the deck options (which essentially is deck settings) will probably be missed by most users, especially if fsrs is on by default and people won’t have to go to the deck options anymore to turn it on.
I don’t think we need to worry about a checklist like this reaching people who have never looked at their Deck Options.
Your learning steps should be shorter than 1d – If they haven’t gone to Deck Options, they haven’t changed from the defaults, so this will never be an issue.
You are using default parameters – FSRS with default parameters is pretty darn good, so I don’t think we need to chase folks down to alert them about that.
You haven’t used “Compute Minimum Recommended Retention” – Again, that means they are using 90% DR. If their CMRR is above that, isn’t there a bigger problem? Is it even possible for CMRR to be above that?
What about big red/yellow triangle next to deck name/due cards count in deck browser, and with redirection to given issue? This will grab attention not only of users with bad settings, but also of these who never opened deck settings.
Fascinating! It’s hard to imagine what confluence of factors leads to 90% DR falling into the “work harder to learn less” zone. Has CMRR been run on the test data set? What’s the range of results you get out of that?
The problem with CMRR is that we don’t know the “ground truth”, so we can’t compare the output to that. Which means that any improvements of CMRR are based solely on common sense, rather than on minimizing/maximizing some metric.
Nope. It’s calculating workload/knowledge specifically when using FSRS scheduling. The user had different scheduling, so his workload/knowledge would be different
Idk what the numbers say, but I guess if it helps to study by at least 20% more efficiently (which I believe it does), then I guess it is important enough.
I’m not aware of anything showing that changing these 3 things result in a 20% increase in efficiency. No one is questioning the value of FSRS overall – but I’m not sure there’s value in featuring this checklist so prominently, like some sort of urgent alert, in the Decks list or Browse window.
The point I was trying to convey is that, optimizing parameters affects scheduling which makes cards due earlier (or later). Inefficiently scheduled cards lead to higher “again” count (or too frequent good count) which extends your study session making it less efficient. So default vs personal parameters could make you study less since its more efficient (properly scheduled). I do believe this gain is significant enough isn’t it? If it would have not been making the difference, why it would have been recommended to optimize it then?
(please correct me if I am wrong, I am new to this)
Unfortunately this part of the UI has not made the svelte migration yet.
I would be wary of putting too much effort into something that might have to be thrown away and rewritten in the near future.
This looks great! Regarding investing time and effort into it, I want to hear @dae’s thoughts
EDIT: on second thought, this placement creates a problem since settings aren’t adjusted for decks, but rather for presets. So if someone is confused about the whole deck vs preset thing, this can confuse them further.
I realise that may be an issue, but if you are confused by decks vs presets you are probably not going to do any better understanding the difference in “Options”.
My idea behind this way of displaying it is that it’s a screen everyone will frequently see, so we will catch users who never look in Options.
If you provide “Quick Fix” buttons they may not even have to open Options for many issues.
You could also make the wording more clear if you think it is a problem. e.g. “The learning steps in the Preset this Deck uses […] changing this will also affect 3 other decks”
Edit:
It also allows you to show deck-specific checks e.g.
/!\
Your retention for this deck is significantly lower than others using the same preset.
It might be a good idea to move this deck to its own preset See the manual for more information…
/ Your learning steps are configured correctly 
Has CMRR been run on the test data set? What’s the range of results you get out of that?
