How to prevent users from misusing Hard? Ideas are welcome

I don’t know about other programs, but Anki has everything in plain sight.
Imgur
Windows users can get help/guidance in many programs using the F1 key.

2 Likes

If you post details of the idea on the Anki subreddit, maybe an interested developer will develop it, I think DerIshmaelite got at least 2 add-ons made that way (one by me).

I was just trying to put the box around each group of buttons so it was clear that the three to the right should be used when you get the answer right or knew the answer.

The spacing can be fixed if I spend some additional time on the css

1 Like

Adding a tutorial may fall into the ‘next,’ ‘next,’ and ‘next’ until the start. I usually do not have the patience to follow tutorials. Adding tooltips to the buttons is excellent and should be simple to add. Sometimes, a tooltip saying hard should only be pressed for incorrect answers should be okay if this will not conflict with add-ons or show tooltips (as Color Confirmation addon).

@Expertium , I am curious why you are so worried about the misuse of the hard button. I wonder how much time someone can use Anki missing the hard button without feeling something is not working as it should and looking for help.

Are you a software developer, and can you implement this in Anki? If you can, you can create pull requests with the proposed ideas not discarded by Elmes. If you are not a developer, you can group and assemble tasks (on the issue tracker) that someone could implement. I say this because I would not know the best approach if I were motivated to help you fix this right now:

  1. A pull-requests adding tooltips?
  2. A pull request adding a tutorial? (How should this tutorial be/presented to the user? What should this tutorial include? When should this tutorial be given?). Depending on how it is done, implementing a tutorial could be a complex task, and the tutorial contents should be decided beforehand.
  3. What has Elmes already rejected regarding other proposed solutions?
1 Like
  1. It was said before that dae doesn’t want to work on the review screen before more of the reviewer code is shared between all the client apps.
  2. This won’t be any useful for mobile clients as users normally do not tap and hold UI elements.

In the 20K collections, we’ve found people who use Hard+Again, Again+Easy without ever feeling like something has gone wrong.

Also, it’s more of a psychological problem than intellectual. Pressing Hard makes you feel, “Just a small mistake, I didn’t get that completely wrong”. Part of why I think 2 button mode should be done.

Also, don’t think dae will accept a PR for tooltips/tutorial. These are already rejected ideas IIRC.

I’m not good enough at coding to implement this myself. As for rejected proposals, scroll all the way to my first message that started this thread.

It affects FSRS, and it’s not something we can solve just by adding more parameters to it. FSRS can’t treat “Hard” as “fail” for some users and as “pass” for other users. It’s not designed to do that, and me and LMSherlock can’t think of any reasonable way to make it do that. Which means we need to solve this outside of FSRS - make it clear which buttons count as “fail” and which count as “pass”.
The only “solution” that I can think of is making 2 different versions of FSRS and somehow dynamically switching between them based on…something. This would be very janky and very complicated, and I wouldn’t even consider it tbh.

4 Likes

I primarily use the iOS version of Anki. How do you see this working on a touch screen device?

This only a concept. Don’t think this concept will be this as-is

…that’s a very good question :sweat_smile:

I guess moving Again further away and giving buttons some sort of colored background really is the best solution.

There isn’t enough space if it’s a small screen. I still think the best solution is giving people only 2 options: Again and Good.

2 Likes

Then this will a need of design overhaul of the whole, entire Anki environment, which I oppose that idea because the buttons are perfect enough.

This is just a concept, but I don’t think this is feasable

I don’t what “perfect” means here. 4-buttons can be misused in various ways other than what is being talked about here. Have you seen the Pass/Fail thread? Besides, it wasn’t very hard to do in AnkiDroid (we already have a two button mode). And as I’ve said before, dae is unlikely to change anything before all the clients start using the same code.

@expertium has a good idea IMO if you ignore the fact that it will look ugly. But if we do it eventually, I’d instead just get a 2-button mode being made default. More options ≠ Better.

1 Like

I get what you mean now, and I also see the same scope of that thread in here as well.
Got it from here - Pass/Fail Grading as Default - #20 by suiyuan

1 Like

Dae’s opposition comes from the fact that sometimes users will have a good interval that they think is too short or too long. In those maybe 5% of cases, we want to give people more options. My point is, most users end up misusing these buttons and in the long term it will just make their overall intervals worse. Trying to fix 5% cases will end up ruining all.

I think a Pass/Fail mode will increase user satisfaction and not decrease it. In the long term, most people will not going to be looking at “that one card with 7 day interval which was too easy it could’ve been a 12 day interval”. Overall stats matter more, and 2 button is good for that. But for that to happen we need two elements in place first,

  • More of the reviewer code being shared between clients to avoid busy-work across the ecosystem.
  • FSRS being made the default scheduler. SM-2 has this ease hell problem if you recall, and with default ease being 250%, each press of Again will decrease it but there’s no Easy button to counteract that. One advice that I used to see is “use Easy more and more” apart from low-key Anki. But this can’t be done with 2-button mode.
1 Like

I thought there were more proposals than the ones you had pointed out; as someone said it was the third time:

Well observed. In this case, each platform would require a different implementation. It could be a tooltip for desktops, and Ankidroid could be an app tutorial. As it was said, no work on the UI would be done without sharing more code between each platform, so we had to wait:

I do that depending on the day and my patience.
https://www.nihongoshark.com/post/when-should-i-hit-the-good-button-in-anki

Well, I could find someone not using hard button as it should be:
https://www.reddit.com/r/Anki/comments/qz7ab2/when_to_really_use_the_hardagain_buttons/

2 Likes

AnkiDroid had a tutorial but it was never enabled before it got completely removed. It was said that we need better design and not good instructional onboarding.

I don’t hate the design changes Expertium is proposing, but I just seriously doubt they’ll have a really great impact on people. In the link you shared, the problem wasn’t that the user was unaware of what they were doing, they simply didn’t care.

Actually, we are having another discussion on learn ahead limit, and I find similarities between how people think in both cases. Most people know they are not helping themselves by taking these sub-optimal decisions, but they continue do those things. “But I’m still getting some benefits”.

By the way, @expertium have you not seen people with 1d+ learning steps despite there being a warning? I can bet this will turn out to be the same. The fact that “Again” moves review cards to relearning state or move you back to the first (re)learning step should be hint enough on what “Again” is doing. IMO for most people this is not even a intellectual issue.

1 Like

I sure did, a lot.

I’m not sure how what you said here is connected to my idea of increasing the distance between Again and Hard and giving them differently colored borders/backgrounds.

I’m saying people will still misuse the buttons.

I bet that people would misuse buttons less frequently if they looked like this
Answer buttons design to prevent misues

1 Like