-
Notifications
You must be signed in to change notification settings - Fork 11
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Take "hard" and "easy" frequency into account #14
Comments
If we do add these, it would probably also make sense to re-introduce the ease penalty for lapses (perhaps make it optional). Otherwise the predictions likely wouldn't make sense. Right now our model basically assumes that "hard", and "easy" ratings balance each other out to produce the specified average ease. With the addition of the settings above, users would be able to evaluate how certain review behaviors affect their long term projections in a more detailed way (e.g. whether it makes sense to only use "good" and disregard all other ratings, as is sometimes recommended). |
Ah, scratch the first part in my previous comment. Completely missed that we do take the ease penalty into account already. |
This will be useful. Adding 'hard' and 'easy' can be easily fetched from the database, in fact, the SQL execution already fetches these but doesn't use them. A few points to consider when implementing this:
|
Yeah, I think a single spinbox would be best. And it might be sufficient to look at young and mature cards only. That's where both hard and easy buttons are consistently available, and that's where they likely have the biggest effect.
I think one way we could go about this is to have an advanced performance settings pane that is collapsed/disabled by default. On enabling it, the settings in the simple preferences pane would be disregarded, and users could instead specify fine-grained percentages each for "again", "hard", "good", and "easy". Again, it would probably make things much easier for us to only consider cards in the review state since not all answer buttons are available in other states. If we do include other states, we would have to map ratings to the options actually available (e.g. counting "hard", "good", and "easy" all as "good" when only "good" and "again" are available in that simulated repetition). |
A nice addition would be some way for the simulations to take "hard" and "easy" ratings into account – either by user-configurable count and/or by looking at the average distribution of ratings in the review history up to that point.
This would allow to draw conclusions on some of the most heavily debated topics concerning deck settings like the idea of an ease hell, etc.
(Thanks to eotini23 over on Patreon for the idea!)
The text was updated successfully, but these errors were encountered: