[Userscript] Expected Daily Reviews

[NeonLotus]

Thank you! This is such a simple but useful script

[Kennis]

Any shot of this getting a new update? I think this script is really useful

[logiczsniper]

Hey @Kumirei! This is actually my first comment/post on the forums

I had an idea similar to this, and when I saw your user script already exists, I thought we could collaborate? The idea is to track the success rate in all reviews and calculate some heuristics for a very detailed breakdown of reviews not only over the next week but next few months. The main benefit is a more accurate view of your next week, but also as you say in this extension:

if the expected number is higher than the number of reviews you are comfortable with doing, then you should perhaps not do any lessons.

We could use this data to build a lot of cool features such as:

• Identify ‘hotspot’ days of reviews that may come n weeks/months down the line, so then we could warn users to only do their daily lessons and nothing extra because it will likely further the severity of the ‘hotspot’.
• Suggest days to do extra - when the expected review days are more ‘quiet’ than others.
• Display something like the GitHub contributions heatmap to quickly visualize the next few months towards the bottom of the review forecast, so the 'hotspot’s and ‘quiet’ days are also visible there.
• Help users stabilize their reviews to a particular goal; if they want ~80 reviews daily, maybe we could use this data to help them hit closer to that mark.

In conclusion I think that we could use a bit of statistics to calculate a more accurate and useful review forecast to allow users to better plan their lessons and reviews. This is all with the assumption that we could use the WK API to fetch all items in the SRS with their next review dates lol

Lemme know what you think, if this idea is flawed then I’d rather not waste my time, and I expect that more experienced users such as yourself would have some pretty good insights on this.

Thanks!

[Tensor]

This script doesn’t seem to be working for me. I’m taking a look.

[Tensor]

I have been inspired by the ideas above and have the following working locally. I am seeking feedback from the community on current progress. Please see attached picture:

UI hangs over too much, but this builds on idea of the original script while making some significant changes.*

Let’s say I do 200 reviews right now. What’s expected to show up tomorrow? Well, it’s whatever is leftover + whatever is coming up (new24h) + whatever I did today that returns tomorrow.

So, instead of one number, there are these three. If returns + new24h > 200, then you probably should lower your lesson count or do more than 200 reviews / day. I decided on 200 because that’s my max given other studies/priorities. These numbers help the user see progress is being made even if there’s a big pile of reviews.

If this looks good, I can figure out how to share it.

---

Note that “returns” is sampled probabilistically, because you may hit more or less low SRS stage reviews, but it should be good enough for an idea of how big that pile is.

*The original script cares about zero’ing reviews, counting inter-day reviews for low SRS items more than once–this is a different philosophy that is perfectly fine. Maybe my script should be published separately (?)