Predicting Number of Reviews

It can be hard to get a feeling for how many reviews you’ll do in the future if you are adding X amount of new items in a day. This is because it will take several weeks for the reviews to propose through Adept and Seasoned ratings.

In my experience around 10X is around how many reviews I’ll get, and that’s with a 70-80% accuracy. So if I do 10 new items I’ll get around 100 reviews, if I do 20, then 200. Right now I’m doing 5 new grammar reviews and I have approx 50 grammar reviews a day.

Additionally, I find it takes me around 1 hour to complete 100 reviews (reading or Cloze). More focussed I can do 100 reviews in 40 minutes. This is just going off of screen time measurements though. In general I can sustain around 1.5 - 2 hours of Bunpro daily depending on if I’m particularly motivated.

I’m curious what other people’s experiences are with these numbers.

3 Likes

That’s pretty consistent if you keep your accuracy rate constant.

For the record, if it helps others to make an estimate, I range everyday with an accuracy between 82% ~ 90%. Last year I was doing 20 new items a day and in the long run my daily reviews were around 300. Nowadays I’m doing 12 new items a day and my reviews are around 150.

I think the difference in our numbers lays in that timing doesn’t escalate linearly with the SRS level. I’m really an empty head when it comes to new items and fail them several times on the first days :laughing:
This means that for every new item I’m having more reviews that pump up these stats, even if my accuracy is okay thanks to the older items that I remember well.

Something I’d really like to see on this platform is a diagram showing our progress over time specifically tracking when we pass or fail each element. This would help us better understand when we are more prone to forgetting specific grammar points in the long run.

1 Like

Yeah, 10x to 13x new items per day is expected. I have been contanstly adding 35 new words per day and review vary between 300 - 400 and did not really deviate from that.

2 Likes

About the number of reviews I’m also curious… Are you all changing the setting to fetch new pending reviews from hourly to daily? Or keeping it at its default?

I’m using the hourly one to force me to review new words several times a day, but I’m not sure if you usually end up with more reviews one way or another. I feel that waiting a whole day would make me fail more, with reviews piling up pretty quickly.

I have it set to daily. Beginner items still showing up based on the hour, so if I’m failing an item I’ll see it multiple times a day.

Daily just means that I’m not getting a constant trickle of new items through the day and it’s a lot easier to “complete” a day. And that’s more motivating.

Oh wait, so beginner items keep appearing several times a day even after setting the review stack to update daily? I was using the hourly one just because of those beginner items :sweat_smile:

Thanks for the heads up, now I’m definitely trying the daily setup.

There’s a tool that someone made for this for WaniKani:
https://observablehq.com/@indutny/wanikani-estimator

It looks like it could be modified to work with Bunpro.
I don’t have an account (or time to look into it at the moment), but I thought I’d mention it here in case anyone else fancies giving it a go.

1 Like

Bunpro’s SRS has 11 intervals. This means that every lesson you take generates at least 11 reviews in the future. So as you keep on and on with your lessons, you’ll get 11x as many reviews every day on average

1 Like

Yes, however the time lag from Expert+ onwards doesn’t kick in for a long time and it doesn’t account for accuracy rates. But it is true that this heuristic lines up with other people’s experiences.