I have a large backlog, and I spent the last week trying to etch at this backlog little by little. I have set the decks to be all sorted according to Relative Overdueness so that it begins with cards that I have most likely forgotten are about to forget. My desired retention was set at 95%. [image…

As you can see, there is a pattern - I perform much better with young cards than with mature cards. Is this expected behaviour :question: No, this is not intended. And does this have to do with interference from newly learned content with the old content inside my memory (like basic physiology …

No, this is not intended. Do you know of any potential causes for why is that the case my true retention for mature cards are clearly not within error range of the desired retention :question: I mean obviously backlog is one of them. Maybe I have run into a bunch of difficult cards in the past w…

You could try filtering reviews in the default Stats window to see what kind of cards may have been young when you reviewed them recently.

[image] Comparing anecdotal actual retention to FSRS average predicted retention FSRS I have this 10000 cards-sized deck. I made a custom session to review all of them at once. Before I did my average predicted retention of it was 98.2% (desired retention was set at 95%) …

Nope! I ask you to read what I mentioned again. I have noticed there is a big discrepancy in my retention for mature cards compared to my young cards. That is a different issue to when I made a 10000 card haul and seeing whether the average predicted retention was indeed accurate or not.

[image] DerIshmaelite: Do you know of any potential causes for why is that the case my true retention for mature cards are clearly not within error range of the desired retention FSRS can schedule the cards for when you are most likely to remember them, but it cannot force you to remember the…

Oke :confused: it was just a really bad day at the office then. I reoptimized my scheduler since it was a while since I last optimized (1-2 weeks?).

Both are about the accuracy of FSRS. [image] Comparing anecdotal actual retention to FSRS average predicted retention Around 80%-90% is where FSRS performs the best. We’re still trying to figure out how to fix that. We might change the shape of the forgetting curve again, for the third time, but…

Here are my stats in the past year [image] V

Backlog: Low Retention Rate for Mature Cards

Anki FSRS

Keks April 20, 2024, 3:53pm 6

You have already asked similar questions and received answers to them.

Topic		Replies	Views
Comparing anecdotal actual retention to FSRS average predicted retention FSRS	13	1325	April 5, 2024
Need help on optimizing Anki for long-term vocabulary retention Help	7	249	January 4, 2026
Is what I am doing logical (FSRS Weights)? FSRS	16	632	April 2, 2024
FSRS not achieving desired retention after 2 months FSRS	11	448	May 8, 2025
Filtered deck cards - Card reviews don't seem to go down? FSRS	8	371	July 24, 2024

Backlog: Low Retention Rate for Mature Cards

Related topics