# FSRS 5: Ideas and Sims **URL:** \ **Category:** FSRS\ **Created:** [October 28, 2024, 8:00am UTC](https://forums.ankiweb.net/t/fsrs-5-ideas-and-sims/51278 "2024-10-28T08:00:50Z")\ **Posts on this page:** 11\ **Page:** 1
**Author:** ![DerIshmaelite](https://sea2.discourse-cdn.com/flex002/user_avatar/forums.ankiweb.net/derishmaelite/32/17983_2.png) [@DerIshmaelite](https://forums.ankiweb.net/u/DerIshmaelite)\ **Post date:** [October 28, 2024, 8:00am UTC](https://forums.ankiweb.net/t/fsrs-5-ideas-and-sims/51278/1 "2024-10-28T08:00:50Z")
## Current FSRS Ideas and Simulations at work (list being updated…) 1-Seed Collection Testing > [@FSRS 5: Optimizer](https://forums.ankiweb.net/t/fsrs-5-optimizer/51229/22): > > Right now I’m just gonna wait for like 30 hours to test 1 vs 10 seeds. That stuff takes forever even on a relatively small dataset EDIT: update, I finished doing it with 2 seeds on 200 collections, some with (relatively) few reviews, and some with a lot of reviews. Note that I deliberately selected collections close to the extreme ends - either a lot of reviews or not a lot. Here is the graph of the differences as a function of the number of reviews: For each collection, RMSE(2 s… 2-Preset Size > [@Number of Cards worthy of a Preset](https://forums.ankiweb.net/t/number-of-cards-worthy-of-a-preset/50948/2): > > That’s something me and LMSherlock would like to investigate, but since our dataset doesn’t have deck/preset names, we can’t find out which cards belonged to which decks, and can’t measure how much splitting things into a million presets helps. 3-Short Term Memory Model > [@FSRS 5: \<1d Scheduling and Learning Steps](https://forums.ankiweb.net/t/fsrs-5-1d-scheduling-and-learning-steps/50242/132): > > If I figure out the short-term memory model, I will decrease the minimum stability limit in the same version. > [@FSRS 5: \<1d Scheduling and Learning Steps](https://forums.ankiweb.net/t/fsrs-5-1d-scheduling-and-learning-steps/50242/149): > > Found it: The median short-term stability of a card whose first rating is ‘again’ is 38 seconds. The median short-term stability of a card whose first rating is ‘again’ and second rating is ‘good’ is 24 minutes. 4-Sorting Order (PSG, PRL) > [@Ordering Request: Reverse Relative Overdueness](https://forums.ankiweb.net/t/ordering-request-reverse-relative-overdueness/50051/493): > > So, turns out PSG cannot maintain desired retention at the specified level. I don’t think we should continue this. I will tell Dae to just implement reverse relative overdueness. P.S. I know sorata will disagree and say that we should try more sort orders. I will not be doing that. 5-Improving Calculations for Retrievability and Difficulty [[Enhancement] Improving the function for calculating difficulty · Issue #352 · open-spaced-repetition/fsrs4anki](https://github.com/open-spaced-repetition/fsrs4anki/issues/352) [[Feature Request] Improve Calculation of Retreivability · Issue #703 · open-spaced-repetition/fsrs4anki](https://github.com/open-spaced-repetition/fsrs4anki/issues/703) 6-Influence of Increasing Collection Size on Retention (Interference - Retrograde Inhibition) 7-Influence of Time of Day (Fatigue) on Scheduling Time Intervals ---
**Author:** ![Expertium](https://sea2.discourse-cdn.com/flex002/user_avatar/forums.ankiweb.net/expertium/32/24176_2.png) [@Expertium](https://forums.ankiweb.net/u/Expertium)\ **Post date:** [October 28, 2024, 10:19am UTC](https://forums.ankiweb.net/t/fsrs-5-ideas-and-sims/51278/2 "2024-10-28T10:19:32Z")
1. Here are the results so far ![image](https://us1.discourse-cdn.com/flex002/uploads/anki2/original/3X/a/f/af3f6053c5a75367f859aaab2db12173e6019045.png) I’ll finish this test, but it seems pretty clear that it’s not worth it even for people who don’t have a lot of reviews, and **definitely** not worth it above 5k reviews. 2. Still waiting for the dataset 3. That’s Jarrett’s territory, not mine. I don’t have much to say 4. Dae will implement reverse relative overdueness, but probably not anything beyond that. And probably not my idea with only having 3-4 sort orders and “Custom”, sadge 5. Both of those are dead ends. We _might_ be able to reduce RMSE by 1-2% (relatively), but not more than that, and even that will be difficult to do. That is, unless someone smarter than me and Jarrett decides to work on FSRS, but I doubt that will happen any time soon. ---
**Author:** ![DerIshmaelite](https://sea2.discourse-cdn.com/flex002/user_avatar/forums.ankiweb.net/derishmaelite/32/17983_2.png) [@DerIshmaelite](https://forums.ankiweb.net/u/DerIshmaelite)\ **Post date:** [October 28, 2024, 10:45am UTC](https://forums.ankiweb.net/t/fsrs-5-ideas-and-sims/51278/3 "2024-10-28T10:45:12Z")
> 6-Influence of Increasing Collection Size on Retention (Interference - Retrograde Inhibition) > 7-Influence of Time of Day (Fatigue) on Scheduling Time Intervals ---
**Author:** ![Expertium](https://sea2.discourse-cdn.com/flex002/user_avatar/forums.ankiweb.net/expertium/32/24176_2.png) [@Expertium](https://forums.ankiweb.net/u/Expertium)\ **Post date:** [October 28, 2024, 11:58am UTC](https://forums.ankiweb.net/t/fsrs-5-ideas-and-sims/51278/4 "2024-10-28T11:58:32Z")
1. I’m not sure what you mean. 2. I’ve experimented with that and got nowhere, and I don’t plan to continue working in that direction. ---
**Author:** ![DerIshmaelite](https://sea2.discourse-cdn.com/flex002/user_avatar/forums.ankiweb.net/derishmaelite/32/17983_2.png) [@DerIshmaelite](https://forums.ankiweb.net/u/DerIshmaelite)\ **Post date:** [October 28, 2024, 12:08pm UTC](https://forums.ankiweb.net/t/fsrs-5-ideas-and-sims/51278/5 "2024-10-28T12:08:55Z")
I think that the size of the collection and the rate at which it grows (number of new cards per day) influences the retention of older learned cards. **New cards interfere with retention of older cards** → Retroactive Inhibition [Retroactive inhibition | psychology | Britannica](https://www.britannica.com/science/retroactive-inhibition) [Learning theory - Stages, Acquisition, Retention | Britannica](https://www.britannica.com/science/learning-theory/Stages-of-learning#ref586992) This also applies in the opposite direction. **Older cards interfere with learning of new cards** → Anterograde Inhibition ![image](https://us1.discourse-cdn.com/flex002/uploads/anki2/original/3X/7/b/7bfe9a5aeb1891f1782ca4fc15cb28d24787d127.png) So I believe that **collection size (Older cards)** and **Number of New Cards per day** are variables that should be looked into. ---
**Author:** ![anon77623737](https://avatars.discourse-cdn.com/v4/letter/a/7993a0/32.png) [@anon77623737](https://forums.ankiweb.net/u/anon77623737)\ **Post date:** [October 28, 2024, 11:24pm UTC](https://forums.ankiweb.net/t/fsrs-5-ideas-and-sims/51278/6 "2024-10-28T23:24:30Z")
I was just doing my reviews today and I mixed up 連絡 and 連続. This one is due to visual similarity. However, more importantly, I have even mixed up visually dissimilar vocab because the meanings are so close to each other, despite the appearance of the word being completely different. Similar to the chair-table example, although it was for more abstract vocab, not nouns. ---
**Author:** ![Expertium](https://sea2.discourse-cdn.com/flex002/user_avatar/forums.ankiweb.net/expertium/32/24176_2.png) [@Expertium](https://forums.ankiweb.net/u/Expertium)\ **Post date:** [October 29, 2024, 12:36pm UTC](https://forums.ankiweb.net/t/fsrs-5-ideas-and-sims/51278/7 "2024-10-29T12:36:34Z")
Alright, I finished testing 1 vs 10 seeds @L.M.Sherlock @vaibhav @sorata For each collection, RMSE(10 seeds) is calculated simply as min(RMSE(seed 1), RMSE(seed 2)…RMSE(seed 10)). I deliberately chose 100 collections with a small numer of reviews and 100 collections with a large number of reviews. Difference in RMSE and the number of reviews. ![Figure_1](https://us1.discourse-cdn.com/flex002/uploads/anki2/original/3X/e/0/e090c99c379278c0f6c55bad8aa3c9d7b0ae5cc7.png) Distribution of differences ![Figure_2](https://us1.discourse-cdn.com/flex002/uploads/anki2/original/3X/2/b/2b8a092365f6ed34e88918661c236d70f2a183cd.png) Improvement and the number of seeds used ![image](https://us1.discourse-cdn.com/flex002/uploads/anki2/original/3X/a/a/aad682f733cf36dc1027d3939980394979dc0c8e.png) As you can see, it is absolutely not worth it even for people with \<5k reviews (for other people the average is even lower). Note that this is different from what I proposed [here](https://forums.ankiweb.net/t/anticipated-features-in-24-10/51056/38). It’s not the same idea. ---
**Author:** ![DerIshmaelite](https://sea2.discourse-cdn.com/flex002/user_avatar/forums.ankiweb.net/derishmaelite/32/17983_2.png) [@DerIshmaelite](https://forums.ankiweb.net/u/DerIshmaelite)\ **Post date:** [October 29, 2024, 12:42pm UTC](https://forums.ankiweb.net/t/fsrs-5-ideas-and-sims/51278/8 "2024-10-29T12:42:21Z")
So is the idea done for good ❓ What happens if you test against 100 seeds ❓ ---
**Author:** ![Expertium](https://sea2.discourse-cdn.com/flex002/user_avatar/forums.ankiweb.net/expertium/32/24176_2.png) [@Expertium](https://forums.ankiweb.net/u/Expertium)\ **Post date:** [October 29, 2024, 1:01pm UTC](https://forums.ankiweb.net/t/fsrs-5-ideas-and-sims/51278/9 "2024-10-29T13:01:11Z")
I will get grey hairs before the optimizer finishes. And the improvement will be small anyway, probably around -0.2% ---
**Author:** ![DerIshmaelite](https://sea2.discourse-cdn.com/flex002/user_avatar/forums.ankiweb.net/derishmaelite/32/17983_2.png) [@DerIshmaelite](https://forums.ankiweb.net/u/DerIshmaelite)\ **Post date:** [October 29, 2024, 1:19pm UTC](https://forums.ankiweb.net/t/fsrs-5-ideas-and-sims/51278/10 "2024-10-29T13:19:54Z")
> probably around -0.2% Yeah, no that is not worth the effort. ---
**Author:** ![system](https://us1.discourse-cdn.com/flex002/uploads/anki2/original/1X/8f1279ababc5879d54e4838989f606cfe55af8c7.jpeg) [@system](https://forums.ankiweb.net/u/system)\ **Post date:** [November 28, 2024, 1:20pm UTC](https://forums.ankiweb.net/t/fsrs-5-ideas-and-sims/51278/11 "2024-11-28T13:20:19Z")
Closed on November 28, 2024, 1:20pm UTC This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.