fsrs4anki-helper: [BUG] A serious and strange bug when rescheduling

Describe the bug Read these two posts: https://www.reddit.com/r/Anki/comments/16dk35h/fsrs_for_anki_parameter_sanity_check/ https://www.reddit.com/r/Anki/comments/16d56xp/spits_coffee/ They have two opposite problems - abnormally low number of reviews and abnormally high number of reviews. The author of the first post says that not only the Burden is very low, but the actual number of cards to review is very low as well. This means that it’s not just a bug with the Burden formula, but a much more serious issue with rescheduling itself.

To Reproduce Unfortunately, it’s unclear how to reproduce it.

Additional Context Settings and parameters of the user with an abnormally high Burden:

"deckName": "global config for FSRS4Anki",
"w": [0.1358, 0.5044, 18.4034, 18.4034, 4.8368, 1.1219, 0.8531, 0.0, 1.3932, 0.1, 0.8975, 2.0978, 0.1428, 0.3228, 1.3624, 0.2382, 2.7056],
// The above parameters can be optimized via FSRS4Anki optimizer.
// For details about the parameters, please see: https://github.com/open-spaced-repetition/fsrs4anki/wiki/The-Algorithm
// User's custom parameters for global
"requestRetention": 0.9, // recommended setting: 0.75 ~ 0.95
"maximumInterval": 36500,

Settings and parameters of the user with an abnormally low Burden:

Parameters: [0.438, 0.7791, 1.0778, 12.3026, 5.8574, 0.6821, 0.8479, 0.0594, 1.2555, 0.1, 0.9616, 2.8513, 0.011, 0.1572, 1.9999, 0.5743, 3.69]
Requested retention: 0.9
Max interval: 36500