[Feature Request] Loosen the clampings for w[9] about fsrs-optimizer HOT 5 CLOSED

According to https://supermemo.guru/wiki/Stabilization_decay, w[9] is usually in the range of [0.2, 0.8]. I have loosened it to [0.1, 0.8].

from fsrs-optimizer.

Supermemo formulas aren't exactly the same as ours.

from fsrs-optimizer.

At least for stabilization decay, FSRS and SM are the same. And for easy cards, I'm afraid it's due to the underestimation of initial stability.

from fsrs-optimizer.

I just checked parameters of another user, he also has the same issue with w[9].
[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]
I recommend you to check optimal parameter for all users who have submitted their data, to see how often this happens.

from fsrs-optimizer.

My main worry is if w[9] is too small, the long-term interval will be very large, which could scare new users.

from fsrs-optimizer.