What significance do the constants in multiplier formula for score calculation have?

[FL44ME]

uhabits/android/uhabits-core/src/main/java/org/isoron/uhabits/core/models/Score.java

Line 66 in aee0da2

double multiplier = pow(0.5, frequency / 13.0);

Do the constants 0.5 and 13.0 have any practical significance in the multiplier formula or are these arbitrary smoothing factors?

@iSoron @hiqua

[hiqua]

It's basically one constant: pow(0.5, 1/13.0) to the power of frequency.

I think it's arbitrary, in this case it's around 0.95, so if frequency is 1, the old score should account for 95% of the new score.

See also #689 ("How is the score calculated?").

[FL44ME]

Thanks. As a follow-up: I feel that this method used for numerical habits should also have a reverse target feature (such as when trying to limit something). While for most bad habits that one would like to completely end, a 0 for target value works, something like weight is dynamic and people could be trying to both gain or lose weight. While reversing the structure of the question is a simple fix, a consistent pattern (higher score = better) on the dashboard across all habits would allow the user to easily get a feel of their scores quickly without having to pay particular attention to different habits. Is this feasible?

[hiqua]

Sounds like you're talking about #1101

[FL44ME]

Sounds like you're talking about #1101

Thanks for pointing it out

[iSoron]

This answer is correct. That particular constant was chosen so that the score reaches 80% after one month, 96% after two months and 99% after three months. I though about making this configurable, but there was very little user demand for it.