Currently, grow penalty happiness is flat for every threshold:
The consequence of that is the difference between 50% and 49% happiness is big, while there is no difference between 49% and 36%, which makes no sense and it feels artificial.
Proposal: Make the grow penalty from unhappiness linear:
Thanks to that every difference in happiness would result in the same difference in growth modifier (except 10% and below and 50% and above). The function feels more natural than artificial thresholds.
The function: Growth modifier (in %) = Happiness (in %) * 2.5 - 25, but should never be less than 0% and 100%, so clamp(growth_modifier, 0, 100)
On average the growth penalty would about the same as before, but it would smoothly decrease instead.
Link to calculations: https://docs.google.com/spreadsheets/d/1jpm5R2CGyeqcMf--m6rxxLrtCwNwGpuVFcP-pAnONPE/edit?usp=sharing
The consequence of that is the difference between 50% and 49% happiness is big, while there is no difference between 49% and 36%, which makes no sense and it feels artificial.
Proposal: Make the grow penalty from unhappiness linear:
Thanks to that every difference in happiness would result in the same difference in growth modifier (except 10% and below and 50% and above). The function feels more natural than artificial thresholds.
The function: Growth modifier (in %) = Happiness (in %) * 2.5 - 25, but should never be less than 0% and 100%, so clamp(growth_modifier, 0, 100)
On average the growth penalty would about the same as before, but it would smoothly decrease instead.
Link to calculations: https://docs.google.com/spreadsheets/d/1jpm5R2CGyeqcMf--m6rxxLrtCwNwGpuVFcP-pAnONPE/edit?usp=sharing