Awesome catch, you're absolutely right on the negative chance. My original intention was to have the chance be greater the lower the age bracket, so for example 1-4=3. Late night coding means I forgot to remove that negative. The intention for that was to replace the (age-4) with: ((age-4)*-1)
Both of those errors were pretty easily resolved by just having a variable with the inverted age then replacing the (age-4) and the age in (age*5) with ageinverted. It's var ageinverted = (age-4)*-1 if you're curious. All of this is fixed in v0.9.7