This is inaccurate: to a man, all women are attractive; however, all women are also scary. These are almost entirely independent of each other, although there may appear to be a slight positive correlation. Neither variable is ever zero. These variables are also not inherent properties of either person: they apply to (man, woman) tuples.
The effect of alcohol is interesting: add it to the male and his estimate of the scariness of any given female decreases; add it to the female and her perceived scariness increases.
We men are simple beings. We’ll approach those women for whom the attractiveness – scariness number is positive in decreasing order of magnitude.
Other important factors, such as availability, can be modeled as mutations to one or both of attractiveness and scariness.
And no, I’m not telling any of you what your comparative attractiveness and scariness positions are in my perception.