Procrastination and resurrecting an old post…

Back in the days when I had a Xanga blog (it’s been long since deleted), I worked on this particular equation. I was tremendously amused by it and started a bit of a trend among my other blogging friends to develop similar equations. Bear in mind that I was also in a Masters’ level math/statistics course and working with this kind of stuff daily, so it seemed funnier to all of us geeks than it might to the average reader. Sigh. Anyway, I intend to give it a run-through and do some tweaking on it based on another 2½ years of experience…

************************************

Let L_uv = Event that I like man X

A_x = Attractiveness of X
PL_x = My perception that X likes me (note: not whether he really does or not)
F_x = How fascinating I find X
E(t,s) = My emotional state depending on t = time and s = external stressors
MP_x = Mate potential of X

Then P(L_uv) = (0.005)*A_x + (0.01)*(PL_x + F_x + E(t,s)) + (0.02)*MP_x

where

A_x = [(Looks, 0-20) + (Intelligence, 0-100) + (Fiscal Status, 0-10) + (Articulateness, 0-20) + (Confidence, 0-50) - ("He's too perfect!" freak-out, 0-50)]⁺

PL_x = [(Special attention X pays to me, 0-20) - (Special attention X pays to other women, 0-20)*(Number of other women) + (How much X listens to me, 0-20) + (How interested X is in understanding me, 0-20) + 20*(Avg. # of times he calls/seeks time with me weekly) - (Stalker factor, 0-40) - 40*{(Avg. # of times he calls/seeks time with other women weekly) + (Avg. # of times he talks to me about other women weekly)}]⁺

F_x = [(Power, 0-20) + (Social Status, 0-10) + (Fame, 0-10) + (Foreignness, 0-10) - (Redneck Pride, 0-20) - (Materialism, 0-30)]⁺

E(t,s) = [(Long-term sinusoidal function E₁(t), 0-75) + (Short-term sinusoidal function E₂(t), 0-25) - (Increasing linear function E₃(s), 0-100)]⁺

MP_x = [(Responsibility, 0-20) + (Christian? 0 or 1)*(Knowledge of his faith, 1-5)*(Love of God, 1-10) + (Health, 0-10) + (Stability, 0-20) - (How much he wants a mom/provider, 0-100)]⁺

And once again:

P(L_uv) = (0.005)*A_x + (0.01)*(PL_x + F_x + E(t,s)) + (0.02)*MP_x

Note that originally I weighted all the five factors evenly, but upon reflection decided to adjust my weights, multiplying the weight for MP_x by 2 and for F_x by 0.5. This should still give me a valid probability for P(L_uv). I still need to run some numbers on it, but I’m thinking that this is a pretty good model.

Wow. Am I ever a nerd! Would you believe, I spent some time deciding whether I was going to attempt a Bayesian analysis of this with conditional probabilities, or if I wanted to construct a values hierarchy or perhaps set this up as a network problem to solve for maximum flow (which in this case would be attraction to X) . . . I’m still tempted to set it up as an influence network and run it on a program I have access to that would give me a sensitivity analysis and point out what are the key nodes.

**********************

OK, this is current KJ again… now I should get back to my homework–trying to get a video into my powerpoint presentation and writing a bullet background paper about Lebanon. Anything else? Probably, but I’m not thinking of it at present…