On true love, statistics and incorrect applications of the Drake Equation.

So here’s a link to a paper that has been sent my way a few times over the past few days, either through Twitter, Facebook links, IM or carrier pigeon. Please read it here.

Let’s put aside the obvious questions of “Why are a bunch of people sending Mehal a paper on the mathematical probabilities of love” (answer: I don’t know, ask them) and “Shouldn’t you be doing something more enjoyable right now?” (answer: Insert appropriate yo-mama joke here), and focus on three glaring errors I can readily see in this paper.

1. Not all variables are independent variables. I suppose this is a problem with the Drake Equation as well as this person’s essay. If we have two events, with probability p1 and p2 of occurring, the probability that they both will occur is p1 * p2 IF AND ONLY IF p1 and p2 are independent. 50% of all Americans are male. 50% of all Americans have Y-Chromosomes. But this does not imply that a randomly selected American has a 25% chance of having both Y-chromosomes and a mantenna array. I don’t know enough about astronomy to whether all the variables in the Drake equation are truly independent but Backus’ statistical analysis of London’s women is fraught with error. For example, depending on what activity Backus is doing, he may find a greater than 26% chance of meeting a college-educated person. Maybe he should consider hitting up trivia night.
   
2. Lack of rigor: The author has spent a lot of time explaining some variables (percent of women that have college degrees, percent of college educated women that are attractive) but chooses to completely ignore others (percent of women that find him attractive, percent of women that are single). Bull-honkey. If you’re going to write a statistics-focused paper, take in to account all statistics before coming up with your final number.
3. Empirical contradiction: If you read Aristotle’s treatises on physics, two things come to mind. First, he’s an eloquent writer whose work bridges a multi-millennial gap and allows the reader to effectively visualize and contemplate his points. Second, he is absolutely 100% wrong about some fairly basic things. Heavy bodies do not fall to the earth faster. An arrow flies in a parabolic trajectory, not a straight line. If you come up with a hypothesis that is refuted by empirical observation, you either discard your hypothesis or find some explanation for why the empirical observations are incorrect. As I write this, there are 3 wedding invitations on my desk (congratulations to you all) and a couple e-Invites, all from people who had high standards for love and have found people who met them. If Backus thinks that my observations are incorrect, Occam’s Razor places the burden of proof on him.
   

In short, a brief plea to the internet. The next time someone wants to generate a paper on the probabilities of finding true love, at least get it peer reviewed. And remember that love is the most confounding of all variables.