Monday, September 20, 2010

prediction

As they say, prediction is hard, especially about the future.

Fortunately, "they" in this case are not me, and "they" happen to be drastically wrong. Prediction is easy. Most especially about the future. In fact, I've managed to put together an amazingly precise prediction engine that can, with great accuracy, forecast events in my life.

When constructing a prediction, it's important that you first understand the concept of a probability. Probabilities are numerical quantifications of random events. For example, tonight my neighbors will probably make a number of random noises, due to the quantity of events they host behind their fence.

As well as being expressed as whining statements, probabilities can also be expressed as numbers between 0 and 1. Take, for example, a few tried-and-true probabilities:

Probability that a squirrel will attack me at some point on my bike ride to work this week: .25

Probability that my neighbors will wake me up tomorrow at 6:00 AM coughing like wounded elephants: .65

Probability that on any given day I run into a female acquaintance: .05

Probability that said woman, knowing next to nothing about me, will launch into a tirade about how I'm too picky about women and need to "lower my standards": 60,000,000

Please note that some probabilities can, in fact, exceed 1, but at that point they lose all meaning, unless the meaning is meant as a passive-aggressive jab.

Anyways, as I said, given my detailed knowledge of probabilities, I've been able to construct very accurate and reliable predictions.

Take, for example, the subject about which I was recently berated. This woman I ran into obviously has no clue how probability works. I present exhibit A:


You can clearly see that the problem is not that my "standards", if they do, in fact, exist,1 prevent me from dating a woman. I am prevented from dating women because I am only interested in women not interested in me. This can be clearly seen from the figure.

This is, of course, the reason that the only girlfriends I've ever had have been women completely and entirely uninterested in me.2

As a corollary to The First Grand Law of Probability, let me present you with exhibit B:


This conclusively proves, furthermore, that probability also prevents me from being interested in women who are available. There is nothing that can be done. Math itself is the cause of it, and I would not dream of going against math.

One might ask one's self what would happen if one of those unavailable women suddenly became available, and what impact that would have on the probability of my being interested in her. Happily, I can answer that quandary with solid empirical evidence: as expected, the probability also experiences a dramatic change in value, from 1 to 0.

Now, lest you've just trudged through this post chafing at the fact that I only use this publication as a vehicle for complaining about my dating life, let me assure you that I also use this as a vehicle for complaining about my financial life. As proof, exhibit C:


But don't feel too badly for me folks. With the knowledge that my investments are bad with a probability of 1, I'll be using this to start a fund to short all of my picks. Money tree, and uninterested trophy women, here I come.3

1. This is not intended as an insult to women I have recently taken out. If you happen to be one of them.
2. No, really. Ask them. If nothing else, this argues for my amazing powers of persuasion.
3. There's just no way I could make it through this post without mentioning the divide in the statistical world between the Frequentists and the Bayesians, which is sort of like the divide between slobbering monkeys and Sean Connery. The only thing you really need to know is that the Bayesians are right. And better dressed. And more attractive. And excellent lovers. Even if Frequentists rule the world.

No comments: