Why Statistics is Bullshit

Background Information

We calculus aficionados have long protested that statistics is bullshit. However, our proclaims have always been regarded as simply elitist and narcissistic. The stat buffs have always viewed the calc lovers as nefarious in both attitude and demeanor. But this is going to all change. I am going to settle the score once and for all. Tune in my vivacious readers, get comfortable and take a deep breath: I’m going to take you for a mathematical adventure.

Personal Interest

Due to a series of unfortunate events and four years after high school calculus, I find myself taking Probabilities & Statistics (Math 67) here at UC Irvine. Wha wha, cry me a river — big deal — but how is this all related you ask? Well, it’s pretty simple actually — I will attempt to demonstrate why statistics is pretty much bullshit and can even be misleading to the masses. Statistics is not only a discipline which has a long history of discouraging betting and gambling but consequently through its biblical teachings of odds and probabilities has almost single handedly destroyed the Vegas Dream. Anything which discourages our dirty, filthy and lowlife right of Vegas culture should be abolished. That is my personal standing on the matter. If you disagree, then well, we might have a problem here.

A Proof by Example

They say fight fire with fire. And alas, here it is:

Let’s assume there’s a population P and a new disease D. Wait, wait come back — before you run away from this entry because I have desecrated it with some mathematics — I want you to be well aware that I’m not going to make you do any math; you just have to read and follow the logic. Simple as that. Okay, so now that you’re a little more comfortable, let’s continue. Where we were? Oh right, fighting fire with fire:

Our Fuel:
Let’s assume that scientists and researchers have developed a new test to see if you are infected by a new disease. Unfortunately, the test results are not perfect (they never are) but close. So let’s say that the test is 99% accurate. That looks like some good odds, right? And let’s say this disease is very rare such that only 0.01% (i.e. 0.0001) of the population is actually infected. And for simpleton’s sake, let’s say our population is a nice round number like 100,000,000 people. So, let’s chalk up our variables…

Our Lighter:
P = 100,000,000 (our total population size)
D = 0.0001 (percentage of population who is actually infected)

Our Ignition:
Now, the numbers game: this means that 10,000 people have the disease and that 99,990,000 do not. Makes sense so far. Of the 10,000 people who have the disease who take the test 9,900 will test positive and 100 will test negative (i.e. the test is 99% accurate). Similarly, this means of the 99,900,000 who do not have the disease, 999,900 will test positive and 98,990,100 will test negative.

So of the 9,900 + 999,900 = 1,009,800 people who test positive, only 9,900 (or 0.98%) actually have it. And of the 100 + 98,990,100 = 98,990,200 who test negative, 100 (or 0.0001%) actually have it.

Our Ashes:
What does this mean? It means that even though the tests are 99% accurate, if you test positive you still have less than a 1% chance of actually being infected. Yes, the calculations are correct and no, I am not trying to fool you with some fuzzy math. It’s just that statistics is bullshit.

A Mature Conclusion

Suckers. We got you good. You suck!

Acknowledgments & Shout outs

This dissertation could not have been possible without my “FUCK STATS” comrade, Nicole. Her undying support and violent encouragement to destroy all things stats really paved the way for this entry and public involvement. You my girl, Nicole!

Modern Day Capitalism

Nicole and I have already begun designing and printing “FUCK STATS” t-shirts for all ages and sizes. What better gift to a newborn then the reassuring confidence to “FUCK STATS”? What better wedding gift to reaffirm the notion of “FUCK STATS” embroidered on an elegant cotton t-shirt to remind the bride and groom that 62% couples are infidel within the first 3 months of marriage with a 87% chance of death or divorce? What kinder thought than to surprise your significant other with our upcoming Valentine’s Day Special Edition “FUCK STATS” t-shirt to show your committed support to your relationship despite the statistical inevitability that 99.7% of relationships among youth end in bitter dispute with an even greater probability that’s it’s not you, it’s someone else.

Interested parties should inquire within. I personally intend on sporting my official “FUCK STATS” t-shirt and hoodie to my Probability & Statistics final. Look for me — I’m going to constantly raise my hand and ask obnoxious questions like if we’re allowed to phone a friend or use 50/50. It’s going to be a good quarter.

Adieu. Navid.