Tilted Forum Project Discussion Community - View Single Post

BigBen · 03-11-2005, 10:24 AM

My undergrad degree was in Economics, with a minor in stats.

I can save you thousands of dollars in tuition and countless hours of studying. Pay close attention:

Graphs, Numbers and Statistics are bullshit. I say that with a 99% confidence interval.

When someone feels strongly about something, they can look around in their environment, count the number of times something fits into their belief, and then create an equation (or worse, a fucking graph) to tell someone that they are right.

Only hard science benefits from statistics. If you are doing physics, engineering, biology, chemistry and the like, then the standard Bell curve is necessary and appropriate. In social sciences, when dealing with people's actions and the complexities therein, all statistics do is cover the person's motives...

I will tell you something, and you tell me if I am wrong. The graph in question you have shown was presented in a context where:
a) the presenter does not like graffiti
b) they are proposing that funds be increased to clean up that graffiti

Therefore, showing an "exponential growth" curve like that shows a tremendous amount of payoff if we could cover up graffiti in 24 hours.

Notice the x-axis as well. That time representation is not linear. You are being lied to with numbers, plain and simple. if the number looked (24 hours, 48 hours, 72 hours) and increments of 24 hours equidistant from the origin, that 'curve' would look drastically different.

Post Script: Yes, it does look like a big fuck-up using the Microsoft Graph grenerator function in the MS Office package. Fucking amateurs.

Sorry, but people lying makes me angry.

03-11-2005, 10:24 AM	#6 (permalink)
BigBen Comedian Location: Use the search button	My undergrad degree was in Economics, with a minor in stats. I can save you thousands of dollars in tuition and countless hours of studying. Pay close attention: Graphs, Numbers and Statistics are bullshit. I say that with a 99% confidence interval. When someone feels strongly about something, they can look around in their environment, count the number of times something fits into their belief, and then create an equation (or worse, a fucking graph) to tell someone that they are right. Only hard science benefits from statistics. If you are doing physics, engineering, biology, chemistry and the like, then the standard Bell curve is necessary and appropriate. In social sciences, when dealing with people's actions and the complexities therein, all statistics do is cover the person's motives... I will tell you something, and you tell me if I am wrong. The graph in question you have shown was presented in a context where: a) the presenter does not like graffiti b) they are proposing that funds be increased to clean up that graffiti Therefore, showing an "exponential growth" curve like that shows a tremendous amount of payoff if we could cover up graffiti in 24 hours. Notice the x-axis as well. That time representation is not linear. You are being lied to with numbers, plain and simple. if the number looked (24 hours, 48 hours, 72 hours) and increments of 24 hours equidistant from the origin, that 'curve' would look drastically different. Post Script: Yes, it does look like a big fuck-up using the Microsoft Graph grenerator function in the MS Office package. Fucking amateurs. Sorry, but people lying makes me angry. __________________ 3.141592654 Hey, if you are impressed with my memorizing pi to 10 digits, you should see the size of my penis.