• Welcome to New Hampshire Underground.
 

News:

Please log in on the special "login" page, not on any of these normal pages. Thank you, The Procrastinating Management

"Let them march all they want, as long as they pay their taxes."  --Alexander Haig

Main Menu

FUN WITH STATISICS

Started by microtone, February 07, 2010, 11:26 AM NHFT

Previous topic - Next topic

microtone

First off the etymology of the word 'STATISTICS' , means "The Science of The State", so you shouldn't be surprised where this post is headed......

"There's Lies, Damn Lies, and then there's Statistics"...

A fallacy in the law of large numbers and the 'probability of death' was pointed out by Richard Von Mises in his PROBABILITY, STATISTICS and TRUTH which entails people jumping to rather unreliable conclusions... (not to be confused with Ludwig Von Mises other cousin D.A. Gilles and his OBJECTIVE THEORY OF PROBABILITY)

Using figures of 1,000,000 people and 23 different insurance companies they calculated a 'relative frequency of deaths' as 940:85,020=0.01106 for 41y.o. insured persons.

Now where it get's interesting is people assume that means the chances of a 41 year old dying in the next year is .011. Which is incorrect. It might be adequate for an Insurance Company to calculate rates based on that 'collective'. But as individuals there are other factors, and we all will belong to different statistical 'collectives'.

It might help someone calculate relative risk. But it's really not a valid statistic to apply to individual cases. When it comes to "adjusting for other factors", frankly the math involved in accounting for just a single variable is a bit of a stretch, I don't know. When there's multiple factors at play, I going to state my opinion as: At best, it's a guess. And usually, it's plain wrong.

Moving on to the LOTTERY... Hey! wasn't that suppose to pay for educastion ?

Okay! Buy 1 lottery ticket, have one chance," buy more and increase the odds ?" Buy 2 tickets and DOUBLE your chances ? Buy 3 and TRIPLE ? Actually using the classical approach(Richard) if the odds are 1:1,000,000 the chances of winning are: .000001, when you buy 2 tickets the odds increase to 1:999,999 =.000001 even if you use 2:1,000,000 your odds increase to a staggering .0000002.

Now looking to the Objective Theory, it can be said in plain english. There's only one winning number therefore buying more than one ticket can't improve the odds. You only have one chance (only one winning number) so it's a waste to buy more than one ticket.

Moving on to the war on tobacco. Looking at the facts, alcohol is more damaging than cigarettes, and if there really was a public health concern, alcohol should be banned instead. I love to pull this one on my 'drug warrior'  friends.

Is alcohol safe or healthy ? NO.  "But we have scientific proof: the government told us so.."

While there is undeniable empirical evidence that smoking damages the esophageal cells, is this caused by those dirty cancer sticks or 'total particulate dust matter' ? In fact, your better off moving out of a polluted area than quitting smoking.

Do both, please, but then the booze is gonna get you.....