probability and statistics
Statistics. When you draw the right card, you can quote them with glee. When you draw the wrong card, they are worthless. And people commonly say that you can make statistics mean anything you want anyway. Which is, ahem, probably true.
Statistics, of course, are largely based on ideas of probability. But what is so fascinating is that you can also make up probability scenarios which are mathematically correct but somehow also seem largely meaningless, or counterintuitive, or just plain WRONG. Here's a great example:
Probability can Bite
If you don't want to read the article, the synopsis is this: If i say "I tell you I have two children and that (at least) one of them is a boy, and ask you what you think is the probability that I have two boys." the answer is 1/3. However, if i say "I tell you I have two children and that (at least) one of them is a boy born on a Tuesday, and ask you what you think is the probability that I have two boys." the probability changes to 13/27....which is quite different, closer to 1/2 than 1/3.
WTF, that doesn't make sense at all!
Fine, ok, the numbers are correct. But you can hardly convince me that it is relevant that one boy was born on a Tuesday. You know, i have always hated probability because of shit like this. But at least it helps me not worry about it so much.
---------------------------------------------------------
I thought about this some more. It helps to actually do the calculation (which i did just to check it...and to make myself understand it. I can't really believe any conclusion from one of these probability problems unless i draw it out myself on a probability tree.) What we have here is basically a variation on the Monty Hall problem. It illustrates nicely a few important things that must be kept in mind when faced with any and all probabilities/statistics --
1. Any information at all changes your sample space (mathematically speaking, that is).
2. The way a question is stated can radically change the answer.
The only thing this problem lacks is something to help us remember that:
Correlation does not equal causation!
For that we need to change the problem to include a baby born dead (on Tuesday) to a mother who is poor, smokes, and doesn't brush her teeth!

1 comment:
dude, you just made my brain explode.
Post a Comment