Blog Thirty: That's Pearl, But Three Is Leather
As this is Blog #30, I was thinking of marking the occasion by pointing out that "hey, it's the silver blog." But silver is 25. Gold is 50. What the hell is 30?
I looked it up; pearl is 30. There is in fact a traditional gift type for every anniversary up to #15, after which you don't have anything specific except on anniversaries that are evenly divisble by ten. This year my third anniversary with Snaby Girl is the leather anniversary. So I know what to get her... leather... huh huh huh. Ohhhhhh yeah.
A leather jacket. What were you thinking, pervert?
I'm having fun watching the election polls. They bounce up and down depending on who's doing the polling but you can split the difference and you'll usually be right. I'm guessing a popular vote split of 37-30-20, Conservative-Liberal-NDP, Quebec obviously being different. I'll even tell you how the election will turn out in seats:
Conservatives: 132
Liberals: 82
Bloc: 64
NDP: 28
If I read one more moron on a message board say "But they only poll, like, a thousand people. MILLIONS of people are voting man. So they can't be right!" I'll scream. A thousand people is a huge number of people, a huge sample. The only reason it's as jumpy as it is is that different regions vote differently.
To get an idea of what I mean, look at it this way. Grab a coin, any coin near you. Do me a favor; flip it ten times, and do that ten times. Ten series of 10 flips. Shouldn't take you long. I'll wait.
Good. Now, you probably had at least one or two sets of ten where you didn't come up with 5 heads or 5 tails. Sure, 5-5 is the usual split. But in 10 flips, you can get 7 heads 3 tails pretty easy, right? I just did it and got (heads-tails) 4-6, 4-6, 7-3, 7-3, 6-4, 3-7, 4-6, 6-4, 6-4, 2-8. I didn't once flip exactly 5 of each, itself kind of flukey; an exact 5-5 split should happen about 23% of the time. But getting 70% heads or tails, or more, should happen 34% of the time.
So let me ask you this; suppose instead of flipping a coin 10 times, you flip it a THOUSAND times. How often would you get 70% heads, or 70% tails. That is, 700 or more heads, or 700 or more tails, out of a thousand flips. Same percentage as before, just a different sample size.
Answer? Never. It would never, ever happen. You could flip coins your whole life. You could hire a million people to flip coins from now until the day the Sun burns out six billion years from now, and it still would not happen. The odds of that happening are 1 in 8,832,839,000,000,000,000,000,000,000,000,000,000. That is a number larger than the number of atoms in the Pacific Ocean.
So a thousand people is a BIG sample. You cannot go too far wrong.
But people have time to change their minds, of course.
I looked it up; pearl is 30. There is in fact a traditional gift type for every anniversary up to #15, after which you don't have anything specific except on anniversaries that are evenly divisble by ten. This year my third anniversary with Snaby Girl is the leather anniversary. So I know what to get her... leather... huh huh huh. Ohhhhhh yeah.
A leather jacket. What were you thinking, pervert?
I'm having fun watching the election polls. They bounce up and down depending on who's doing the polling but you can split the difference and you'll usually be right. I'm guessing a popular vote split of 37-30-20, Conservative-Liberal-NDP, Quebec obviously being different. I'll even tell you how the election will turn out in seats:
Conservatives: 132
Liberals: 82
Bloc: 64
NDP: 28
If I read one more moron on a message board say "But they only poll, like, a thousand people. MILLIONS of people are voting man. So they can't be right!" I'll scream. A thousand people is a huge number of people, a huge sample. The only reason it's as jumpy as it is is that different regions vote differently.
To get an idea of what I mean, look at it this way. Grab a coin, any coin near you. Do me a favor; flip it ten times, and do that ten times. Ten series of 10 flips. Shouldn't take you long. I'll wait.
Good. Now, you probably had at least one or two sets of ten where you didn't come up with 5 heads or 5 tails. Sure, 5-5 is the usual split. But in 10 flips, you can get 7 heads 3 tails pretty easy, right? I just did it and got (heads-tails) 4-6, 4-6, 7-3, 7-3, 6-4, 3-7, 4-6, 6-4, 6-4, 2-8. I didn't once flip exactly 5 of each, itself kind of flukey; an exact 5-5 split should happen about 23% of the time. But getting 70% heads or tails, or more, should happen 34% of the time.
So let me ask you this; suppose instead of flipping a coin 10 times, you flip it a THOUSAND times. How often would you get 70% heads, or 70% tails. That is, 700 or more heads, or 700 or more tails, out of a thousand flips. Same percentage as before, just a different sample size.
Answer? Never. It would never, ever happen. You could flip coins your whole life. You could hire a million people to flip coins from now until the day the Sun burns out six billion years from now, and it still would not happen. The odds of that happening are 1 in 8,832,839,000,000,000,000,000,000,000,000,000,000. That is a number larger than the number of atoms in the Pacific Ocean.
So a thousand people is a BIG sample. You cannot go too far wrong.
But people have time to change their minds, of course.
0 Comments:
Post a Comment
<< Home