Question of statistics...

[edro]

Let's say there are 99 Red balls in a bag and 1 Green ball.

Blindly... if you choose one ball at a time, check the color, then put it back:
1. Would you have a 1% chance of getting the Green ball each time you chose?
2. If you chose 100 times, would you probably choose the Green ball once?
3. If you chose 1000 times, would you statistically choose the Green ball 10 times?

So if something is 99% true, does that mean 1/100 tries is false?

I never took a statistics class... Maybe that should be my next book.

 

[brandonbull]

as you are describing it, you have the right idea

 

[Atheus]

Put the ball back and it's independent events. Every time you pull a ball you have a 1/100 chance of getting the green one, regardless of how many times you picked it before.

 

[mugs]

You'd have a 36.6% chance of picking 100 white balls in a row ( (99/100) ^ 100 ), so you'd have a 63.3% chance of getting 1 green ball in 100 tries.

Edit: 1 or more green balls that is.

 

[yosuke188]

1. Yes
2. (99/100)^100 = probability of not picking a green. Subtract that from 100% and you get about 63% chance of picking a green.
3. (99/100)^1000, same process. 99.99% of picking a green.

 

[edro]

Originally posted by: yosuke188
1. Yes
2. (99/100)^100 = probability of not picking a green. Subtract that from 100% and you get about 63% chance of picking a green.
3. (99/100)^1000, same process. 99.99% of picking a green.

Originally posted by: mugs
You'd have a 36.6% chance of picking 100 white balls in a row ( (99/100) ^ 100 ), so you'd have a 63.3% chance of getting 1 green ball in 100 tries.

Ah ha! Thanks guys! I knew there had to be a spin on it...

 

[Cattlegod]

Originally posted by: edro
Let's say there are 99 Red balls in a bag and 1 Green ball.

Blindly... if you choose one ball at a time, check the color, then put it back:
1. Would you have a 1% chance of getting the Green ball each time you chose?
2. If you chose 100 times, would you probably choose the Green ball once?
3. If you chose 1000 times, would you statistically choose the Green ball 10 times?

So if something is 99% true, does that mean 1/100 tries is false?

I never took a statistics class... Maybe that should be my next book.

1 - yes
2 - 1-(99/100)^100
3 - With certain confidence, yes. but not 100% confidence

 

[edro]

Ok, so because condoms are 99% effective and birth control is 99.99% effective... how statistically rare is it that she is pregnant?

 

[mugs]

Originally posted by: edro
Ok, so because condoms are 99% effective and birth control is 99.99% effective... how statistically rare is it that she is pregnant?

I don't believe either of those numbers are accurate, but in combination they'd be 99.9999% effective.

 

[dullard]

Originally posted by: edro
Ah ha! Thanks guys! I knew there had to be a spin on it...

Yep. The spin is that you have to do the math on how many times you will NOT get what you want. Then subtract from 100% to get the number of times you will get what you want.

Most statistics problems are like that. You have to do the math for the NOT case and then convert back.

 

[edro]

I guess we'll name him Lucky.

 

[fitzov]

I think your second and third questions are worded a bit imprecisely.

If you have chosen 99 times, the probability of getting the green ball is 1/100

but the probability of getting a green ball in 1000 tries is 1/100 * (1000 * 1/100) = 10/100

 

[tfinch2]

Originally posted by: mugs
You'd have a 36.6% chance of picking 100 white balls in a row ( (99/100) ^ 100 ), so you'd have a 63.3% chance of getting 1 green ball in 100 tries.

Edit: 1 or more green balls that is.

You have a 0% chance of picking any white balls.