**Forums**- Cross Site
- The Asylum
- Final Stretch Election Thread

Unplanned

13,228 posts

No. You are wrong.

Go look up how statistical models and polling works.

Go look up how statistical models and polling works.

Just as Planned.

Liberal Conspiracy

TILT

Almalieque wrote:

Another John Stewart reference...

http://www.thedailyshow.com/watch/wed-october-10-2012/frequently-asked-questions-

http://www.thedailyshow.com/watch/wed-october-10-2012/frequently-asked-questions-

Instead of picking a segment made purely for humor, since you're on a Jon Stewart kick, you could have linked to his interview with Mr. Silver. Or, if you really want, the extended interview located in the side bar "related videos" section.

Edit: Actually, I take that back since I was thinking of a different time I saw him somewhere talking more about polling. This interview is more into campaign micro-targeting. The interview is interesting though so I'll leave my links up.

Belkira wrote:

Wow. Regular ol' Joph fan club in here.

The All Knowing

10,152 posts

Timelordwho wrote:

No. You are wrong.

Go look up how statistical models and polling works.

Go look up how statistical models and polling works.

Your claim contradicts math. Your only counter would be a proof on how less than 1/2 of 1% accurately depicts the remaining 99%. Since it's mathematically impossible and the definition of statistics supports that, I would like for you prove that. You can ridicule me in your Nobel speech.

Jophiel wrote:

Almalieque wrote:

Another John Stewart reference...

http://www.thedailyshow.com/watch/wed-october-10-2012/frequently-asked-questions-

http://www.thedailyshow.com/watch/wed-october-10-2012/frequently-asked-questions-

Instead of picking a segment made purely for humor, since you're on a Jon Stewart kick, you could have linked to his interview with Mr. Silver. Or, if you really want, the extended interview located in the side bar "related videos" section.

The point was to show the polls of what people see and not to show how polling works. What's the point of an accurate poll if it gets saturated with contradicting polls? How can you say that polling matters if you can't even differentiate an accurate poll from a bogus polls. Although that was a comedic skit, those polls were meant to be taken seriously.

Tracer Bullet

12,636 posts

Almalieque wrote:

I never said that you couldn't predict with a sample. I said that it is mathematically impossible to accurately predict an outcome with a sample that is less than 1/2 of 1% of the population. There is a difference between statistics and mathematics.

http://www.surveysystem.com/sscalc.htm

Plug in 1 for confidence interval. That means you want the result to be +/- 1% margin of error.

Check 99% for confidence level. That means that the you want the sample's result to fall within the specified confidence level 99% of the time. I.E, if you repeated the sample 100 times, 99 times you'd be within that +/- 1% of the true number.

Enter 225000000 for population, which is approximately the eligible voting population of the US.

You will see that the sample size needed for this level of confidence is only

These numbers aren't some crazy magic they invented just now. This math was first developed like 75 years ago, and involves some intense calculus, and you can double-check it with experiments and computer simulations.

Na Zdrowie

Liberal Conspiracy

TILT

Because (and this does come up in the interview) the point isn't to hang on each individual number. You get a large enough base that the outliers come out in the wash. My favorite quote from Silver was a tweet he made saying "The point of examining poll aggregates is that you don't need to litigate each individual poll". So when Gbaji starts hooting that some poll was D+5 and he says it really should be D+2 or whatever, whether or not he's even accurate sort of doesn't matter because there's so many polls at this point (Silver says in the before linked blog entry that Ohio is a sample size of 17,000+ people right now) that the static gets largely worked out. Likewise, saying "Poll A says O+2 but Poll B says R+1 so no one can be right!!" is foolish because it's not about just polls A & B.

Now, granted, Gbaji wants everyone to believe that*every* poll is flawed because that's the only way he can point to a Romney win. So be it. And you want to continue to insist that polling and poll aggregates can't be accurate because "1/2 of 1%!!" and that's fine as well. Previous election results don't support you and Gbaji is clutching at strands of hope but I can't say it affects me much.

Romney*can* potentially win. I don't think he will but even a 1-in-5 chance is a 1-in-5 chance. But the odds are well against him.

Now, granted, Gbaji wants everyone to believe that

Romney

Belkira wrote:

Wow. Regular ol' Joph fan club in here.

10,538 posts

Alma, you keep saying math does work that way when the way polls work and statistical representation work are based on math...

Here's a suggestion. Since math 'doesn't work that way' Can you please explain the math that shows this doesn't work?

Here's a suggestion. Since math 'doesn't work that way' Can you please explain the math that shows this doesn't work?

01001001 00100000 01001100 01001001 01001011 01000101 00100000 01000011 01000001 01001011 01000101

You'll always be stupid, you'll just be stupid with more information in your brain

Forum FAQ

You'll always be stupid, you'll just be stupid with more information in your brain

Forum FAQ

The All Knowing

10,152 posts

trickybeck wrote:

Almalieque wrote:

I never said that you couldn't predict with a sample. I said that it is mathematically impossible to accurately predict an outcome with a sample that is less than 1/2 of 1% of the population. There is a difference between statistics and mathematics.

http://www.surveysystem.com/sscalc.htm

Plug in 1 for confidence interval. That means you want the result to be +/- 1% margin of error.

Check 99% for confidence level. That means that the you want the sample's result to fall within the specified confidence level 99% of the time. I.E, if you repeated the sample 100 times, 99 times you'd be within that +/- 1% of the true number.

Enter 225000000 for population, which is approximately the eligible voting population of the US.

You will see that the sample size needed for this level of confidence is only

These numbers aren't some crazy magic they invented just now. This math was first developed like 75 years ago, and involves some intense calculus, and you can double-check it with experiments and computer simulations.

You do realize that there is a difference between statistics and math right? Statistics use Math, They aren't one in the same. It is mathematically impossible to accurately determine the outcome by taking a sample that is less than 1/2 of 1% of a population. At best, it is an educated guess. I'm not saying the poll will never be right, but that's the point of statistics, if it weren't ever right, then what would be the purpose of statistics? But, once you crossover into the mathematics domain, it simply isn't true anymore. It's a middle school concept.

10,538 posts

you keep going on about the mathematics, so please explain how mathematics disproves the process as opposed to being the core of it.

01001001 00100000 01001100 01001001 01001011 01000101 00100000 01000011 01000001 01001011 01000101

You'll always be stupid, you'll just be stupid with more information in your brain

Forum FAQ

You'll always be stupid, you'll just be stupid with more information in your brain

Forum FAQ

10,538 posts

on a side note Silver is going to be back on the daily show on the 7th. I'm interested in watching that one.

Alma, go read the signal and the noise.

Alma, go read the signal and the noise.

01001001 00100000 01001100 01001001 01001011 01000101 00100000 01000011 01000001 01001011 01000101

You'll always be stupid, you'll just be stupid with more information in your brain

Forum FAQ

You'll always be stupid, you'll just be stupid with more information in your brain

Forum FAQ

Tracer Bullet

12,636 posts

Almalieque wrote:

You do realize that there is a difference between statistics and math right? Statistics use Math, They aren't one in the same. It is mathematically impossible to accurately determine the outcome by taking a sample that is less than 1/2 of 1% of a population. At best, it is an educated guess. I'm not saying the poll will never be right, but that's the point of statistics, if it weren't ever right, then what would be the purpose of statistics? But, once you crossover into the mathematics domain, it simply isn't true anymore. It's a middle school concept.

Your an moran.

Na Zdrowie

The All Knowing

10,152 posts

trickybeck wrote:

Almalieque wrote:

You do realize that there is a difference between statistics and math right? Statistics use Math, They aren't one in the same. It is mathematically impossible to accurately determine the outcome by taking a sample that is less than 1/2 of 1% of a population. At best, it is an educated guess. I'm not saying the poll will never be right, but that's the point of statistics, if it weren't ever right, then what would be the purpose of statistics? But, once you crossover into the mathematics domain, it simply isn't true anymore. It's a middle school concept.

Your an moran.

Please define a "moran".

Liberal Conspiracy

TILT

Sir Xsarus wrote:

on a side note Silver is going to be back on the daily show on the 7th. I'm interested in watching that one.

Be more interesting if Romney is elected

Belkira wrote:

Wow. Regular ol' Joph fan club in here.

10,538 posts

Jophiel wrote:

Sir Xsarus wrote:

on a side note Silver is going to be back on the daily show on the 7th. I'm interested in watching that one.

Be more interesting if Romney is elected

You'll always be stupid, you'll just be stupid with more information in your brain

Forum FAQ

Tracer Bullet

12,636 posts

Almalieque wrote:

trickybeck wrote:

Your an moran.

Please define a "moran".

Congratulations, you identified 1 out of the 3 purposely incorrect words. You're 33% smart - that's math!

Na Zdrowie

10,538 posts

You'll always be stupid, you'll just be stupid with more information in your brain

Forum FAQ

Guru

1,877 posts

Almalieque wrote:

1/3 isn't 33%

trickybeck wrote:

Your an moran.

#swaggerjacker

GBATE!! Never saw it coming

9,297 posts

Nate Silver was on Bill Maher October 26.

Jophiel wrote:

Last week, I saw a guy with an eyepatch and a gold monocle and pointed him out to Flea as one of the most awesome things I've seen, ever. If I had an eyepatch and a gold monocle, I'd always dress up as Mr. Peanut but with a hook hand and a parrot.

The All Knowing

10,152 posts

Criminy wrote:

Almalieque wrote:

1/3 isn't 33%

trickybeck wrote:

Your an moran.

Pretty sure 1/3 is irrational. But, go ahead, prove me wrong..

Liberal Conspiracy

TILT

Almalieque wrote:

Pretty sure 1/3 is irrational. But, go ahead, prove me wrong..

Any port in a storm.

Belkira wrote:

Wow. Regular ol' Joph fan club in here.

Sage

5,684 posts

Almalieque wrote:

Criminy wrote:

Almalieque wrote:

1/3 isn't 33%

trickybeck wrote:

Your an moran.

Pretty sure 1/3 is irrational. But, go ahead, prove me wrong..

1 and 3 are both integers...

Almalieque wrote:

I admit that I was wrong

God bless Lili St. Cyr

Worst. Title. Ever!

16,884 posts

Almalieque wrote:

Pretty sure 1/3 is irrational. But, go ahead, prove me wrong..

You could at least make it hard...

Can't sleep, clown will eat me.

The All Knowing

10,152 posts

TirithRR wrote:

Almalieque wrote:

Pretty sure 1/3 is irrational. But, go ahead, prove me wrong..

You could at least make it hard...

You can get help with that. just sayin.

10,538 posts

This makes my day. Spend all day saying "the math doesn't allow that" and then demonstrate a **fundamental** lack of math knowledge. Wow.

You'll always be stupid, you'll just be stupid with more information in your brain

Forum FAQ

**B***i*__U__~~S~~
Smileys

**Post**

Anonymous Guests (64)

**Forums**- Cross Site
- The Asylum
- Final Stretch Election Thread

© 2017 ZAM Network LLC