Rachel Maddow Shows Stupidity - Page 2 - Politics and Economics

lol at you if you think she made a chart that appeared on her show. lol at you if you think any TV host creates any graphics that appear on their show

07-28-2015 , 10:02 AM

#28

BrianTheMick2

Long way to go and a short time to get there.

Join Date: May 2012 Posts: 19,424

Quote:

Originally Posted by Low Key

lol at you if you think she made a chart that appeared on her show. lol at you if you think any TV host creates any graphics that appear on their show

Nuh uh. She also writes all of her own jokes and runs the camera as well.

07-28-2015 , 10:06 AM

#29

SmokeyQ123

Carpal \'Tunnel

Join Date: Nov 2007 Posts: 6,583

Quote:

Originally Posted by David Sklansky

She thinks that those who are polling at 4% might actually be at 1% or 7%.

Quote:

Originally Posted by David Sklansky

Cliff notes is that she claimed that a candidate way out of the top 10 with perhaps 1% could actually easily be ahead of some ranked eighth or so with 5% because the poll had a "margin of error" of 3%.

Was the issue that she couldn't do basic arithmetic (as said in the second quote), or that she made a statistical error due to the 3% MOE not applying to someone who is polling at 4%? I know basically nothing about polling, however I've done enough stats and math to feel bad that I don't see what's wrong with the first sentence I quoted.

07-28-2015 , 10:24 AM

#30

Trolly McTrollson

Carpal \'Tunnel

Join Date: Oct 2010 Posts: 37,125

Quote:

Originally Posted by SmokeyJ

I've done enough stats and math to feel bad that I don't see what's wrong with the first sentence I quoted.

Old man yells at Gaussian distribution.jpg

07-28-2015 , 10:26 AM

#31

Jbrochu

banned

Join Date: Jan 2005 Posts: 16,054

Don't be hard on him. Even black kids get more sex than math nerds. That wears a man down.

07-28-2015 , 10:32 AM

#32

seattlelou

Carpal \'Tunnel

Join Date: Dec 2009 Posts: 31,584

The number of choices in the field makes the maths bad right? Assuming all the moe would go from one candidate to another? I suck at math be gentle.

07-28-2015 , 10:34 AM

#33

heehaww

Pooh-Bah

Join Date: Aug 2011 Posts: 5,081

Quote:

Originally Posted by Trolly McTrollson

Standard deviation, how does it work?

To determine how many standard deviations below the mean, add up all the negative squared deviations to get the negative variance, and then take the square root. That's what I learned from the "Poker Is Rigged" thread.

07-28-2015 , 10:41 AM

#34

SmokeyQ123

Carpal \'Tunnel

Join Date: Nov 2007 Posts: 6,583

Quote:

Originally Posted by Jbrochu

Don't be hard on him. Even black kids get more sex than math nerds. That wears a man down.

You seem pretty mad at David

Quote:

Originally Posted by Trolly McTrollson

Old man yells at Gaussian distribution.jpg

Ha. The confidence interval stays the same when approaching a boundary, you just apply it as normal and remove illegal ranges, right? I assume the problem here is the other participants in the race, I just realised that every stats problem I've done is very different from this kind of situation.

07-28-2015 , 10:47 AM

#35

AlexM

ignored

Join Date: Dec 2003 Posts: 20,655

Looks more like ignorance than stupidity to me.

Man publicly calls woman stupid for making an error in math. Now that's showing stupidity.

07-28-2015 , 10:50 AM

#36

Trolly McTrollson

Carpal \'Tunnel

Join Date: Oct 2010 Posts: 37,125

4% with a standard deviation of 1% is going to mostly look Gaussian. There's probably some alternative distribution that would be strictly correct, but if you just edit out illegal ranges, it's a reasonable approximation. I'm not even sure if that's what Dave is yelling about.

07-28-2015 , 10:57 AM

#37

Low Key

Carpal \'Tunnel

Join Date: Jan 2007 Posts: 33,327

"A lesbian read a cue card that may not have been perfectly cromulent," seems to be the point of this thread.

07-28-2015 , 11:09 AM

#38

seattlelou

Carpal \'Tunnel

Join Date: Dec 2009 Posts: 31,584

Quote:

Originally Posted by Low Key

"A lesbian read a cue card that may not have been perfectly cromulent," seems to be the point of this thread.

Loki assuming that a Rhodes scholar just reads cue cards. Nice.

07-28-2015 , 11:28 AM

#39

Lilu7

Made 2+2 Great Again

Join Date: Mar 2014 Posts: 2,123

"The margin of error is usually defined as the "radius" (or half the width) of a confidence interval for a particular statistic from a survey. One example is the percent of people who prefer product A versus product B. When a single, global margin of error is reported for a survey, it refers to the maximum margin of error for all reported percentages using the full sample from the survey. If the statistic is a percentage, this maximum margin of error can be calculated as the radius of the confidence interval for a reported percentage of 50%." -Wiki

07-28-2015 , 11:37 AM

#40

Jbrochu

banned

Join Date: Jan 2005 Posts: 16,054

Quote:

Originally Posted by Lilu7

Someone already provided a link.

"When a single, global margin of error is reported for a survey, it refers to the maximum margin of error for all reported percentages using the full sample from the survey."

https://en.m.wikipedia.org/wiki/Margin_of_error

Basically, you can't apply the full margin of error to every proportion reported.

Pretty sure this is what he's upset about. Or else it's something that flew over my head.

Edit: you ninja edited your post so my response doesn't make sense.

07-28-2015 , 11:42 AM

#41

Lilu7

Made 2+2 Great Again

Join Date: Mar 2014 Posts: 2,123

Hahahaha i knew i was going to get ripped a new one so i did some hmwk and edited

07-28-2015 , 11:44 AM

#42

Jbrochu

banned

Join Date: Jan 2005 Posts: 16,054

Quote:

Originally Posted by David Sklansky

Basically you take the square root of the number of people who selected the longshot candidate. So if 40 out of 1000 pick Christie two standard deviations is about 13 or 1.3%

Is this a shorthand method for estimating the standard error?

When I calculate 2 standard errors based on your numbers (40 out of 1000) it comes up as 1.24% -- so I assume that's what you're estimating?

07-28-2015 , 11:52 AM

#43

seattlelou

Carpal \'Tunnel

Join Date: Dec 2009 Posts: 31,584

Quote:

Originally Posted by seattlelou

The number of choices in the field makes the maths bad right? Assuming all the moe would go from one candidate to another? I suck at math be gentle.

This is pertinent?

Quote:

When comparing percentages, it can accordingly be useful to consider the probability that one percentage is higher than another.[12] In simple situations, this probability can be derived with 1) the standard error calculation introduced earlier, 2) the formula for the variance of the difference of two random variables, and 3) an assumption that if anyone does not choose Kerry they will choose Bush, and vice versa; they are perfectly negatively correlated. This may not be a tenable assumption when there are more than two possible poll responses. For more complex survey designs, different formulas for calculating the standard error of difference must be used.

07-28-2015 , 11:57 AM

#44

SmokeyQ123

Carpal \'Tunnel

Join Date: Nov 2007 Posts: 6,583

Is it right that each candidate % has a distribution that, given enough space and a decent sample size, looks roughly normal? Is the MOE calculated to apply to every candidate provided said normal distribution?

There's gotta be an smp joke in here

Last edited by SmokeyQ123; 07-28-2015 at 12:05 PM.

07-28-2015 , 12:06 PM

#45

seattlelou

Carpal \'Tunnel

Join Date: Dec 2009 Posts: 31,584

One of you mathies should calculated the probability that a candidate polling at 2 percent could be ahead of a 4 percent candidate in an 18 person poll with a 3 percent MOE. It feels remote to me but I have faith in you number bunnies to provide an answer.

07-28-2015 , 12:10 PM

#46

Jbrochu

banned

Join Date: Jan 2005 Posts: 16,054

This is why we shouldn't chase away all the nerds and racists...

07-28-2015 , 12:20 PM

#47

Trolly McTrollson

Carpal \'Tunnel

Join Date: Oct 2010 Posts: 37,125

Quote:

Originally Posted by seattlelou

If the MOE is +/-3%, there's about a 2.5% chance that the two percent guy is actually at >5% I think. I'd need to math it out a bit to get >4%.

07-28-2015 , 12:21 PM

#48

govman6767

Carpal \'Tunnel

Join Date: Sep 2006 Posts: 7,549

Quote:

Originally Posted by adios

Haha maybe she knew she was making an illogical/ignorant argument but she knew her audience was gullible enough to accept it.

Pretty standard for most of what comes out of her mouth...

I listen to her podcast every single day in the hopes of winning a cocktail shaker one of these weeks. But the fact that she bends the damn truth so much or falls into this rut where for 2 weeks straight the only god damn thing she talks about is bridgegate while the world is falling apart....Jesus it's life tilting..

But last week the winner of the weekly quiz got the bacon rub and the cocktail shaker....I shall win the cocktail shaker

07-28-2015 , 12:23 PM

#49

Jbrochu

banned

Join Date: Jan 2005 Posts: 16,054

How many 5 year old kids with chainsaws do you need in order to chop the 18 people in the poll into enough pieces in order for one candidate to be polling at 2%?

07-28-2015 , 12:23 PM

#50

AlexM

ignored

Join Date: Dec 2003 Posts: 20,655

Quote:

Originally Posted by seattlelou

Loki assuming that a Rhodes scholar just reads cue cards. Nice.

Uhm, everyone who gives daily performances on TV is reading from cue cards. Even if they wrote the cards themselves and know what they're talking about down cold, it's still going to be easier to read from a card.

Page 2 of 9

First

1 2 3 4 5 6 7

Last

...

Page 2 of 9

First

1 2 3 4 5 6 7

Last