g********2
发帖数: 6571 | 1 ACTUARIAL REVIEW: Analysis of Recent Polls Shows Trump Win and Possible
Landslide
Jim Hoft Oct 18th, 2016 10:02 am Leave a Comment
Guest post by Joe Hoft
trump clinton speeches
Ann Coulter warned in her book Slander in 2003 that the far left main stream
media always uses polls to push their agenda. Polls can be skewed by
selecting an unreasonable sample size, by asking lead up questions or by
selecting more of a sample population of one side of an issue to achieve a
desired result.
Main stream media skews polls to discourage potential voters from voting and
has done it for years.
A good example of the media trying to shape a vote was in 1980. In a Gallup
poll released on October 26th in 1980, two weeks before the election, Jimmy
Carter was leading Ronald Reagan 47 – 39. Two weeks later Reagan won in
such a landslide that Carter conceded before California was closed.
1980-elec-map
Another example of mainstream liberal media bias was in 1988. A Gallup Poll
from July 26 showed Michael S. Dukakis leading George H. W. Bush by 17
points. Of course Bush went on to thump Dukakis in the general election.
This past week a number of polls show Hillary Clinton leading Donald Trump
by various margins. One poll last week reported by NBC/WSJ showed Hillary
ahead by 11. However, Truthfeed pointed out that NBC/WSJ didn’t mention
that the poll was created by a Hillary Super PAC.
It’s also well known that the Monmouth University poll is run by a Hillary
Huckster who recently was caught manipulating a poll and then lied about it.
Realclearpolitics.com takes an average of these distorted polls to come up
with their analysis of the current race. Their efforts are a great example
of the phrase – ‘garbage in – garbage out’.
Expert Analysis
With all the liberal distortions and dishonesty we decided to have a small
team of actuarial and statistics professionals take a look at a couple of
the recent polls to get their take on the reliability of these polls. They
selected the recent FOX poll from October 14 showing Hillary up by 7 and the
WSJ/NBC poll from October 16 showing Hillary with an 11 point lead.
The first observation is that both polls are heavily skewed towards
Democrats. At a high level, the FOX poll consists of 43 Dems to 36 Reps to
21 Other while the NBC poll shows 44 Dems to 37 Reps to 19 Other.
By selecting more Dems the polls are designed to provide a Dem result.
Our experts next analyzed the data and calculated results using the same
data from the two surveys on a split of 40 Dems, 40 Reps and 20 Other. The
results show that using either sets of data Trump comes out ahead with a
larger margin of victory using the FOX data.
Clearly the polls using data that is heavily weighted towards Democrat
voters is incorrectly skewed.
This year Republicans crushed their previous record in the primaries for
number of votes by 150%. Their old record was 20 million and this year 31
million voted in the primaries.
The Democrats on the other hand had 7 million votes less than their record
year in 2008 with 30 million this year compared to 37 million in 2008.
Also, the primaries were heavily contested on both sides resulting in
factions from each party vowing not to vote for the party candidate. The
impact of these two groups is difficult to judge. The percentage of these
voters that change parties is probably limited. If anything, the Sanders
people will probably be more likely to vote for Trump since he is an
outsider and many of them will never vote for Crooked Hillary.
Finally, it is difficult to determine what the independent voters will do
but many independent voters partook in the primaries to vote for Trump.
Therefore it is more likely that Independent voters vote for Trump as well.
If more Democratic voters vote for Trump than Republicans vote for Hillary
and more Independents vote for Trump than Hillary, both scenarios which are
highly likely, then the results of the general election will likely be a
Trump landslide.
Let’s face it NOBODY likes Hillary and NOBODY trusts Hillary.
Nobody trusted Hillary before WikiLeaks. Trump is filling arenas around the
country but Hillary can barely fill half a high school gymnasium. Her VP
candidate Tim Kaine can’t even get 50 people at an event.
Obama had yard signs everywhere. Hillary has none. This is because
people are embarrassed to support lying Hillary.
— Bill Mitchell (@mitchellvii) October 17, 2016
Maybe the current polls are not just inaccurate – maybe they are dead wrong.
http://www.thegatewaypundit.com/2016/10/actuarial-review-of-recent-polls-shows-trump-win-and-possible-landslide/ |
g********2
发帖数: 6571 | |
l*****o
发帖数: 19235 | 3 看这段就行了
This year Republicans crushed their previous record in the primaries for
number of votes by 150%. Their old record was 20 million and this year 31
million voted in the primaries.
The Democrats on the other hand had 7 million votes less than their record
year in 2008 with 30 million this year compared to 37 million in 2008. |
H****t
发帖数: 1367 | 4 传给床铺看看，他就不再嚷嚷RIGGED。哈哈
stream
【在 g********2 的大作中提到】
: ACTUARIAL REVIEW: Analysis of Recent Polls Shows Trump Win and Possible
: Landslide
: Jim Hoft Oct 18th, 2016 10:02 am Leave a Comment
: Guest post by Joe Hoft
: trump clinton speeches
: Ann Coulter warned in her book Slander in 2003 that the far left main stream
: media always uses polls to push their agenda. Polls can be skewed by
: selecting an unreasonable sample size, by asking lead up questions or by
: selecting more of a sample population of one side of an issue to achieve a
: desired result.
|
l*****o
发帖数: 19235 | 5 the FOX poll consists of 43 Dems to 36 Reps to
21 Other while the NBC poll shows 44 Dems to 37 Reps to 19 Other.
都是D+7
选情还是焦灼状态，如果女克真领先很多，MSM就不会现在这样歇斯底里的撕老床了 |
s******l
发帖数: 6898 | 6 川普赢，民主党作弊，可能同时为真
【在 H****t 的大作中提到】
: 传给床铺看看，他就不再嚷嚷RIGGED。哈哈
:
: stream
|
a****a
发帖数: 3905 | 7 turnout不是精算师能预测的。
12年几个砸了牌子的都是多估了R的turnout。
brexit那么大误差也是turnout估计错误。
stream
【在 g********2 的大作中提到】
: ACTUARIAL REVIEW: Analysis of Recent Polls Shows Trump Win and Possible
: Landslide
: Jim Hoft Oct 18th, 2016 10:02 am Leave a Comment
: Guest post by Joe Hoft
: trump clinton speeches
: Ann Coulter warned in her book Slander in 2003 that the far left main stream
: media always uses polls to push their agenda. Polls can be skewed by
: selecting an unreasonable sample size, by asking lead up questions or by
: selecting more of a sample population of one side of an issue to achieve a
: desired result.
|
s*x
发帖数: 8041 | 8 有没有估算往年primary turn out 和最后大选turnout 的？今年primary 31m vs 30m
。 |
g********2
发帖数: 6571 | 9 This year Republicans crushed their previous record in the primaries for
number of votes by 150%. Their old record was 20 million and this year 31
million voted in the primaries.
The Democrats on the other hand had 7 million votes less than their record
year in 2008 with 30 million this year compared to 37 million in 2008.
【在 a****a 的大作中提到】
: turnout不是精算师能预测的。
: 12年几个砸了牌子的都是多估了R的turnout。
: brexit那么大误差也是turnout估计错误。
:
: stream
|
g******t
发帖数: 18158 | 10 RIGGED在差不太多的时候有用，好比拳击比赛要数点数决定胜负，裁判有搞花样的空间
。老婊砸眼看要被床铺KO了，再RIGGED也不起作用了
【在 H****t 的大作中提到】
: 传给床铺看看，他就不再嚷嚷RIGGED。哈哈
:
: stream
|
|
|
g********2
发帖数: 6571 | 11 你这个脑子真的有问题。
这篇明明指出媒体在用rig民调的手段rig大选，你一点都看不懂？
【在 H****t 的大作中提到】
: 传给床铺看看，他就不再嚷嚷RIGGED。哈哈
:
: stream
|
a****a
发帖数: 3905 | 12 primary是6个月前了。现在希婆拼命抹黑就是要打击R的积极性。R党内建制派也在扯后
腿 。
【在 g********2 的大作中提到】
: This year Republicans crushed their previous record in the primaries for
: number of votes by 150%. Their old record was 20 million and this year 31
: million voted in the primaries.
: The Democrats on the other hand had 7 million votes less than their record
: year in 2008 with 30 million this year compared to 37 million in 2008.
|
l*****o
发帖数: 19235 | 13 你这是闭这眼睛假装看不见老床Rally的人数？
【在 a****a 的大作中提到】
: primary是6个月前了。现在希婆拼命抹黑就是要打击R的积极性。R党内建制派也在扯后
: 腿 。
|
h*******n
发帖数: 8906 | 14 这六个月希拉里给Wikileaks 漏的尿布兜不住了 和党建制派和媒体中伤之类的影响在
wikileaks的email面前不值一提
: primary是6个月前了。现在希婆拼命抹黑就是要打击R的积极性。R党内建制派也
在扯后
: 腿 。
【在 a****a 的大作中提到】
: primary是6个月前了。现在希婆拼命抹黑就是要打击R的积极性。R党内建制派也在扯后
: 腿 。
|
a****a
发帖数: 3905 | 15 Rally人数只能作为铁杆选民的指标。
举个例子，老创在W I初选前拉力一样很浩大，但是该输还是输。
wikileaks爆料不少，但是缺乏一击致命的大料。
【在 l*****o 的大作中提到】
: 你这是闭这眼睛假装看不见老床Rally的人数？
|
n****u
发帖数: 229 | 16 Jack Posobiec @JackPosobiec 3h3 hours ago Manhattan, NY
Everyone thinking the Podesta emails are our October Surprise
This is just the appetizer
Jack Posobiec @JackPosobiec 3h3 hours ago Manhattan, NY
If you think the Podesta emails are bad
Wait til we release hers |
t*******d
发帖数: 12895 | 17 最近的十来次大选，
和党primary turn out高的全胜，
猪党好像有一两次高也不顶。
30m
【在 s*x 的大作中提到】
: 有没有估算往年primary turn out 和最后大选turnout 的？今年primary 31m vs 30m
: 。
|
r*****e
发帖数: 7853 | 18 Is this rigged?
stream
【在 g********2 的大作中提到】
: ACTUARIAL REVIEW: Analysis of Recent Polls Shows Trump Win and Possible
: Landslide
: Jim Hoft Oct 18th, 2016 10:02 am Leave a Comment
: Guest post by Joe Hoft
: trump clinton speeches
: Ann Coulter warned in her book Slander in 2003 that the far left main stream
: media always uses polls to push their agenda. Polls can be skewed by
: selecting an unreasonable sample size, by asking lead up questions or by
: selecting more of a sample population of one side of an issue to achieve a
: desired result.
|
D*******r
发帖数: 2323 | 19 你难道没有见过作弊也没考及格的傻叉？
【在 H****t 的大作中提到】
: 传给床铺看看，他就不再嚷嚷RIGGED。哈哈
:
: stream
|
g********2
发帖数: 6571 | 20 先说说你的“this”是指什么。
【在 r*****e 的大作中提到】
: Is this rigged?
:
: stream
|
|
|
a****a
发帖数: 3905 | 21 这次和往年不一样的地方是Trump本质上不是党内人士。他现在在和党内的支持率在80%
到85%之间已经徘徊很久了。最后的结果和poll的差别，一个是D+8的turnout model是
不是准确，一个是和党那最后的10%会不会最终归队。
【在 t*******d 的大作中提到】
: 最近的十来次大选，
: 和党primary turn out高的全胜，
: 猪党好像有一两次高也不顶。
:
: 30m
|
g********2
发帖数: 6571 | 22 这话说得很有pollster的风度了。
80%
【在 a****a 的大作中提到】
: 这次和往年不一样的地方是Trump本质上不是党内人士。他现在在和党内的支持率在80%
: 到85%之间已经徘徊很久了。最后的结果和poll的差别，一个是D+8的turnout model是
: 不是准确，一个是和党那最后的10%会不会最终归队。
|
p********e
发帖数: 6030 | 23
Dem不是投的人少，而是投给Sanders的都没算。
【在 l*****o 的大作中提到】
: 看这段就行了
: This year Republicans crushed their previous record in the primaries for
: number of votes by 150%. Their old record was 20 million and this year 31
: million voted in the primaries.
: The Democrats on the other hand had 7 million votes less than their record
: year in 2008 with 30 million this year compared to 37 million in 2008.
|
n**********n
发帖数: 984 | 24 精算师有没有考虑到死人投票，投票机吃票，邮差撕票？ |
h*******n
发帖数: 8906 | 25 这个只能靠turnout来overcomez了
【在 n**********n 的大作中提到】
: 精算师有没有考虑到死人投票，投票机吃票，邮差撕票？
|
l********7
发帖数: 2974 | 26 她大概不懂怎么问，应该问： Is this estimation biased?
【在 g********2 的大作中提到】
: 先说说你的“this”是指什么。
|
j******g
发帖数: 2689 | 27 在乱搞女人报告前川普还真的会赢 米切尔给床铺竞选钉了最重要的一枚棺材钉。 |
q***0
发帖数: 225 | 28 哎，又是不懂统计的人来说poll，不理解的就弄些阴谋论。阴谋论你爱信就信，说说
fact。
首先，文章里提到的两次选举都是几十年前，现在的sampling办法比那时好很多。其次
，文章里cherry pick 了两个数字而不是看所有的data。先说Carter-Reagan，poll确
实有误差，但投票前的poll是里根领先，而且那次选举流动性很大，随着伊朗人质事件
的拖延和恶化，卡特的支持率直线下降，而且每辩论一次，里根的支持率就涨一下，
poll变化很快。再说 Dukakis-Bush，文章 cherry pick 了七月份民主党convention刚
开过的一个poll来说事儿，离选举还有三个多月，当然不能反映最后投票了。拿出选举
前的poll看看，是准的。而且Dukakis 在campaign时犯了很多错误 |
j******g
发帖数: 2689 | 29 俺宁可相信谷歌大数据也不会相信这厮的精算。
在乱搞女人报告前川普还真的会赢 米切尔给床铺竞选钉了最重要的一枚棺材钉。 |
