d****i
发帖数: 4354 | 1 RCP(realclearpolitics.com)平均数还算比较靠谱，他们把左中右的民调取平均数，一
定程度上使有党派偏向的民调抵消掉。
目前Ohio RCP数字是48% O, 45.3% R
看起来似乎Obama赢面大，其实还要注意那其余的6.7%。因为根据以往选战总结出来的
铁律，如果在位总统不能达到50%，那么最后未决定的选民80%-90%会倒向挑战者一方。
这个不是我杜撰的，关注选战比较有经验的人都知道，屡试不爽。
所以如果照目前这个数字看，最后结果有可能会是O 48.5%，R 49.5%，其他候选人2%。
而且一般来说，共和党民调总是比最后结果偏低。
所以现在着急的是O阵营。 |
f*********g
发帖数: 1637 | 2 "因为根据以往选战总结出来的
铁律，如果在位总统不能达到50%，那么最后未决定的选民80%-90%会倒向挑战者一方。
这个不是我杜撰的，关注选战比较有经验的人都知道，屡试不爽。"
Just curious, any source link please? |
d****i
发帖数: 4354 | 3 http://www.pollingreport.com/incumbent.htm
【在 f*********g 的大作中提到】
: "因为根据以往选战总结出来的
: 铁律，如果在位总统不能达到50%，那么最后未决定的选民80%-90%会倒向挑战者一方。
: 这个不是我杜撰的，关注选战比较有经验的人都知道，屡试不爽。"
: Just curious, any source link please?
|
f*********g
发帖数: 1637 | 4 This is good link for discussion, it was published in 1989 though.
But let us check real case Ohio, in 2004, Bush lead in RCP average is only
48.8 vs 46.7, i.e. just 2.1 lead, he just alternatively lead in polls in
that year, but in final election, he won 50% over Gore 46.5%.
In current Ohio case, O lead Ohio most time, like 9 out of 10.
It doesn't mean your quoted study is wrong, just means it is too general and
, old history case study not modern enough, especially not for Ohio. |
d****i
发帖数: 4354 | 5 I am just showing you a source since you asked for it. This phenomenon is
still credited to this date. And there has been refinement to the theory.
Firstly 2004 was Bush/Kerry, not Gore.
Things to consider:
GOP tend to slightly under perform in polls.
Romney is leading by almost 1 point in popular vote poll. While Kerry was
behind quite a bit.
Bush approval rating was slightly better than Obama at the same time in 2004
and
【在 f*********g 的大作中提到】
: This is good link for discussion, it was published in 1989 though.
: But let us check real case Ohio, in 2004, Bush lead in RCP average is only
: 48.8 vs 46.7, i.e. just 2.1 lead, he just alternatively lead in polls in
: that year, but in final election, he won 50% over Gore 46.5%.
: In current Ohio case, O lead Ohio most time, like 9 out of 10.
: It doesn't mean your quoted study is wrong, just means it is too general and
: , old history case study not modern enough, especially not for Ohio.
|
f*********g
发帖数: 1637 | 6 I voted in 2000 and 2004, so I knew who was who.
You probably knew by now, the strange election system makes state polls more
relevant than nation polls.
So far for all swinging states, O leads R by about 0.5 points.
By the way, where do you get impression that "GOP tend to slightly under
perform in polls."? |
d****i
发帖数: 4354 | 7 I am glad you know who was who, just trying to help.
A candidate win by EC but lose popular vote if the popular vote is close
enough. 2000 Gore won popular vote by 48.4:47.9, but Bush won EC.
This year Romney is very likely to win popular vote by 2-3 points looking at
current polls, it is very unlikely that he will lose EC.
more
【在 f*********g 的大作中提到】
: I voted in 2000 and 2004, so I knew who was who.
: You probably knew by now, the strange election system makes state polls more
: relevant than nation polls.
: So far for all swinging states, O leads R by about 0.5 points.
: By the way, where do you get impression that "GOP tend to slightly under
: perform in polls."?
|
f*********g
发帖数: 1637 | 8 "This year Romney is very likely to win popular vote by 2-3 points looking
at current polls," fact?
" it is very unlikely that he will lose EC.," again, fact.
Note I am not asking your impression, just poll facts. |
d****i
发帖数: 4354 | 9 The fact is about one week in the future and projecting future result based
on poll is not a precise science, don't forget about that, things do happen.
Romeny win at least 2-3 points in popular vote: because Obama cannot break
47% in polls this close to election and Romney is leading at 47.9. There is
more upside to Romney.
There's only one sample in modern time was 2000 and I already shown the
popular vote difference was so close (half a percentage point). The burden
is on you to provide a counter if you want to discredit that.
【在 f*********g 的大作中提到】
: "This year Romney is very likely to win popular vote by 2-3 points looking
: at current polls," fact?
: " it is very unlikely that he will lose EC.," again, fact.
: Note I am not asking your impression, just poll facts.
|
f*********g
发帖数: 1637 | 10 I asked you "poll facts", you cant answer the simple question other than
your impression. As matter of fact all poll facts do not support your
impressions so far. Anyway I thought you may have some fact based assessment
that I missed but unfortunately you dont have one. Our discussion should
end here. Thanks for discussion. bye. |
|
|
d****i
发帖数: 4354 | 11 Then provide "all the poll facts" that discredit what I say, that popular
vote will be very close if EC winner and popular winner are not the same.
As for Romney winning popular vote by 2-3, that's a prediction, you can
disagree but we will have to wait for the result in 10 days or so.
assessment
【在 f*********g 的大作中提到】
: I asked you "poll facts", you cant answer the simple question other than
: your impression. As matter of fact all poll facts do not support your
: impressions so far. Anyway I thought you may have some fact based assessment
: that I missed but unfortunately you dont have one. Our discussion should
: end here. Thanks for discussion. bye.
|
h****g
发帖数: 2216 | 12 你要是真有兴趣可以在RCP上找篇报道，上面是对EC和PC相关性的分析。简单讲就是如
果候选人赢PC的比例（如果输了，那就是个负数，比如2000的布什）和在第270票的州
赢得比例之间的差距，这个差距我记得多数是小于1个点的。
如果RCP是正确的话 （Romney PC赢0.9），那么obama是有机会的。但是如果Romney真
的赢到了2个点或更多（假设Gallup那批是正确的话），那么obama现在在Ohio等摇摆州
的领先就全是幻觉。
我个人的猜测。如果现在就选，Obama有可能输PC （一个点以内），赢EC。但是第270
票的州的优势应该是相当小（接近布什2000年在FL的优势500多票）。
当然你要是对这些分析不敢兴趣，可以直接忽略我的发言。
【在 f*********g 的大作中提到】
: "This year Romney is very likely to win popular vote by 2-3 points looking
: at current polls," fact?
: " it is very unlikely that he will lose EC.," again, fact.
: Note I am not asking your impression, just poll facts.
|
T**********e
发帖数: 29576 | 13
270
如果R赢2%以上，输EC的机率在1%以下。
【在 h****g 的大作中提到】
: 你要是真有兴趣可以在RCP上找篇报道，上面是对EC和PC相关性的分析。简单讲就是如
: 果候选人赢PC的比例（如果输了，那就是个负数，比如2000的布什）和在第270票的州
: 赢得比例之间的差距，这个差距我记得多数是小于1个点的。
: 如果RCP是正确的话 （Romney PC赢0.9），那么obama是有机会的。但是如果Romney真
: 的赢到了2个点或更多（假设Gallup那批是正确的话），那么obama现在在Ohio等摇摆州
: 的领先就全是幻觉。
: 我个人的猜测。如果现在就选，Obama有可能输PC （一个点以内），赢EC。但是第270
: 票的州的优势应该是相当小（接近布什2000年在FL的优势500多票）。
: 当然你要是对这些分析不敢兴趣，可以直接忽略我的发言。
|
t*b
发帖数: 850 | 14 Rcp is useless
Average of all polls means nothing.
Look at which poll is the most trustable in 2008
Base on that poll is more valuable than average poll |
d****i
发帖数: 4354 | 15 RCP不是最好的，不过是左右派辩论时互相比较可以接受的一个数据来源。
【在 t*b 的大作中提到】
: Rcp is useless
: Average of all polls means nothing.
: Look at which poll is the most trustable in 2008
: Base on that poll is more valuable than average poll
|
d********f
发帖数: 43471 | 16 民主党人一般都投2票，顺便还帮附近墓地里的人投票，所以你们都歇了吧
【在 d****i 的大作中提到】
: RCP(realclearpolitics.com)平均数还算比较靠谱，他们把左中右的民调取平均数，一
: 定程度上使有党派偏向的民调抵消掉。
: 目前Ohio RCP数字是48% O, 45.3% R
: 看起来似乎Obama赢面大，其实还要注意那其余的6.7%。因为根据以往选战总结出来的
: 铁律，如果在位总统不能达到50%，那么最后未决定的选民80%-90%会倒向挑战者一方。
: 这个不是我杜撰的，关注选战比较有经验的人都知道，屡试不爽。
: 所以如果照目前这个数字看，最后结果有可能会是O 48.5%，R 49.5%，其他候选人2%。
: 而且一般来说，共和党民调总是比最后结果偏低。
: 所以现在着急的是O阵营。
|
d****i
发帖数: 4354 | 17 民主党作弊，这个已经都build-in到turn-out model里面了。
而且共和党不同程度上控制了不少摇摆州，比如Florida, Ohio，Wisconsin, PA,
Virginia, Michigan, Iowa, Nevada州长都是共和党。摇摆州中Colorado和New
Hampshire, North Carolina州长是民主党。
当然不是说共和党当州长就能保证选举完全公平，但是多少起点作用。
此外死人投票必须是过去选举卷宗里有死人的名字在。不要以为过去四年GOP都闲着呢
，好几个州，比如Ohio已经整理卷宗去掉了很多已经去世的选民名字。
还有voter ID的推行，虽然对这次选举影响没那么大，明年开始就会逐渐见效了。
【在 d********f 的大作中提到】
: 民主党人一般都投2票，顺便还帮附近墓地里的人投票，所以你们都歇了吧
|
B*******m
发帖数: 512 | 18
more
【在 f*********g 的大作中提到】
: I voted in 2000 and 2004, so I knew who was who.
: You probably knew by now, the strange election system makes state polls more
: relevant than nation polls.
: So far for all swinging states, O leads R by about 0.5 points.
: By the way, where do you get impression that "GOP tend to slightly under
: perform in polls."?
|
B*******m
发帖数: 512 | 19 LOL, 有些黑人老大妈 have voted since 1980, but still don't know who was who.
more
【在 f*********g 的大作中提到】
: I voted in 2000 and 2004, so I knew who was who.
: You probably knew by now, the strange election system makes state polls more
: relevant than nation polls.
: So far for all swinging states, O leads R by about 0.5 points.
: By the way, where do you get impression that "GOP tend to slightly under
: perform in polls."?
|
B*******m
发帖数: 512 | 20 有时间你可以查查Gallup and Rasmussen,还在看CNN和MSNBC的Poll?
assessment
【在 f*********g 的大作中提到】
: I asked you "poll facts", you cant answer the simple question other than
: your impression. As matter of fact all poll facts do not support your
: impressions so far. Anyway I thought you may have some fact based assessment
: that I missed but unfortunately you dont have one. Our discussion should
: end here. Thanks for discussion. bye.
|
