Cement trampoline (merged thread) [Archive]

a_greer

07-31-2004, 07:05 PM

Originally posted by DanMacMan
Well the first of the polling results pre/post convention are out. Surprisingly though, Kerry did not get a whopping 10 - 15 point jump in the numbers that some had predicted. Ouch.
http://pollingreport.com/images/NWKpostcon.GIF usualy after a convention, there is at least a 5 point bounce, no matter the party positions or polotics of the conventioning party. (3rd partys excluded as no one cares what Nader or any other fringe candidate has to say)

this is sad for the dems, but it is only one poll, if all of the major polls dhow the same next week, that will be a fun time to watch the dems pannic.

addabox

07-31-2004, 07:12 PM

You guys are great. Really. Keep it coming.

After Kerry wins, you can post about how the dems must be in disarray because he failed to get the 75% of the vote that you decided he had to get to consider himself really president.

a_greer

07-31-2004, 07:23 PM

Originally posted by addabox
You guys are great. Really. Keep it coming.

After Kerry wins, you can post about how the dems must be in disarray because he failed to get the 75% of the vote that you decided he had to get to consider himself really president. And when bush wins in a landslide, you guys are gonna cry "Florida 2000 take 2" or "(plug in token minority here) was dis-infranchised!" and when the supreme court turns you down again, you will then use dibold, and say that hackers broke the system and stuffed the preverbial ballot box for big company freindly bushies"

think I am streaching it, watch, even if bush wins 45 states, they will sue over the computer thing, I will put money on it.

addabox

07-31-2004, 07:40 PM

Cake

07-31-2004, 07:59 PM

Originally posted by a_greer
this is sad for the dems, but it is only one poll, if all of the major polls dhow the same next week, that will be a fun time to watch the dems pannic.
You don't get it.
The country is polarized. There is only approximately 8% of voters who are undecided.
With that small number, you're not going to get the kind of a bounce that has traditionally occured in the past.

And "landslide"? You are absolutly dreaming.:lol:

Paul

07-31-2004, 08:06 PM

bingo...
much smaller % of minds that can be changed=less "bounce" available...

a_greer

07-31-2004, 08:28 PM

Originally posted by Cake
And "landslide"? You are absolutly dreaming.:lol: I was saying IF, IF it is a landslide, the dems will try tocomplain about the computers, call it a hunch

Wrong Robot

07-31-2004, 08:34 PM

I'm still wondering, greer, are you even old enough to vote?

BRussell

07-31-2004, 08:41 PM

Originally posted by DanMacMan
Well the first of the polling results pre/post convention are out. Surprisingly though, Kerry did not get a whopping 10 - 15 point jump in the numbers that some had predicted. Ouch. A challenger is beating an incumbent president 49-42, and you say ouch? I wonder if any incumbent has ever come back from a deficit like that at this point in time. I doubt it. When there's an incumbent, undecided voters go overwhelmingly for the challenger, so these polls understate the difference.

Anything can happen, but if you had to guess at this point in time who would win, you'd have to say Kerry.

Towel

07-31-2004, 09:55 PM

On Thursday, the NYT actually went to the trouble of investigating this "average 15 point bounce" claim by Matthew Dowd and other Republicans. It's based on two shaky bits: they're talking about the swing in poll numbers, not the increase in people voting for you (so Kerry's would be a 5-pont bounce, not a 2-point bounce, based on your polls); and they count the massive bounce Clinton got in 1992, when Ross Perot dropped out of the race (temporarily) at the same time the Democrats had their convention. A challenger almost always gets a bigger bounce than an incumbent, but usually because he's far behind to begin with. Kerry was already ahead.

And then there's the whole polarized electorate thing. No candidate in this race is ever going to poll higher than 50-52%, and most people who are sure they're going to vote have already made up their minds as to for whom.

Finally, I'd like to see a poll done this week, rather than an "overnight" one dated to the very day of Kerry's speech (the 29th).

pfflam

07-31-2004, 10:53 PM

Originally posted by Towel
most people who are sure they're going to vote have already made up their minds as to for whom. That has to have been the case with Repubs-who are for Bush for ever . . . otherwise there is no explanation as to why they would vote for him . . . its all about some fantasy image of what he is . . . because it sure as hell has nothing to do with his record:

The CHURCH of BUSH (http://www.villagevoice.com/issues/0429/perlstein.php)
Good article about how people adore him, and their so called 'reasons' . . . real examples to...

FormerLurker

07-31-2004, 11:21 PM

I enjoyed the last two lines of the article:
Conservatives see something angelic in George Bush. That's why they excuse, repress, and rationalize away so much. And that is why conservatism is verging on becoming an un-American creed.
So sad, and yet so true.

addabox

07-31-2004, 11:26 PM

Originally posted by a_greer
I was saying IF, IF it is a landslide, the dems will try tocomplain about the computers, call it a hunch

Actually, people who are knowledgeable about the paperless voting machines are complaining about them now.

The big problem is they have no paper verification. No recount is possible. You just have to trust that the machine recorded your vote correctly. And because the source code is a corporate secret, there is no way for governments to verify that there isn't any malicious code.

Oh, and surprise! The Bush admin, Jeb in Florida and republicans across the nation are fighting tooth and nail against a relatively inexpensive add-on that would provide a voter verifiable paper trail on the existing machines.

And it turns out that Diebold's machine programs were on an unsecured FTP site, and the memory cards used in the machines were unencrypted. Oh, and in Georgia, all the cards had the same password. Which means if I could get my hands on some of those cards, I could change the tallies on my laptop without any evidence that it had happened.

Which are some of the same reasons that California is suing Diebold for fraud and breach of contract.

Anyone who cares about democracy should be deeply concerned about the potential for foul play in a system that doesn't allow for independent verification. Which of course means a_greer thinks it's swell.

BRussell

07-31-2004, 11:40 PM

Voting Machine Controversy (http://www.commondreams.org/headlines03/0828-08.htm)COLUMBUS - The head of a company vying to sell voting machines in Ohio told Republicans in a recent fund-raising letter that he is "committed to helping Ohio deliver its electoral votes to the president next year.

The Aug. 14 letter from Walden O'Dell, chief executive of Diebold Inc. - who has become active in the re-election effort of President Bush - prompted Democrats this week to question the propriety of allowing O'Dell's company to calculate votes in the 2004 presidential election.

ShawnJ

08-01-2004, 12:42 AM

Newsweek took half the poll before Kerry gave his speech, but BRussell nailed it anyway.

applenut

08-01-2004, 12:55 AM

no one who was undecided watched the frickin convention anyways. people who liked kerry watched.

this whole convention thing is a waste of money and an embarrassment. talk about the poor and then go spend nearly 100 million on a fucking week long pat on the back.

Anders

08-01-2004, 12:55 AM

Regarding electronic voting:

Use a couple of hours to see this. (rtsp://video.c-span.org/project/c04/c04070704_esystems.rm) After seeing it I would not feel sure in a pure electronic system, but to use it for assisting in the voting process it would serve a good purpose. You use the computer to make your choice and the machine make it clear for you what you voted for (name, picture whatever.) and print the ballot (not a receipt) for you. That would be the way ahead with the use of computers in the voting process.

Aquatic

08-01-2004, 01:02 AM

Voting machines should NEVER EVER be closed-source. EVER. That is one thing Kucinich hit hard upon when he came to my town, and I liked that. This is one thing that can NOT be let slide. I'll tell ya why. In NH, Republicans paid telemarketers to jam phone banks for an agency that helped people without transportation to get to vote. If they do that, why not tamper with e-voting? It's even easier and harder to prove. Diebold sucks asscheese.

ShawnJ

08-01-2004, 01:26 AM

Originally posted by applenut
this whole convention thing is a waste of money and an embarrassment. talk about the poor and then go spend nearly 100 million on a fucking week long pat on the back.

Interesting point about conventions. But still, 100 million to get out the message, rally the faithful, and try to convince those straddling the fence-- I think I can accept that if it means the difference between the goals of the Kerry administration and another half-trillion dollar tax cut or expensive, unnecessary war. The cost is high, but the benefits could be worthwhile-- *if* the convention helps Kerry, which does seem so.

applenut

08-01-2004, 01:40 AM

Originally posted by ShawnJ
Interesting point about conventions. But still, 100 million to get out the message, rally the faithful, and try to convince those straddling the fence-- I think I can accept that if it means the difference between the goals of the Kerry administration and another half-trillion dollar tax cut or expensive, unnecessary war. The cost is high, but the benefits could be worthwhile-- *if* the convention helps Kerry, which does seem so.

i really don't see any evidence of that. a few percentage points in a meaningless poll.

the convention in the perspective of "getting the message out" was a failure. the major networks barely gave it notice, ratings were an all-time low and its not like the people who actually made the effort to tune in are the people kerry needs to win.

the faithful are already rallied, the message is still not out there, and no one straddling the fence watched.

unfortunately i think some people here are giving the general voting population too much credit. issues are not what will win. appeal will.

Anders

08-01-2004, 01:59 AM

In a perfect world the dems would have taken that 100.000.000 and prevented hundreds of 911´s (http://allafrica.com/stories/200407290858.html) and made one commercial explaining why there wasn´t a convention this year.

But one of the reasons to have a convention (now that the election of JK wasn´t a question) is to rally the troops and say "Look how many we are. You are not alone. Now go out and sell the message". The message is not only for the viewers but also for the volunteers.

Paul

08-01-2004, 04:26 AM

http://www.electoral-vote.com/graph.png (http://www.electoral-vote.com/info/graph.html)
the race isn't as close as national polls indicate...

the graph would be better if it took the margin into account in the swing states tho...

From here:Electoral-Vote (http://www.electoral-vote.com/)

Common Man

08-01-2004, 08:03 AM

Come election day Bush will win big. My prediction is Bush 55%, Kerry 45%. Kerry is going to drop fast. Mainstream America does not identify with him. His campaign is based on creating class warfare.

groverat

08-01-2004, 10:22 AM

Originally posted by Common Man
His campaign is based on creating class warfare.

I kind of agree with this. Bush's campaign is based on convincing the poor that there is not a class war.

I hate both of these guys so much I think I'm going to vote for anarchy and see how many polling centers I can sabotage (non-violently, of course).

(Hello FBI and DOJ, I am just kidding. I'm not a terrorist. Don't sing at me!)

groverat

08-01-2004, 10:27 AM

I love that Village Voice article. It is great stuff.

Conservatives are at the point now that pointing out any criticism is a very reluctant and painful move because they know (and not even deep down) that once they start listing the things that make him a bad president they aren't likely to stop in time to save their Bush-love house of card from tumbling.

bunge

08-01-2004, 12:23 PM

Originally posted by a_greer
think I am streaching it, watch, even if bush wins 45 states, they will sue over the computer thing, I will put money on it.

No matter who wins I'm in favor of suing over computerized voting.

Towel

08-01-2004, 01:32 PM

Originally posted by applenut
the convention in the perspective of "getting the message out" was a failure. the major networks barely gave it notice, ratings were an all-time low and its not like the people who actually made the effort to tune in are the people kerry needs to win.Everyone's been saying that network ratings were at an all time low, especially the networks. But in reality, at least 24 million (http://www.freerepublic.com/focus/f-news/1182121/posts) people watched on Thursday night, three million more than saw Al Gore's speech in 2000. What's more, these early ratings numbers don't even include PBS, who reported 3 million viewers of their own for Monday and Tuesday nights, or CSPAN. So when the final numbers come out, I wouldn't be surprised if 30 million Americans saw Kerry's address, a big bounce from four years ago.

The networks got the shaft because they chose to. I mean, holy self-fulfilling prophecy. Meanwhile, the cable news networks and PBS all reported huge increases in viewership, even on the early nights.

Splinemodel

08-01-2004, 10:39 PM

Originally posted by a_greer
And when bush wins in a landslide, you guys are gonna cry "Florida 2000 take 2" or "(plug in token minority here) was dis-infranchised!" and when the supreme court turns you down again, you will then use dibold, and say that hackers broke the system and stuffed the preverbial ballot box for big company freindly bushies"

think I am streaching it, watch, even if bush wins 45 states, they will sue over the computer thing, I will put money on it.

Florida 2000, eh?

I am soooooo tempted to put my presidential vote on eBay.

I don't think I could stoop that low, though, because my party actually has an agenda backed by ideals, and I'm not going to sacrifice my vote to help some jerk who kowtows to a self-inflicted baseline of "moderation" or "values" or whatever the fuck they say matters.

That was fun to write. Anyway, vote 3rd party. . . any of them. Otherwise you're throwing away your vote.

DanMacMan

08-02-2004, 07:30 AM

More numbers, this time from Gallup/USA Today/CNN, published 8/1;

"Suppose that the presidential election were being held today, and it included John Kerry and John Edwards as the Democratic candidates, and George W. Bush and Dick Cheney as the Republican candidates._ Would you vote for John Kerry and John Edwards, the Democrats, or George W. Bush and Dick Cheney, the Republicans?" If undecided: "As of today, do you lean more toward Kerry and Edwards, the Democrats, or Bush and Cheney, the Republicans?"

7/30-31
Bush/Cheney - 50
Kerry/Edwards - 47
Neither - 1

7/19-21
Bush/Cheney - 47
Kerry/Edwards - 49
Neither - 2

hardeeharhar

08-02-2004, 09:21 AM

and the margin my dear friends is 3%...

this is shit.

you have half the nation thinking that the US should swing its power around like a dirty cock (a dirty cock is always more likely to catch a venereal disease) and the other half thinking that the power should be wielded with a little (but not much) more thought than that.

when you go to the polls think of the us' cock and how painful venereal diseases are... ok?

Common Man

08-02-2004, 12:40 PM

The trend will continue to move in favor of President Bush., By election day there will be no challenge.

Bush 55%
Kery 45%

hardeeharhar

08-02-2004, 01:08 PM

Common Man said he was going away away away.

You can never kill the common man for if you do you become common.

Common Man cannot vote.

Common Man is amazingly omnipresent.

Seeing all, yet seeing nothing.

What happens on Dec. 3rd?

Wait and see, I say.

DanMacMan

08-03-2004, 08:04 AM

Originally posted by hardeeharhar
What happens on Dec. 3rd?
What exactly is the significance of December 3rd? The election is November 2nd. By the way, some more poll numbers;

http://pollingreport.com/images/postconv.GIF

SDW2001

08-03-2004, 09:27 AM

According to this CNN/USA Today/Gallup Poll, Kerry/Edwards experienced a negative bounce after the convention.

Poll Article (http://usatoday.printthis.clickability.com/pt/cpt?action=cpt&title=USATODAY.com+-+Poll%3A+No+boost+for+Kerry+after+convention&expire=&urlID=11186062&fb=Y&url=http%3A%2F%2Fwww.usatoday.com%2Fnews%2Fpolitic selections%2Fnation%2Fpresident%2F2004-08-01-poll-kerry_x.htm&partnerID=1660)

The explanation that the electorate is so polarized is a possible one. I also know that the first thing out of a lot folks mouths on this story is going to be "The Republicans are just trying to create unrealistic expectations to make themselves look better". OK, that may be the case, but a 1 point drop, while the opponent gains four points? Something doesn't seem right to me there.

Even the Newsweek poll, which paints a much rosier picture for Kerry, admits that the bounce is the smallest in the history of the poll. Polarized electorate or bad omen for Kerry?

History of convention bounces: Source: Gallup

Year Candidate Bounce

2000 Gore +8
2000 Bush +4 Winner

1996 Clinton +5 Winner
1996 Dole +3

1992 Clinton +16 Winner
1992 Bush +5

1988 Dukakis +7
1988 Bush +6 Winner

1984 Mondale +9
1984 Reagan +5 Winner

1980 Carter +10
1980 Reagan +8 Winner

1976 Carter +9 Winner
1976 Ford +5

1972 Nixon +7 Winner
1972 McGovern +0

1968 Nixon +5 Winner
1968 Humphrey +2

1964 Goldwater +5
1964 Johnson +3 Winner

__________________________________________________ ___
Losses by the incumbent:

1976: Carter defeats Ford. Carter gets 9 point bounce (Ford=+5)
1980: Regan defeats Carter. Reagan gets 8 point bounce (Carter=+10)
1992: Clinton defeats Bush. Clinton gets 16 point bounce (Bush=+5)

Losses by the challenger:

1964: Johnson defeats Goldwater. Goldwater gets +5 bounce

1972: Nixon defeats McGovern. McGovern gets +0 bounce

1984: Reagan defeats Mondale. Mondale gets +9, still loses.

1996 Clinton defeats Dole. Dole gets +3 bounce

I think the above data can support some conclusions in particular: No candidate who received a negligable post-convention bounce or was "out bounced" by a large margin has been able to win. The possible exception is Bush 43 himself, where Gore got twice the bounce.

If we just look at challengers to an incumbent, the picture is worse for Kerry. No challenger has won without a large bounce. Period. History predicts that Kerry will not and cannot win. Then again, Bush has defied history, so why can't Kerry? Also, it should be said that a large bounce by the incumbent does not equal victory. My argument is that the challenger must experience a substantial bounce to win.

Regardless of who you support, what do you think about the above?

BRussell

08-03-2004, 09:54 AM

I'd like to see not just "bounces" but also absolute numbers.

My guess is that in most cases, the incumbent has been ahead of the challenger - sometimes way ahead, like Bush v. Clinton. There, the bounce was critical for the challenger to take the lead. This year, Kerry has been tied or ahead of Bush in most polls, and so the bounce really isn't needed to get ahead.

I don't remember where I read it, but I don't think that any incumbent has ever won when the polls were where they have been for Bush's approval ratings and the Bush v. Kerry numbers.

All I know is that I'm going to love every minute of it for the next 3 months.

SDW2001

08-03-2004, 10:07 AM

Originally posted by BRussell
I'd like to see not just "bounces" but also absolute numbers.

My guess is that in most cases, the incumbent has been ahead of the challenger - sometimes way ahead, like Bush v. Clinton. There, the bounce was critical for the challenger to take the lead. This year, Kerry has been tied or ahead of Bush in most polls, and so the bounce really isn't needed to get ahead.

I don't remember where I read it, but I don't think that any incumbent has ever won when the polls were where they have been for Bush's approval ratings and the Bush v. Kerry numbers.

All I know is that I'm going to love every minute of it for the next 3 months.

A good point, though the polls are pretty deadlocked, and sometimes Bush has lead. Bush's approval rating also hovers near 50%, but sometimes at 47%, which is hard to draw a conclusion from.

dmz

08-03-2004, 10:24 AM

Kerry is a dork. He looks like a dork, talks like a dork: full of sound and fury, signifying nothing.

What does this idiot stand for except some intellectually deprived, Michel More -esque pretzle logic?

More half-warmed, status quo rehtoric.

He has 0 chance of winning, even with the press spinning as hard as they can, all the Dickead Clarke catharsis trips, and some of the best smear that Hollywood can produce.

Anyone remember Bob Dole?

Anders

08-03-2004, 10:36 AM

Originally posted by dmz

He has 0 chance of winning,

Bookmarked :)

Towel

08-03-2004, 10:40 AM

According to Gallup, despite the lack of a "bounce", Kerry and the DNC were well received (http://gallup.com/content/?ci=12580). More people claim to have watched at least some of it than any of the three previous conventions; fully 74% of self-described Democrats, and even 55% of Independents and 46% of Republicans watched at least "some". The numbers for "more likely to vote for him afterwards" are almost identical to those for Bush and Gore in 2000 (44% more, 30% less). 52% of viewers thought Kerry's speech was "Excellent" or "Good", with more Excellent marks than either Bush or Gore's 2000 address. Perhaps most importantly, Independents were evenly split betwen thinking thaat Convention went too far with its partisan attacks or not far enough, with the majority thinking it was about right. All in all, I'd have to agree with all the non-GOP pundits about the conlfuence of factors that created the concrete trampoline; people not liking Kerry or the DNC is not among them.

Anders

08-03-2004, 10:52 AM

One thread about this topic must be enough.

BRussell

08-03-2004, 11:02 AM

Originally posted by dmz
Anyone remember Bob Dole? I remember Bob Dole. Clinton had an average 10-12 point lead over him for the entire election season. That's been typical for challengers who have lost to incumbents. On the other hand, Kerry has been deadlocked with Bush, and if anything, slightly ahead.

http://www.cnn.com/ALLPOLITICS/1996/polls/cnn.usa.gallup/tracking/09.02-12.jpg

Paul

08-03-2004, 11:51 AM

Originally posted by Paul
http://www.electoral-vote.com/graph.png (http://www.electoral-vote.com/info/graph.html)
the race isn't as close as national polls indicate...

the graph would be better if it took the margin into account in the swing states tho...

From here:Electoral-Vote (http://www.electoral-vote.com/)
(the image above is "live" updating, take a look at the bounce)...
Thus the states to watch are Minnesota, Wisconsin, Iowa, Missouri, Michigan, Ohio, Pennsylvania, and Florida. Forget the national polls. The popular vote doesn't matter, as we saw in 2000. It is these eight states and another half dozen or so smaller battleground states where the action will be.
after playing with some functions from this (http://www.electoral-vote.com/aug/aug03.xls) excel file I made a "national" ratio between the candidates...

Kerry 45.94%
Bush 45.22%
Nader 1.99%
Other/ Undecided 6.84%

From the day before:

Kerry 45.72%
Bush 45.05%
Nader 1.99%
Other/ Undecided 7.24%

Change:

Kerry 0.22%
Bush 0.17%
Nader 0%
Other/ Undecided -0.40%

So, there was a "bounce" but it was pretty balanced for both candidates when the whole country* is taken into account.

(*well, at least the changes in AR, FL, IA, MI, MN, MO, NV, NH, NM, OH, OR, PA, TN, WA, WV, and WI)

SDW2001

08-03-2004, 01:07 PM

Originally posted by Towel
According to Gallup, despite the lack of a "bounce", Kerry and the DNC were well received (http://gallup.com/content/?ci=12580). More people claim to have watched at least some of it than any of the three previous conventions; fully 74% of self-described Democrats, and even 55% of Independents and 46% of Republicans watched at least "some". The numbers for "more likely to vote for him afterwards" are almost identical to those for Bush and Gore in 2000 (44% more, 30% less). 52% of viewers thought Kerry's speech was "Excellent" or "Good", with more Excellent marks than either Bush or Gore's 2000 address. Perhaps most importantly, Independents were evenly split betwen thinking thaat Convention went too far with its partisan attacks or not far enough, with the majority thinking it was about right. All in all, I'd have to agree with all the non-GOP pundits about the conlfuence of factors that created the concrete trampoline; people not liking Kerry or the DNC is not among them.

That's not what either of these threads was about. The convention didn't even get good ratings. I honestly don't think many people were paying attention.

trumptman

08-03-2004, 02:47 PM

Originally posted by BRussell
I remember Bob Dole. Clinton had an average 10-12 point lead over him for the entire election season. That's been typical for challengers who have lost to incumbents. On the other hand, Kerry has been deadlocked with Bush, and if anything, slightly ahead.

http://www.cnn.com/ALLPOLITICS/1996/polls/cnn.usa.gallup/tracking/09.02-12.jpg

Man, Clinton had such a lead in that election. Dole couldn't excite the base and Clinton had SOOO much money. I used to describe his commercial runs as carpet bombings because they were on every channel 20 times a day.

Nick

Anders

08-03-2004, 02:52 PM

Originally posted by Paul
(the image above is "live" updating, take a look at the bounce)...

after playing with some functions from this (http://www.electoral-vote.com/aug/aug03.xls) excel file I made a "national" ratio between the candidates...

Kerry 45.94%
Bush 45.22%
Nader 1.99%
Other/ Undecided 6.84%

From the day before:

Kerry 45.72%
Bush 45.05%
Nader 1.99%
Other/ Undecided 7.24%

Change:

Kerry 0.22%
Bush 0.17%
Nader 0%
Other/ Undecided -0.40%

So, there was a "bounce" but it was pretty balanced for both candidates when the whole country* is taken into account.

(*well, at least the changes in AR, FL, IA, MI, MN, MO, NV, NH, NM, OH, OR, PA, TN, WA, WV, and WI)

Unless al the polls were made by the same company as the previous in all states using excatly the same method what you found out doesn´t mean anything.

faust9

08-03-2004, 03:17 PM

Originally posted by Anders
Unless al the polls were made by the same company as the previous in all states using excatly the same method what you found out doesn´t mean anything.

Untrue. Metadata can be used successfully. Here, the polling questions before and after probably didn't change (very unlikely). These polls show general sentiment which can be evaluated. Combining all the for Kerry, for Bush and For Nadar sentiment is not scientifically unsound. If the sum of the polling data showing pro candidate sentiment before and after the convention is used one can safely show there was an increase in candidate sentiment after the DNC.

Now, I don't know if the correct statistical analysis where used to generate the .22% increase but the idea that all data collected must come from the same source is not true. In fact, scientists rely on data collected from various organizations all the time to draw conclusions.

Towel

08-03-2004, 03:32 PM

Originally posted by SDW2001
That's not what either of these threads was about. The convention didn't even get good ratings. I honestly don't think many people were paying attention. I might gently suggest that you aren't paying attention to my posts. :) Ratings (http://www.abcnews.go.com/wire/Entertainment/ap20040730_1718.html) were up from four years ago. Yes, fewer people watched on the 3 networks, but they asked for it by hardly broadcasting anything. Cable news viewership doubled - doubled!, and PBS reported a 20% increase from four years ago. CSPAN hasn't reported its ratings yet, but they were surely higher, too. All together, despite fewer people watching the Big 3, the overall number of viewers was up at least 20% from four years ago.

And take a look at those Gallup numbers again. Well over half of all those polled reported watching "some" or "a lot" of the convention, and it wasn't just the Democrats.

I think people are paying attention. The "liberal" networks seem to be doing their best to downplay that for some reason, though.

Anders

08-03-2004, 04:01 PM

Originally posted by faust9
Untrue. Metadata can be used successfully. Here, the polling questions before and after probably didn't change (very unlikely). These polls show general sentiment which can be evaluated. Combining all the for Kerry, for Bush and For Nadar sentiment is not scientifically unsound. If the sum of the polling data showing pro candidate sentiment before and after the convention is used one can safely show there was an increase in candidate sentiment after the DNC.

Now, I don't know if the correct statistical analysis where used to generate the .22% increase but the idea that all data collected must come from the same source is not true. In fact, scientists rely on data collected from various organizations all the time to draw conclusions.

Sorry but you are wrong. Look at the general polls. Some of them show a certain leaning all the way like right now where some poll continue to have Bush as the leader while others have Kerry. If you look at the polls over time they might all show the same tendency (a shift of x point in favor of candidate Y). But when you compare the numbers from poll X at time Y with poll Z at time Q you are comparing oranges with apples. Even if, say, all polls show a drop in Kerry vote you could get the result that Kerry gained a lot if the first poll company made low-kerry leaning polls and the second made high-kerry leaning polls.

In short: Never use two polls made from different companies to say anything about developemnt over time.

faust9

08-03-2004, 04:27 PM

Originally posted by Anders
Sorry but you are wrong. Look at the general polls. Some of them show a certain leaning all the way like right now where some poll continue to have Bush as the leader while others have Kerry. If you look at the polls over time they might all show the same tendency (a shift of x point in favor of candidate Y). But when you compare the numbers from poll X at time Y with poll Z at time Q you are comparing oranges with apples. Even if, say, all polls show a drop in Kerry vote you could get the result that Kerry gained a lot if the first poll company made low-kerry leaning polls and the second made high-kerry leaning polls.

In short: Never use two polls made from different companies to say anything about development over time.

No I'm not wrong because combining the various polls produces a single figure showing the sum of sentiment across various polls. The sum can be compared before and after because the various pooling agencies are not likely to change methodology during a presidential election season. It doesn't matter that poll A has a different value than poll B because both polls end up being statistical points anyway.

What one has to do is find the standard deviation of these various polls and determine is 70% of the polls fall within one SD to say the data is statistically valid. Was this done? I don't know. Additionally, is the .22% inside or outside the SD range? I don't know. .22% is easily within the error range thus making the figure circumspect but the combining of polls is not wrong here. Skewed data points are compensated for during a correctly performed analysis. Also, the likelihood of one poll showing Kerry 20 points ahead of Bush is slim to none. These polls all generate numbers in a tight range sop the fear of one point trowing the entire result into disarray is false.

Now as an exercise you could play with your own scenario where one poll is significantly higher then the others. See how much of a change is required to shift the SD significantly when grouped with 50 others. See how much one data point must change to shift a pack of data points 0.22%. If the statistical analysis is done correctly then the shift required for one point to skew the rest significantly is fairly large.

I'm not saying the 0.22% change is a good data point, I'm countering your idea that data need come from the same source to build a statistical model. It doesn't.

Also if you compare Poll A+ Poll B at time X to Poll A + Poll B at time Y you'll compensate for methodology differences between the two polling agencies. You can't compare Poll A time X to poll B time Y. I don't think the latter was done. I think the first method was used.

SDW2001

08-03-2004, 04:37 PM

Originally posted by Towel
I might gently suggest that you aren't paying attention to my posts. :) Ratings (http://www.abcnews.go.com/wire/Entertainment/ap20040730_1718.html) were up from four years ago. Yes, fewer people watched on the 3 networks, but they asked for it by hardly broadcasting anything. Cable news viewership doubled - doubled!, and PBS reported a 20% increase from four years ago. CSPAN hasn't reported its ratings yet, but they were surely higher, too. All together, despite fewer people watching the Big 3, the overall number of viewers was up at least 20% from four years ago.

And take a look at those Gallup numbers again. Well over half of all those polled reported watching "some" or "a lot" of the convention, and it wasn't just the Democrats.

I think people are paying attention. The "liberal" networks seem to be doing their best to downplay that for some reason, though.

OK, and I will suggest that you realize you're talking two different issues. Kerry did not get a significant bounce. That's the point. I don't care how many people indicate it was "well-received". That doesn't mean they are going to vote for him.

SDW2001

08-03-2004, 04:42 PM

Originally posted by faust9
No I'm not wrong because combining the various polls produces a single figure showing the sum of sentiment across various polls. The sum can be compared before and after because the various pooling agencies are not likely to change methodology during a presidential election season. It doesn't matter that poll A has a different value than poll B because both polls end up being statistical points anyway.

What one has to do is find the standard deviation of these various polls and determine is 70% of the polls fall within one SD to say the data is statistically valid. Was this done? I don't know. Additionally, is the .22% inside or outside the SD range? I don't know. .22% is easily within the error range thus making the figure circumspect but the combining of polls is not wrong here. Skewed data points are compensated for during a correctly performed analysis. Also, the likelihood of one poll showing Kerry 20 points ahead of Bush is slim to none. These polls all generate numbers in a tight range sop the fear of one point trowing the entire result into disarray is false.

Now as an exercise you could play with your own scenario where one poll is significantly higher then the others. See how much of a change is required to shift the SD significantly when grouped with 50 others. See how much one data point must change to shift a pack of data points 0.22%. If the statistical analysis is done correctly then the shift required for one point to skew the rest significantly is fairly large.

I'm not saying the 0.22% change is a good data point, I'm countering your idea that data need come from the same source to build a statistical model. It doesn't.

Also if you compare Poll A+ Poll B at time X to Poll A + Poll B at time Y you'll compensate for methodology differences between the two polling agencies. You can't compare Poll A time X to poll B time Y. I don't think the latter was done. I think the first method was used.

I advise looking a variety of polls. The most optimistic shows a four point bounce for Kerry. The most pessimistic shows a 1 point drop. Either way I don't think it's good for Kerry, even taking into account the polarized electorate.

Northgate

08-03-2004, 05:52 PM

Zogby has released polls in 16 battleground states. As usual in such a tight race, there is good news and bad news for all the candidates (except Nader, for whom all the news is bad--he is dropping practically everywhere). First the good news for Kerry. He is ahead, albeit by small margins, in 13 of the key 16 states in which the election will be won or lost, including a 3% margin in Florida (which is still within the margin of error). What's more, Michigan and Pennsylvania, two states hard hit by job losses, have been solidly in his column for well over a month now. In addition, Kerry and Bush are still statistically tied in Tennessee for the third poll in a row. Making Tennessee a battleground state is good news for Kerry. Bush was supposed to win this one without a fight.

Finally, the suspense is over. We now know whether there was a post-convention bounce for Kerry. Answer: there was. Of course, there always is, and it evens up after the other convention, but for the moment, Kerry has received a boost. He is now ahead in the electoral college 328 to 210, with all 50 states and D.C. reporting for duty (we had two ties until today). The Wall Street Journal's estimate using a different methodology is 318 to 220. Close enough for government work.

Paul

08-03-2004, 06:24 PM

Originally posted by Anders
Unless al the polls were made by the same company as the previous in all states using excatly the same method what you found out doesn´t mean anything.

I'm no statistician and there are a few problems with my numbers, but basically what I was saying is that the NATIONAL poll is meaningless...
it is the individual states that matter:
Pre DNC Kerry 291 _ Bush 237
http://www.electoral-vote.com/jul/jul27.png (http://www.electoral-vote.com/jul/jul27.html)
Post DNC Kerry 328 _ Bush 210
http://www.electoral-vote.com/aug/aug03.png (http://www.electoral-vote.com/aug/aug03.html)

The country has became MORE polarized and Bush "lost" a bunch of red outlines to blue ones but "bounce"? SDW is right: hardly. But measuring it nationally won't help to gauge what is going on...

On a somewhat related note: do these polls accurately take into account people who have never voted before? I'm betting that a lot of people who would otherwise be disinterested in politics are being motivated to vote because of the close election last year and post 9/11 sentiment in the country. I also get the feeling that the vast majority of these people are primarily motivated by their dislike of the current administration.

Would this be taken into account with most polls?