Debatable

Tonight is the debate, 9pm. I expect you all to watch, or else! Sure, the debates are closed to all but the two main parties, the camera angles and audience are pre-picked, the hairdos perfectly tuned - but this is the best we've got.

So a few of pre-debate polls for you to chew on . The first is a brand spanking new L.A. Times poll in which Bush is leading Kerry by 5 points.

This article in Investor's Business Daily describes a few polls and mentions that everything is close, which it is, but if we look here it's pretty obvious that, according to the polls, Bush is ahead, however slightly.

So watch the debate tonight! I know I will, If I can find a television.

Tell me something I don't know...

...which would be impossible, since obviously I know everything. Except this, this I didn't know:
A new poll shows that Americans distrust politicians! Amazing.

Daily Show-Off

Haha, aren't I clever! I have added "off" to the end of a television show. Now, instead of Daily Show, it is Daily Show-off!!! Haha! Clearly, I have superior intellect.

And why is the Daily Show a show-off? Superior intellect need not justify itself! Ok, ok, its because this new survey shows that people who watch the Daily Show are more likely to have a more complete knowledge of current affairs than people who watch other late night TV. The Daily Show wanted the survey done because of a little spat between O'Reilly and John Steward where O'Reilly called Stewart's audience "stoned slackers".

Also of interest is this NYT article about overseas voters getting ballots late - probably good to keep track of.

Poll Troll

Well, it wasn't enough for Wired and Chris Mooney to have a go at the state of polling, now Reuters has gone and written up an article too. Geeze, maybe you guys should have a State The Obvious party.

I like parties. I'll come. But only only if you serve ice cream cake.

Poll Dancing

Ok you ravenous newfilter-eaters, it's chow time, and the state of polling is the main course.
First, two conflicting polls, one showing Bush ahead and the other showing Kerry ahead.

The first is from CNN and shows Bush ahead 52 percent to 44 percent, with a 4 point margin of error.

This poll from the Economist shows Kerry ahead with 46 percent to 45 percent, with Nader at 1 percent and the margin of error 2 points.

I don't want you to think that I am doing this thing where I post conflicting poll results because I want it to look like I'm trying to be balanced. I'm not. Actually, I want you to think I'm doing it because the Universe is a horrible, horrible place in which nothing makes sense and Half Life 2 still isn't out yet and the heat death is still at least another billion years off.

So, on that note, I've got another couple state polls that don't jive, this time from Ohio. One is from a Republican outfit and the other is from a Democratic one? Can you guess which is which? This is NOT a trick question.

And last, but of course not least, another fun electoral vote website at the Christian Science Monitor, with stuff you can click on. A nice thing about the above link is that it will show you how the numbers went in 2000 as well as more current statistics (though not as current as electoral-vote.com, of course).

And so, the elaborate waltz that is polling continues. There, I justified this article's title.

Polling and Cellphones: Update

Great article today on Wired about a new survey that shows how cell phones aren't being utilized by pollsters. Could have a small influence on this election, but it seems that by 2009 pollsters will have to seriously reconsider the way they conduct their polls.

Hey, I've got some ideas. For starters, they could use polling robots. Then again, maybe that wouldn't be a good idea, since Asimov's laws of robotics dictate that eventually all robots will turn against their masters and crush them until the last minute when they all gets souls or someone shuts them down or something. But still, at least they could find some way to use lasers, or quantum teleporation, or both.

Blog Is Not A Four Letter Word

According to a new NY Times poll (scroll down), all this blogging might actually be good for politics, surprise surprise. Me, I thought it was going to cause the apocalypse. Hey, maybe it still could. So maybe it could bring the four horsemen - but it would still be good for politics. Oh, and listen to the NPR interview - fascinating stuff, if you're a geek like me.

Also of note is another snippet about the whether poll results can change stock numbers. I mean, duh, everything changes stock numbers. Haven't these guys ever heard of the chaos theory? What? Chaos theory is so 1992 you say? Well, whatever, I watched Jurassic Park and skimmed the book, so I think I'm pretty qualified to talk about it.

And last but not least is an absolutely fantastic and completely must-read article (keep in mind that it's written by Chris Mooney, who as usual has an axe to grind) about the sort of "creative polling" I've been trying to tell you all about. It'll tell you why you can't trust the polls, and might as well not even check them at all.

Anyway, that's what this site is for! Poll Pollution - we tell you what to think so you don't have to! Isn't that nice?

The Art of Polling

Got two new articles for you guys today.

One, from the Wall Street Journal basically implies that polling is in no way an exact science. Its more of an art - not like painting, but in the same way anthropology and history aren't really exact sciences. The sort of thing I've been trying to say all along.

Also, an interesting article about voter registration in some key states from the New York Times. The gist of this one is that pollsters are missing alot of the Democratic registrants, but this might be the sort of thing that isn't proven till the election actually happens.

That said, the first debate is happening on Thursday, and everyone should watch.
Unless is pre-empts Joey, in which case I'm gonna be pissed.

Swing

All right everyone. I've hope you've got your double-moca-hyper-chino-whatever from Starbucks because today you're gonna need it. Remember electoral-vote.com, the wonderful site that shows you how recent polling within individual states has effected the electoral college numbers?

Well, compare today's numbers to yesterday's. Whoa! It's like watching Punk'd for the first time! Wow Ashton, you sure threw me for a long one!

Actually, its like watchinig anything with Ashton Kutcher in it - like that scene in Butterfly Effect where he wakes up with no arms. Brilliant cinema. Basically, its a big shock. How can these numbers change so much in polls released just a day apart, and conducted almost during the same time as eachother?

Well, children, something is rotten in the state of polling. These grand shifts are a serious matter. Short of some major news, there is no way public opinion could change so much over just such a brief period of time. So something is wrong, and I think I have an idea of what that is - I'll explain over the next few days.

All 50 States and a Poll

Great news today - the American research group has come out with polls in all fifty states. It looks pretty accurate and coincides with what we've been seeing, for the most part.

Also of interest is a Bloomberg poll of Florida, in which Nader has garnered 5 percent. Thats pretty high, at least 2 or 3 percent above most other recent Florida polls, like the one done by the American Research Group, for instance. But worth keeping track of, no doubt.

Also I have added a poll to the page to corespond with the article last week about the effect of cell phone use on polls where only landline phones are used by the pollsters. It might be telling if you web savvy readers use cell phones as your primary means of communication.
Which type of telephone do you use the most?

I told yah so!

First of all, a new poll from George Washington University shows President Bush with a slight lead over John Kerry. In the new poll, which was conducted from September 12th - 16th, George Bush is ahead with 49 percent to John Kerry's 45 percent, with a 3.1 margin of error.

Nothing to see here folks, overall, Bush has been either slightly ahead of Kerry or neck and neck. But the interesting thing is the shift that has been happening within individual states - take a looksie at the electoral vote count here. New state polls - including ones in Florida and New Jersey, have put Kerry back above Bush for the electoral vote, for now. Now Kerry is the one who is holding a tiny lead in several key states, not Bush. I expect it to shift back and forth several times before the election.

Note that neither candidate has the required 270 electoral votes they would need to carry the actual election. You might ask, why is this number 270? Or, you might not ask, but I'm going to tell you anyway. If this explanation is too much for you, I'll try to lay it out as simple as possible:
There are 538 electoral votes up for grabs, based on state population. A candidate needs 270 of them, a majority, to actually win.

The funny thing is, that while some states have their own laws governing electors, nowhere in the U.S. Constitution does it say that they have to put their electoral vote towards the winner of the popular vote in their state. Not that this happens very often, but there is nothing preventing electors (well, besides honor and righteousness) from certain states from just putting their vote towards anybody they liked.

Weird, huh? Well, don't blame me, blame that wacky U.S. Constitution.

Comparison

Once again electoral-vote.net has some cool new numbers related to just how bitterly contested this election is going to be. If we check here we initially see a country relatively secure for George Bush, but if we look at how many electoral votes are in the barely Bush catagory it shows just how much the situation could change as we approach the actual election. For instance, earlier last week Florida was actually starting to seem as if it was pretty secure for George, but more recent polls have brought his lead back to within the margin of error.

Bush has a definite lead for the time being, but any small shift in favor could have drastic ramifications on his numbers.

Election Protection

For those of you who are early risers, I've got a brief article for you about youth voting. It goes well with what we've discussed the last few days about the difficulity of polling the younger voters. At any rate, the article seems like it can only be good news - the younger generation are registering to vote in larger numbers than ever before.

Also, this weekend I worked AV for an event put on by a group called Election Protection (It was held at Columbia Law School, where I happen to have gainful employment). Really cool stuff going on there, mostly related to ensuring voter's rights.

Check them out here.

Poll Confusion?

Well guys, it seems that we've got our answer to that whole consensus question we had a few days ago: Gallup just released another poll.

This new poll puts Bush well ahead of Kerry, with 55 percent to 42 percent and a 4 point margin of error. Thats a 13 point lead - compared that to the Pew poll a few days ago, where the two of them were neck and neck!

Lets think back to the last election now - we had the same thing happen with Gore's numbers at this point, and he ended up winning the popular vote.

Why can this be? Is it that polling is an unreliable science? Of course! Polling isn't even really a science, unless you count it with those fake sciences like phrenemology, astrology, or, uhhh, psychology. Oftentimes the methedology pollsters use is kept secret, away from public scrutiny.

And when it does come out, what do we find? Well, this for example. Apparently the sample the Gallup poll took leans heavily to the right. Now, this is clearly a left wing blog - it would be nice to see if we could get similar data from the recent Pew poll and compare them, perhaps both were weighted towards one side in particular.

How can this sort of thing happen? Well, a large section of Democratic voters are young people, and this article proposes that many pollsters are missing the younger generation by ignoring people that use their cell phones as their primary means of communication (or people like me, a likely voter who doesn't even own a phone at all!).

Unfortunately both of these articles seem to have a liberal slant - maybe the polling numbers will shift back towards Kerry and we will get some conservative viewpoints as to why the polls are so biased ;).

Neck and Neck

Extra! Extra! Read all about it! Bush and Kerry are neck and neck again! It feels like June, and May, and all those other months...all over again. Except without me sleeping until, like, 5pm.

A Pew research center poll released today has shown that Bush and Kerry are now tied at 46 percent among registered voters, with Bush at 47 percent and Kerry among 46 percent for likely voters. The poll had a 3.5 point margin of error.

The reasons this is so signifigant is because the Pew Center, known for it's relative high standards and non-partisan conduct, had a poll shortly after the convention that gave Bush the standard bounce. Has Bush really lost his bounce, nationally? What do we need now to have this proved?

How about consensus? The only other poll as recent as the Pew one is one from Democracy Corps, and if you look at the little D next to the name it'll tell you why we certainly can't use that poll to build consensus (Here's a clue - there is a reason they didn't call is Republican Corps). We'll see what we get from Zogby or Gallup in the next few days.

Also, since it seems that Bush is coming ahead slightly for likely voters, this seems as good a time as any to explain the difference between registered and likely voters, if any of you don't have that figured out. Registered voters are people that are, simply, registered to vote. Like, you know, alot of college students, because they did it at the DMV. Likely voters are people who actually are going to vote.

Like my grandparents, or Jeb Bush.

States of Matter

All right guys. Time to shift focus and zoom in on a few state polls. Since I went to great lengths yesterday to explain the importance of the electoral college and individual states, today I'll focus in on some key (and no so key) state races and see what we can gleam from that.

Pennsylvania: Here, an ABC poll essentially puts the two candidates in a dead heat. Bush has 49 percent to Kerry's 48 percent, with a margin of error of 3.5 for likely voters, and 3 for registered voters. Pennsylvania holds 21 electoral votes.

Michigan: In a CNN poll Kerry is leading Bush 50 percent to 44 percent, with 1 percent for Nader and a 4 point margin of Error. Michigan's tasty filling is 17 electoral votes.

New York: According to two different New York polls, Kerry is leading Bush in New York. A Quinnipac University poll has Kerry leading Bush 47 percent to 41 percent, with Nader at 4 percent, with a margin of error of 3. A Marist college poll had Kerry at 48 percent, Bush at 40 percent, and Nader at 4 percent, with a 3.5 margin of error.

So what does all this mean? The New York polls are telling because Bush is doing surprising well in a heavily democratic state. Michigan is interesting because, if you'll check the headline, CNN is calling it too close to call - though in other polls with similar numbers have been labled as "widening leads". And Pennsylvania is worth looking at because a tip in either direction will have a signifigant effect on how the candidates spend their money in the upcoming weeks.

Economic Indicator

Also released today was the TIPP Economic indicator - which shows consumer optimism dipping slightly since August, but Presidential leadership numbers even higher than last month.
Funny, that.

Dead Heat?

Ahh, you think, I must be referring to the Joe Piscapo/Treat Williams film featuring the final performance of Vincent Price. Good stuff. And it had a zombie pig in it.

Because of course there is no reason I could be referring to the election. After all, George Bush is pulling ahead, as several new poll numbers show. Indeed, were one to look at a group of the latest poll numbers, we would see that this is true. But no, you say! Why, wasn't it yesterday that I talked about about the numbers were so darn deceptive?

Quiet, you! Yesterday, if you'll recall, I discussed the importance of consensus in identifying polling trends and such. And you'll find, that while there is no real consensus as to the extent of Bush's gains, there was consensus as the fact that, overall, he is ahead.

Ok, so end of story. You wanna know if you can sleep for the next six weeks. The election is just a popularity contest, right? Or rather, a popular vote contest? Harhar!

Surely not, you say. That's rediculous, you say. Please leave me alone and stop calling my house, you say. Because you know, of course, that it's the electoral college that really makes the difference here. We wrote about it yesterday, remember?

But there is more too it. I'm gonna show you some graphs and charts now. Ross Perot would be proud. Lets look here. This is basically similar to the Gallup summary of electoral college numbers, with one really cool difference - it alters the size of the states based on how large their electoral college contribution is. Suddenly Florida looks alot more important, huh?

The idea here is that the popular vote right now, overall, is not that important. In a race as close as this, we need to keep a close eye on actual electoral vote numbers based on who is ahead in individual states. Mark my word, we will see some more fluctuations before Nov. 2nd, whether it ends up coming ahead for Bush or Kerry in the end.

Ok, you can go back to watching the Alf talk show now.

Numbers Showdown

Calm down guys. Take your hands off your ears, don't worry. I know you've been hearing alot lately, from both camps - assault weapons ban, National Guard memos, Nader in Florida - so lets push all that to the side, for now.

We don't need that stuff. It's not going to tell us the one thing we want to know about the candidates: boxers or briefs. Am I right guys? Tell me I'm right.

Ok, I'm wrong. I'm trying to make a point here: we're getting to a sort of nether zone in the election period - after the conventions, but before the debates. Just over a month and a half to go. A point when everyone starts to really care about numbers. So lets look at a few of them, and then lets tear 'em apart.

First up, the Gallup Showdown. This one is great because it shows us electoral college numbers based on in-state polling. Right now Bush is up at 309, and Kerry has 229. If you take a look at the map, you'll see a division between the swing states - of the eight of them Gallup has listed, only Oregon and Iowa are going to Kerry. Ohio, North Carolina, Wisconsin, Pennsylvania, Missouri, and Florida are predicted for Bush.

Next lets look at a couple of other recent polls that finished the 9th of September, both of which show Bush ahead. The Newsweek poll shows Bush's lead slipping , to six points, while a Time poll still shows Bush with a 11 point lead over Kerry.

Compare this with another more recent poll by Investor's Business Daily. In this poll, Kerry and Bush are neck and neck even if you factor in Nader.

If you check the polling questions here side by side (you might have to scroll a bit), you'll find that, in this instance at least, they're all pretty similar.

So then what exactly going on here? Is Bush keeping his bounce? Or losing it? I think neither. The best thing about contrary poll numbers, especially when there are alot of them, is that they show the wavering, unsteady political climate.

For example, scientists cannot absolutely prove a great deal of their theories, so they do the next best thing: build consensus through repeated experimentation. When we have a bunch of polls like this, without consensus, it only shows one thing: That nothing is for certain.

Except, of course, another six seasons of that new TV hit Joey.

Keep cool with Coolidge!

Found a great Op/Ed piece over at the nytimes.com about the state of polling - really funny stuff. Sheds some light on how much attention is wasted on some of these polling numbers...

As usual, registration required.

Bounce Part 2

Brevity is the soul of wit, so I'll be brief: Bush got a bounce.
A small bounce, nonetheless, but he got it, and Kerry didn't, so it makes one wonder. A CNN/USA Today/Gallup poll conducted from September 3-5 found that Bush had gained a 2 point bounce - but with a Margin of Error of 3.5, this may not mean so much.

The more signifigant fact, I think, is that Kerry's numbers decreased a bit during the RNC. A Bush bounce will likely all but dissapear in the coming
weeks, as almost all bounces historically have (and Bush's is smaller than most). But a slip for Kerry could have more damaging implications for his campaign.

Some Democratic pundits are saying that Kerry should use this as an opportunity to take the same aggressive tack that many Republicans took during the convention, but I've got a better
idea. Since they've been complaining that he's been spending too much time windsurfing, he should make a real go of it.

After all the broohaha about the candidates' wartime service, I think one thing is coming out that we didn't know about America: no body really cares so much about service. We like our leader's to have served in the
army, sure, but if we've learned anything from Bush - and Clinton too - actual combat experience isn't such a big deal anymore.

But what is a big deal? EXTREME SPORTS! And if Kerry could improve his windsurfing - well, I'm not saying anything, I'm just saying. I forsee a future in which each candidate takes up increasingly dangerous extreme sports in order to cull the favor of the populace. Windsurfing now, Paintball next - perhaps the 2008 will be decided by competitive base jumping? Wait and see, I tell you. Wait and see.

Slight Increase for Bush in Polls Leads to Slight Increase of Articles about Slight Increase for Bush in Polls

Earlier this week Bush saw a slight increase in Poll numbers based on a ABC News-Post poll , leading Democratic strategists to come out with a real revelation: that there is a small but unmistakable shift in the political envirornment.

The two are still tied in the afformentioned poll at 49 percent (unless you add the leaners, in which case Bush drops a percentage point. (Margin of Error plus or minus 3 points for the ABC poll).

I'll be interested to see if this is part of a Convention related bounce for Bush, and will be sure to add some commentary on that following the RNC, but the Yahoo News AP article I linked to is really quite useless - small shifts in numbers tend to occur all the time, and often have less to do with the actual popularity of the candidate then on a myriad of other factors, possibly including our poor performance in Olympic Trampoline events.

A glance at a whole group of recent Bush vs. Kerry matchups show that the two are pretty much in the same state they were in at the start of the summer : neck and neck.

Just another piece of evidence to support my case for the dismemberment of our current electoral college in favor of an American Gladiator jousting event that would decide the presidency.