Well, I Guess This Would Explain Why Rick Ruffin Has a Column

by Robert Koehler on October 29, 2008

Repent, America! Repent!

{ 41 comments… read them below or add one }

1 The Goat October 29, 2008 at 12:31 pm: That was awesome. I needed a good laugh today.
2 kimcheeone October 29, 2008 at 12:35 pm: Maybe it should be changed to Repent, Greedy Bastards of the World.

Yeah I know everyone is blaming the U.S. for this…they have a lot of blame, but what about all those who invested in the U.S. bonds, companies, stocks and those mortgages/sub-prime loans most which were repackaged as grade A investments…

When times were good, the return was great and they didn’t question what was going on. There are some companies in the U.S. that saw the writing on the wall and didn’t squander their investors money away. Now times are bad and they blame the U.S.

SO REPENT WORLD REPENT
3 Wedge October 29, 2008 at 12:51 pm: Seeing this guy’s head explode on Nov. 5 is a good enough reason for me to vote for McCain.

Marmot: Do you mean KT runs Ruffin to make this guy look good? It’s debatable who’s the bigger asshat.
4 Robert Koehler October 29, 2008 at 12:54 pm: I mean that the writer is the KT’s assistant managing editor.
5 Jewook October 29, 2008 at 1:20 pm: “First, be a leader of the world, not just of your country.”

It almost sounds like he wants Obama to follow in the footsteps of Hitler and Emperor Hirohito and try to rule the world with an iron fist.
6 parker October 29, 2008 at 1:34 pm: The KT’s assistant managing editor seriously uses foolsdie@koreatimes.co.kr for an email address?
7 user-81 October 29, 2008 at 1:56 pm: Seeing this guy’s head explode on Nov. 5 is a good enough reason for me to vote for McCain.

Gallup has Obama and McCain in a statistical dead heat.

http://www.gallup.com/poll/111568/Gallup-Daily-Presidential-Race-Narrows-Slightly.aspx

None of this is consistent with what we would expect from random statistical variation. These considerations reinforce the point I made on Friday. In all likelihood, something else is going on here. The pollsters have different “visions” of what the electorate is, and these visions are inducing such divergent results.

This is why I would urge caution when interpreting all this polling data. We’re talking about disagreements among good pollsters. I take all of these firms seriously whenever they produce new numbers. They are disagreeing with one another in ways that can’t be chalked up to statistical “noise.” That gives me great pause.

http://www.realclearpolitics.com/horseraceblog/2008/10/more_on_the_polls_1.html
8 iheartblueballs October 29, 2008 at 3:12 pm: Gallup has Obama and McCain in a statistical dead heat.

I’m sure what you meant to say is that of Gallup’s three models, one of them (likely voters traditional) shows a statistical tie for today’s tracking.

Surely you also meant to include that Gallup’s other two models (likely voters expanded and registered voters) had Obama leads of 51-44 and 50-43 for today’s tracking, which definitely do NOT qualify as statistical ties, but DO in fact fall right in line with most of the other national trackers.

And then I know you also meant to include these other tracking polls for today…

ABC/Post – Obama +7
ARG – Obama +5
Battleground – Obama +3
Diageo/Hotline – Obama +8
IBD/TIPP – Obama +4
Ipsos – Obama +6
Pew – Obama +15
Rasmussen – Obama +5
Research 2000 – Obama +6
Zogby – Obama +4.3

…as well as these battleground state polls, which are far more valuable than national polls when it comes to electoral college math:

Colorado – Obama +9
Florida – Obama +4.8, Obama +7
Georgia – McCain +1
Indiana – McCain +2, Obama +1
North Carolina – Tie
New Hampshire – Obama +11, Obama +16
Nevada – Obama +4, Obama +10
Ohio – Obama +9, Obama +4
Pennsylvania – Obama +8, Obama +12, Obama +7
Virginia – Obama +9
Wisconsin – Obama +9

I know you meant to include all this other information, because if you didn’t, it would appear to someone that follows polling results that you had in fact cherry-picked the one poll out of about 30 that had the tiniest sliver of good news for McCain, and presented it as somehow being representative, while ignoring the other 29 that when looked at in sum, make that sliver completely irrelevant.

And if you did that, well that would just make your post seem completely ridiculous and void of any useful information or context, wouldn’t it?
9 user-81 October 29, 2008 at 4:26 pm: I’m sure what you meant to say is that of Gallup’s three models, one of them (likely voters traditional) shows a statistical tie for today’s tracking.

I’m sure I included a link that would explain all that.

And if you did that, well that would just make your post seem completely ridiculous and void of any useful information or context, wouldn’t it?

I’m also sure I might have posted that comment because it’s so much fun to watch ihbb get his panties in a bunch whenever someone mentions the possibility of McCain winning.

But he will win or come within one or two percentage points of losing. The polling process is not prepared for this unique election and their failure will come out in the actual election. “Traditional” Gallup provided a glimpse that it’s closer than most polls show.

Gallup’s “traditional” likely voter model, which Gallup has employed for past elections, factors in prior voting behavior as well as current voting intention. This has generally shown a closer contest, reflecting the fact that Republicans have typically been more likely to vote than Democrats in previous elections.

http://www.gallup.com/poll/111568/Gallup-Daily-Presidential-Race-Narrows-Slightly.aspx

Gallup’s “expanded” likely voter model determines likely voters based only on current voting intentions. This estimate would take into account higher turnout among groups of voters traditionally less likely to vote, such as young adults and minorities.

http://www.gallup.com/poll/111121/Gallup-Daily-Likely-Voters-Expanded.aspx

Traditional voters would be the people energized by Palin. Would Obama voters be any more energized than the anti-Bush voters in 2004? They came out in record numbers but the Bush supporters did, too.

It’s not going to be a runaway Obama victory. Expect a McCain or Obama win to be a photo finish. Stock up on antacid for the next few days.
10 iheartblueballs October 29, 2008 at 5:30 pm: You’re embarrassing yourself, please stop.

New voter registrations are breaking 4 to 1 in favor of Democrats. In 13 battleground states, Democrats added 1.4 million new voters, and Republicans lost 60k. Many states are reporting large increases in early voting, and expecting large increases in turnout on election day.

Of course the traditional method is going to undercount Obama support, because he’s the much bigger beneficiary of new voters. Any analysis done using 2004 models and prior voting record is bound to favor McCain given the massive surges in new voter registrations for this election. Early voting statistics have already shown large increases in minority and first-time voters, and every single state that provides early voting statistics has shown an advantage for Democrats, when in 2004 early voting favored Republicans nearly across the board.

I could care less if you post your opinion that McCain will win or “lose a close election” (nice backpedalling by the way). You’re setting yourself up to be mocked extensively, but whatever, I’ve been laughing at Brendon for 6 months for his numbnutted prediction of 135 electoral votes for Obama.

But at least provide something to back it up if you expect to be taken seriously. Thus far you’ve provided nothing but cherry-picked needles in haystacks while ignoring a landslide of data in the other direction.

Simple question: What electoral map do you see McCain carrying?

I look forward to seeing how many states that McCain is currently 7-10 points down in you must include, because there are several.
11 user-81 October 29, 2008 at 5:55 pm: You’re embarrassing yourself, please stop.

No I’m not. I’m having fun watching the irreverent and imperturbable Iheartblueballs get his panties in a bunch.

On November 4 I will be proved right or wrong about McCain winning or losing by only one or two percentage points. There’s no reason for you to write half a page each time to contradict me. Your words won’t affect me being right or wrong.

But seriously, it’s not going to be a runaway Obama victory. Expect a McCain or Obama win to be a photo finish. Stock up on antacid for the next few days.
12 user-81 October 29, 2008 at 6:03 pm: IHBB, I think you’re right that Brendon Carr is a numbnuts for predicting 135 electoral votes, unless he meant 135 votes plus at least 135 more.

IOW even if McCain squeaks by in the popular vote he will probably lose the electoral vote. I think Brendon Carr said that in the hope that his prediction would demoralize the Obamaniacs so much that they would blow a circuit and roll up into a ball of despair and be unable to even walk out the door on election day.

But seriously, it’s not going to be a runaway Obama victory. Expect a McCain or Obama win to be a photo finish. Stock up on antacid for the next few days.
13 The Goat October 29, 2008 at 6:18 pm: I think IHBB should write for one of the dailies.
14 Sonagi October 29, 2008 at 6:31 pm: I’m having fun watching the irreverent and imperturbable Iheartblueballs get his panties in a bunch.

Panties in a bunch? You just got smacked down. Hard. Glad you enjoyed it.
15 user-81 October 29, 2008 at 6:50 pm: Panties in a bunch? You just got smacked down. Hard. Glad you enjoyed it.

Smacked down hard? I deliberately picked the one poll where McCain and Obama are in a dead heat and he practically blew a blood vessel pointing out that it’s just one cherry picked poll. He could have done that in one sentence.

But seriously, McCain is going to do a lot better than you think. Forget the 10 point spread and I think there’s a better than 50-50 chance he will actually think.

But what would be wrong with that? I know everyone here hates Bush but McCain is not Bush. Why get so upset about the prospect that McCain might win? It’s really scary how cult-like Obama supporters can be. Even IHBB who seems to despise everything except blue balls looks like he has a tingle up his leg for Obama.
16 user-81 October 29, 2008 at 7:01 pm: New voter registrations are breaking 4 to 1 in favor of Democrats.

Now that’s cherry picked data. In 5 EV Nevada, yes. In 27 EV Florida, no. In Virginia, no.

http://www.washingtonpost.com/wp-dyn/content/article/2008/10/05/AR2008100502524.html?hpid=topnews

And that doesn’t mean as much as you might think. In Florida in 2004 Bush won the state by more than 381,000 votes “even though Republicans trailed in the voter registration count that year by nearly 368,000″.

http://www.latimes.com/news/politics/la-na-registration6-2008oct06,0,2248014.story

nice backpedalling by the way

I said a long time ago I think the polls are wrong because there’s an unusual set of circumstances and it will be a very close race. Only recently did I boldly say McCain would win but if I say that McCain might lose by only one or two points that’s similar to what I’ve been saying all along.

I could care less if you post your opinion that McCain will win or “lose a close election”

Actually you seem to care a lot.
17 R. Elgin October 29, 2008 at 7:04 pm: . . . everyone here hates Bush but McCain is not Bush

Yes, but Mr. McCain is sponsored by the same party that promoted that loser and that is more than enough reason to reject whatever they offer. America would be better represented by a party that has not lost its raison d’être and prizes ideology over competence in government.

Sic semper tyrannis . . .
18 user-81 October 29, 2008 at 7:20 pm: Let’s look at the MoE (margin of error a little bit).

http://www.realclearpolitics.com/epolls/2008/president/us/general_election_mccain_vs_obama-225.html

Reuters/CSPAN/Zogby has Obama at 49 and McCain and 44. MoE is 2.9, so Obama could be down to 46.1 and MCain up to 46.9. A statistical dead heat.

Rasmussen has Obama at 51 and McCain and 46. MoE is 2.0, so Obama could be down to 49 and MCain up to 48. Advantage Obama if it holds.

Gallup Traditional has Obama at 49 and McCain and 47. MoE is 2.0, so Obama could be down to 47 and MCain up to 49. A statistical dead heat.

Gallup Expanded has Obama at 51 and McCain and 44. MoE is 2.0, so Obama could be down to 49 and MCain up to 46. Advantage clearly Obama if it holds.

Diageo/Hotline has Obama at 50 and McCain and 42. MoE is 3.6, so Obama could be down to 46.4 and MCain up to 45.6. Advantage Obama if it holds.

ABC News/Wash Post has Obama at 52 and McCain and 45. MoE is 2.5, so Obama could be down to 49.5 and MCain up to 47.5. Advantage Obama if it holds.

Ipsos/McClatchy has Obama at 50 and McCain and 45. MoE is 3.4, so Obama could be down to 46.6 and MCain up to 48.4. A statistical dead heat.

IBD/TIPP has Obama at 48 and McCain and 44. MoE is 3.0, so Obama could be down to 45 and MCain up to 47. A statistical dead heat.

GWU/Battleground has Obama at 49 and McCain and 46. MoE is 3.1, so Obama could be down to 45.9 and MCain up to 49.1. A statistical dead heat.

Pew Research (which was highlighted in a link I gave, IHBB claim of me cherry picking notwithstanding) has Obama at 53 and McCain and 38. MoE is 3.5, so Obama could be down to 49.5 and MCain up to 41.5. Advantage clearly Obama if it holds.

So FIVE of TEN have the two in a statistical dead heat, unless I’m doing the numbers wrong by subtracting/adding the MoE from both candidates. (Yes, I know that it goes up and down for EACH of the candidates but the “down” range for McCain and the “up” range for Obama is irrelevant to the overlap).

I was just messing with people’s heads about Gallup, but until I crunched these numbers I had no idea that HALF of the major polls had a DEAD HEAT. Wow. A Bradley effect (if it exists) could make a difference in the popular vote. A strong turnout by Republicans excited about Palin could make a difference.

Registrations of Mickey Mouse could make a difference when Mickey doesn’t show up to vote. Or the once excited college student doesn’t show up to vote.

Really, would a McCain presidency be so bad?
19 user-81 October 29, 2008 at 7:36 pm: New voter registrations are breaking 4 to 1 in favor of Democrats.

“You know what they call a candidate who’s counting on a lot of new voters? A loser.” (James Carville)

In 2004 new voter registration favored PresidentSenator Kerry.

Nationwide, at least two polls in the last week showed that newly registered voters favored Democratic presidential candidate John F. Kerry by double-digit margins. The Massachusetts senator holds an even greater lead, the polls found, among voters 29 and younger, many of whom will be voting for the first time.

And Kerry went on to lose by 3 million votes. 10 million more people voted in 2004 than in 2000. The left was well positioned against Bush and they came out in record numbers. But the Bush supporters came out in even larger numbers than that.

Palin will bring those people back. Until I noticed those FIVE of ten polls in a dead heat I thought that a 1 or 2 percent margin was a bold prediction. I need something bolder: a McCain popular win but electoral loss (but I still want some credit for a McCain loss by just one or two points).

Really, would a McCain presidency be so bad? I’m sure Palin will learn a lot in her first year.
20 Linkd October 29, 2008 at 10:32 pm: Does your vow still stand, user-81?

http://www.rjkoehler.com/2008/10/25/open-thread-72/#comment-198159

If I’ve followed along right, I think it was made before you went all wishy-washy on the outcome.
21 user-81 October 30, 2008 at 2:35 am: It stands.
22 user-81 October 30, 2008 at 3:27 am: It stands only if IHBB’s vow also stands.

http://www.rjkoehler.com/2008/10/25/open-thread-72/#comment-198165

And this is only about the popular vote. I never claimed McCain would win the electoral vote.
23 setnaffa October 30, 2008 at 6:12 am: user-81, I’ll claim it’s going to be a McCain blowout…

Though I am not a prophet from God–so if I’m wrong I hope you won’t take me out and stone me!!

I think it’s going to be 300+ EC votes for McCain…

The recent Joe the Plumber incidents, video releases of Obama bemoaning Constitutional limits against his Marxist fantasies, and McCain suddenly becoming forceful have affected the polling. Well, that and the pollsters traditional return to semi-honest methodologies as the election draws nigh…

I want to see the USA survive. I want to see the victims of BDS get the inpatient care they need. And I want to be the one who decides which charitable donations I make…

If liberals want to spend more money on the poor or whatever, let them do it with their own money. Joe Biden gave less money to charity last year than most folks who work in donut shops…

Forcing other people to donate to your favorite cause is not the American way.

And before the anyone starts up with “…well, Bush did this…” or “…well, McCain dis that…”, let me ask everyone to focus like a laser beam and answer three questions:

1. What percent of your own money did you give to a charity last year.

2. Do you think some nameless, faceless bureaucrat from either political party could make a better choice of where to spend it than you did?

3. How much should you pay that bureaucrat to make those choices for you, knowing they might pick a group you despise?

PS, I know the GOP is as full of corruption as the Democrat Party. I just happen to like their take on Marxism a little bit better…
24 user-81 October 30, 2008 at 9:18 am: user-81, I’ll claim it’s going to be a McCain blowout…

Big deal unless you back it up with a promise to leave the Hole like me and IHBB did. Are you willing to do that JTB-in-Texas?

I shocked even myself with the five of ten top polls at RealClearPolitics having the race at a statistical dead heat. Since IHBB and Sonagi haven’t responded then maybe I shocked them too. Today I checked again and it’s even worse for Obama.

Rasmussen, Gallup traditional, Diageo/Hotline, Reuters/CSPAN/Zogby, IBD/TIPP, GWU/Battleground, and Ipsos/McClatchy have a statistical dead heat. That’s two more than yesterday since it includes Rasmussen (50-47 with a 2.0 MoE) and Diageo/Hotline (49-42 with a 3.6 MoE).

Seven of the TEN COMMANDING POLLS have a statistical dead heat. Where’s your messiah now, Iganos?
25 iheartblueballs October 30, 2008 at 9:38 am: By all means user-81, keep on shocking yourself by DOUBLING the MoE. Hell, go ahead and quadruple it and then you can claim McCain is LEADING in all those polls.

Once again, embarrassing. If you don’t understand polling and MoE, stop pretending as if you do.

Do you really think betting books around the world would be making McCain a 10-1 underdog if seven of the TEN COMMANDING POLLS had the race as a dead heat?

Try mixing a little common sense into your pie-in-the-sky analysis. That is if you’re not too busy making up fake polling data using your fake margin of error. Jesus, North Korean vote counters are more honest.
26 user-81 October 30, 2008 at 10:12 am: By all means user-81, keep on shocking yourself by DOUBLING the MoE.

Hey, calm down. Relax. Cheer up. I did specifically ask if “I’m doing the numbers wrong by subtracting/adding the MoE from both candidates“. I guess my mistake was misreading the MoE as plus-or-minus and not the whole range. So a 3.6 percent MoE is ±1.8 and not ±3.6. That makes sense now. Thanks.

Once again, embarrassing. If you don’t understand polling and MoE, stop pretending as if you do.

Who is pretending? I even asked if I was calculating that correctly. Wow, you really get upset when it comes to your messiah. Someone might imagine you crushing those blue balls out of rage.

But you should know that even if he wins the world’s problems will not be solved overnight. There might even be difficult challenges in the months ahead as the world tests his mettle. Just a heads up.

(I don’t think McCain can pull of an EC victory so I’ve made peace with an Obama victory even if I didn’t vote for him).
27 inkevitch October 30, 2008 at 10:30 am: IHBB, bookies hedge against bets previously made. Idiots often group and make bets that their country will win the world cup and within their country their odds will be short, if they made trhe same bet in another country they would get better odds.

So if everyone is seeing the polls and thinking it is a sure thing they will take the Obama odds no matter how poor, leading to massively overinflated odds for McCain. Just saying bookie odds mean squat. The polls and voter registration, voter turnout, campaign funding and volunteers leads me to otherwise agree with you.
28 user-81 October 30, 2008 at 10:48 am: After looking at Rasmussen’s site linked from RCP I am again confused about the MoE calculation (that’s right IHBB I am not claiming to be an expert). According to them the 2.0 MoE is a ±2.0, not ±1.0 like I conceded.

Daily tracking results are collected via telephone surveys of 1,000 likely voters per night and reported on a three-day rolling average basis. The margin of sampling error—for the full sample of 3,000 Likely Voters–is +/- 2 percentage points with a 95% level of confidence. Results are also compiled on a full-week basis and crosstabs for the full-week results are available for Premium Members.

Now I’m back to my original question: Is the MoE subtracted and added to each candidate’s numbers or not?
29 iheartblueballs October 30, 2008 at 11:18 am: My pointing out your ridiculous analysis based on manufactured data has nothing to do with the relative merits of either McCain or Obama. Anyone who knows polling and can reasonably analyze other metrics involved in forecasting elections could just as easily call you out for the bullshit you’re peddling. I’d be doing the same if you claimed Obama were going to win by 15, and you used cherry-picked data to try and legitimize it.

But you should know that even if he wins the world’s problems will not be solved overnight.

Do you have any more overstatements of the obvious, or is that all?

There might even be difficult challenges in the months ahead as the world tests his mettle.

Really? Will the sun rise tomorrow as well?
30 user-81 October 30, 2008 at 11:30 am: I’m just having fun with you about the Obama messiah comments, IHBB. It really is fun watching you get so worked up about something besides people who believe in religion.

Since you said I was doubling the margin of error can you give me a real-world example of where the margin lies? Using the IBD/TIPP poll with Obama at 48 and McCain at 44, with a margin of error of 3.0, I thought that meant the sample indicated actual support for Obama was likely somewhere between 45 to 51 and actual support for McCain was somewhere between 41 and 47. The overlap meant a dead heat that couldn’t be confidently determined from the polling sample.

As calmly as you can (remember: Obama is 85% likely to win no matter what I say) can you please tell me where I am wrong and what the correct interpretation would be?
31 Linkd October 30, 2008 at 11:38 am: If I remember my stats correctly, it’s like this: there is a TRUE average, which is a number that cannot be known, unless EVERY voter is polled. The polls come up with AN average, which has some mathematical chance of being close to the true average, based on the all-important assumptions that the sampled population is actually representative of the total population.

When something is reported as 21% plus/minus 2%, 19 times out of 20 in a sample set of 1000, it means that if you went out and asked 20 different sets of 1000 random people, then 19 of those sets would return an average score of between 19 and 23%, with those 19 scores distributed fairly evenly around 21%. ONE of those sets would return an average outside of this range.

Nothing in the above paragraph makes any reference to the TRUE average, in this case, what percent of the total population would vote for Obama. (Note, how pollsters choose their sample sets to account for likely voters as opposed to all eligible voters I don’t know – but all such fiddlings take a sampled statistic further away from a true statistic).
32 user-81 October 30, 2008 at 11:40 am: ABC News/Washington Post says that early voting is headed in Obama’s favor:Vote preference among these early voters is 59-40 percent, Obama-McCain, widening to
68-31 percent in the 16 battleground states and 71-28 percent in the eight toss-ups states
as designated by ABC’s Political Unit (Colorado, Florida, Indiana, Missouri, Nevada,
North Carolina, Ohio and Virginia).I think Democratic supporters who were kept from voting in Ohio in 2004 and Florida in 2000 have taught everyone a lesson.
33 Linkd October 30, 2008 at 11:43 am: Of course, now that I’ve written that, I’m getting snippets of contradictory recollections from that 2003 stats class…but damned if I’m going to go back and open up that textbook again. Think I’ll just wait for the election results to bring in Obama. What’s considered a landslide in America? 52%
34 user-81 October 30, 2008 at 11:45 am: I’m an arachnologist not a statistician so I don’t know the lingo.

If you take out “actual support” and put in “true average” does it make sense?
35 iheartblueballs October 30, 2008 at 11:51 am: Margin of error of 3% means the total movement of BOTH candidates cannot exceed 3.0 in your example.

So the range for Obama would be 46.5-49.5 and the range for McCain would be 42.5-45.5.

So even if you take Obama’s bottom range of 46.5 and McCain’s top range of 45.5, you still have an Obama edge. Thus, it does not fall in the MoE.

So if everyone is seeing the polls and thinking it is a sure thing they will take the Obama odds no matter how poor, leading to massively overinflated odds for McCain. Just saying bookie odds mean squat.

That can be true, but rarely for long. Overinflated odds at one book always bring action to the opposite side, generally evening them out. With numerous books offering lines, over time the odds will settle, particularly in this case with a period of months for wagers to be placed.

You can even ignore the books and just look at the markets, which are not like the books in that there’s no incentive to place equal action on both sides and collect the vig. The largest market at Intrade currently has McCain at 15%, and that includes a huge effort by a single investor to artificially inflate McCain’s market price. He did so by buying contracts above market value to inflate McCain’s value even when they were available for cheaper elsewhere.

The CEO of intrade detailed it on the site:

http://www.intrade.com/jsp/intrade/misc/blog/#jd_3
36 user-81 October 30, 2008 at 12:18 pm: Margin of error of 3% means the total movement of BOTH candidates cannot exceed 3.0 in your example.

Thanks. If that’s true then only GWU/Battleground and and maybe IBD/TIPP would have them overlapping and that’s just barely.

But I’ve been trying to find something to support my earlier interpretation or your interpretation and the students at Tufts seem to agree with my interpretation (I’m choosing this one because it’s the clearest of the explanations I’ve seen):

The biggest mistake most people make when they try to interpret polling data is that they disregard the margin of error. The MoE is what makes every poll, to one extent or another, “fuzzy.” It’s what prevents us from ever knowing the precise, actual percentages of the public’s preferences. Most reputable polls carry an MoE of plus or minus three points at a 95 percent confidence level. So let’s take the latest FOX News poll, which puts Obama at 49 percent and McCain at 40 percent nationally. With an MoE of three percent at a 95 percent confidence level, we can say that we know for a fact that there is a 95 percent chance that Barack Obama’s support is as high as 52 percent or as low as 46 percent, and that McCain’s support is as high as 43 percent or as low as 37 percent. What FOX pegged as a 9 point difference could be as small as 3 points: 46-43 percent.

http://www.tuftsdaily.com/1.842962
37 user-81 October 31, 2008 at 3:52 am: Linkd or anyone… Can you support or refute with a link whether my earlier interpretation of the margin of error (that it is a ± for each polled value, so that you apply it to both the Republican candidate’s polled value AND the Democratic candidate’s polled value) or IHBB’s interpretation that by applying the MoE to both candidates I was “doubling” it? (I think he was saying the margin of error is a number applied to the range of both candidates.)

I was willing to take IHBB’s word for it because he was so adamant and this is not my field but when I went looking for verification it seemed to support my earlier interpretation. Now I’m not sure.

The polls today have inched toward Obama (by like 1/10 of a percent) but I think that’s within the MoE regardless of how it’s interpreted. Only five of ten today would be “a statistical dead heat” if my interpretation was right. And one of them is Fox.
38 iheartblueballs October 31, 2008 at 7:27 am: http://www.cjr.org/behind_the_news/margin_of_ignorance.php

Your interpretation is wrong because you don’t understand what actually constitutes a “statistical dead heat.”

A +/-3% in a 48-44 Obama lead includes the possible ranges of 45-51 for Obama and 41-47 for McCain. But it is not a statistical dead heat nor within the margin of error, which you claimed. A statistical dead heat occurs when the difference between the %s (4 in this case) is equal to or smaller than the MoE.

The further you get away from the baseline % in either direction, the less likely that outcome is. And what you’re doing is taking the lowest possible range for Obama and highest possible range for McCain, and using those as the baseline, despite the fact that the odds of that occurring in one poll is statistically very small, but the likelihood of that exact deviation occurring in every poll is infinitesimal. That is no more legitimate than doing the opposite (which would make it 51-41 Obama) and claiming that it’s an Obama 10-point blowout. Taking the McCain high-end and the Obama low-end on every poll is ridiculous statistical manipulation to support your conclusion.
39 R. Elgin October 31, 2008 at 8:14 am: I think Democratic supporters who were kept from voting in Ohio in 2004 and Florida in 2000 have taught everyone a lesson.

Yes, the system has been gamed too much and the Republican Party should be disbanded or change its name to “Kaos”.
40 user-81 October 31, 2008 at 8:23 am: IHBB:
Your interpretation is wrong because you don’t understand what actually constitutes a “statistical dead heat.”

You said in that my interpretation was wrong because I “doubled the margin of error”.

By all means user-81, keep on shocking yourself by DOUBLING the MoE. Hell, go ahead and quadruple it and then you can claim McCain is LEADING in all those polls.

Did I double it or not? Your link seems to support the way I calculated it: adding and subtracting the MoE to get a range for each candidate. If those candidates overlap then it is “tied” and “you can’t tell who is in the lead”.

As an example, let’s take the Newsweek numbers CNN political analyst Carlos Watson attempted to explain on July 12. In that poll, 51 percent supported Kerry-Edwards, while 45 percent supported Bush-Cheney. The poll had a margin of error of plus or minus 4 percent, with a confidence interval of 95 percent. The simple six-point gap between the two tickets led Watson to believe that the difference surpassed the margin of error.

However, the margin of error was plus or minus four points for each number. If we subtract four points from the Kerry-Edwards ticket and add four points to the Bush-Cheney ticket, the result is Bush-Cheney at 49 percent, two points ahead of Kerry-Edwards at 47 percent.

“Margin of error, Smargin of Smerror,” you say to yourself, the Kerry-Edwards six point advantage is still a lead, regardless of these calculations. Well, not really.

Dick Bennett, president of the polling firm American Research Group, told me, “If [the numbers] are within the margin of error, you can’t tell who is in the lead.” He added, “In theory, it’s tied.”

So did I double the MoE or not? If I did can you explain how it is a doubling?

A +/-3% in a 48-44 Obama lead includes the possible ranges of 45-51 for Obama and 41-47 for McCain. But it is not a statistical dead heat nor within the margin of error, which you claimed.

What is not within the margin of error? In #18 I talked about ranges based on margin of error and pointed out half or where Obama’s and McCain’s overlapped, thinking that that was called a “statistical dead heat” (but your link says they are theoretically tied).

A statistical dead heat occurs when the difference between the %s (4 in this case) is equal to or smaller than the MoE.

I thought if the ranges overlapped then it was called a “statistical dead heat” but I would be the first to admit using a wrong term if it’s wrong. What is it called then if the ranges overlap? Is it a “statistical tie”? Is it an undetermined outcome? What it is called?

And what you’re doing is taking the lowest possible range for Obama and highest possible range for McCain, and using those as the baseline,

I did not do that. I talked about overlapping ranges is all. I never used McCain’s high range as a baseline and Obama’s low range as a baseline.

I started with an outlier poll that had Obama and McCain at 49-47 (which would be a dead heat according to your definition) and then I also noted that half the polls had Obama and McCain with overlapping ranges IF my ±MoE calculation was right.

despite the fact that the odds of that occurring in one poll is statistically very small, but the likelihood of that exact deviation occurring in every poll is infinitesimal.

My major point was to say that despite the polls people are citing I don’t feel this is going to be a runaway Obama victory and this might even be reflected in the polls themselves. But I don’t even think McCain can win the electoral vote. And I think that there are some social (and structural) factors this time that are “unusual” enough that the polls might be wildly off because they don’t properly reflect it. I’ll save that for another comment.

I am NOT saying “look at these polls, they clearly point to a McCain victory if you twist your head right!” Don’t attribute that to me.

Taking the McCain high-end and the Obama low-end on every poll is ridiculous statistical manipulation to support your conclusion.

What was my conclusion? If I even had a conclusion it was just that it’s not a runaway victory for Obama.
41 iheartblueballs October 31, 2008 at 9:26 am: You’re making this far more complicated than it actually is.

(Candidate A poll %) minus (Candidate B poll %) = X.

Margin of error = Y (this is a single percentage, not a range)

If X is greater than Y, then Candidate A has a statistically significant lead over Candidate B.

X is equal to or less than Y, then the race is considered to be a statistical tie.

What is it called then if the ranges overlap?

Whatever you want to call it. Pollsters and statisticians find nothing significant about overlapping ranges, and therefore don’t call it anything. You can call it Marilyn if you like. With the avg MoE in the 3-5% range, overlapping ranges are extremely common and signify nothing.

..then I also noted that half the polls had Obama and McCain with overlapping ranges IF my ±MoE calculation was right.

Of the 10-15 concurrent national weekly/tracking polls coming out daily over the last several weeks, there are usually one or two within the MoE. Just as there are usually one or two showing a 13-15 pt lead for Obama. At no time has it ever approached half being within the MoE. If you throw out the low outliers within the MoE, and throw out the high outliers with 13-15 pt leads, you’re left with 8-10 polls all showing a 5-7 pt average lead for Obama.