Comments on FiveThirtyEight: Politics Done Right: This Post is Already Written

艾葳酒店經紀公司提供專業的酒店經紀, 酒店上班小姐,八大行業,酒店兼職,傳播妹,或者想要到打工兼差、...

2009-09-11T02:08:06.361-04:00

艾葳酒店經紀公司提供專業的酒店經紀, 酒店上班小姐,八大行業,酒店兼職,傳播妹,或者想要到打工兼差、打工,兼差,或者八大行業,酒店兼職,想去酒店上班, 日式酒店,制服酒店,ktv酒店,禮服店,整天穿得水水漂漂的,還是想去制服店當上班小姐,水水們如果想要擁有打工工作、晚上兼差工作、兼差打工、假日兼職、兼職工作、酒店兼差、兼差、打工兼差、日領工作、晚上兼差工作、酒店工作、酒店上班、酒店打工、兼職、兼差、兼差工作、酒店上班等,想了解酒店相關工作和特種行業內容,想兼職工作日領、假日兼職、兼差打工、或晚班兼職想擁有快速賺錢又有保障的工作嗎???又可以現領請找專業又有保障的艾葳酒店經紀公司!

艾葳酒店經紀是合法的公司工作環境高雅時尚，無業績壓力，無脫秀無喝酒壓力，高層次會員制客源，工作輕鬆，可日領、現領。
一般的酒店經紀只會在水水們第一次上班和領薪水時出現而已，對水水們的上班安全一點保障都沒有！艾葳酒店經紀公司的水水們上班時全程媽咪作陪，不需擔心！只提供最優質的酒店上班,酒店上班,酒店打工環境、上班條件給水水們。心動嗎!? 趕快來填寫你的酒店上班履歷表

水水們妳有缺現領、有兼職、缺錢卡奴的煩腦嗎?想到日本留學缺錢嗎?妳是傳播妹??想要擁有高時薪又輕鬆的夜間兼職工作,打工機會和,假日打工,假日兼職賺錢的機會嗎??想實現夢想卻又缺錢沒錢嗎!??
艾葳酒店台北酒店經紀招兵買馬!!徵專業的酒店打工,想要去酒店的水水,想要短期日領,酒店日領,禮服酒店,制服店,酒店經紀,ktv酒店,便服店,酒店工作,禮服店,酒店小姐,酒店經紀人,
等相關服務幫您快速的實現您的夢想~!!

牙醫,植牙,假牙|矯正|牙周病,牙醫診所、植牙,紋身,刺青,創業,批發,TATTOO,皮膚科,痘痘,...

2009-04-13T02:11:00.000-04:00

牙醫,植牙,假牙|矯正|牙周病,牙醫診所、植牙,紋身,刺青,創業,批發,TATTOO,皮膚科,痘痘,雷射、脈衝光、除斑,中醫,腫瘤,腎臟病,僵直性脊椎炎,飛梭雷射,肉毒桿菌,玻尿酸,痘痘,脈衝光,醫美,毛孔粗大,醫學美容,seo,關鍵字行銷,關鍵字、自然排序,網路行銷,關鍵字、自然排序,關鍵字行銷、seo,關鍵字廣告,網路行銷,seo,關鍵字行銷,關鍵字廣告,關鍵字,自然排序,部落格行銷,網路行銷,網路爆紅,牛舌餅，婚紗，台中婚紗，腫瘤，腎臟病。

^^ nice blog!! thanks a lot! ^^徵信, 徵信社,徵信, 徵信社, 徵信...

2009-03-11T06:02:00.001-04:00

^^ nice blog!! ^@^徵信, 徵信, 徵信, 徵信社, 徵信社, 徵信社...

2009-03-11T06:02:00.000-04:00

情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,...

2009-02-14T01:41:00.000-05:00

ModelerThanks for the explanation and link. That's...

2008-07-12T20:40:00.000-04:00

Modeler

Thanks for the explanation and link. That's what I needed to know!

I knew a man to whom bad things constantly happened. He was steadfastly sure that nothing he was doing was causing these outcomes. He was (I take it in my newfound verbal perspicacity) precise but inaccurate - and he's now dead.

Douglas L. BarberHere's a good article on accuracy...

2008-07-12T17:46:00.000-04:00

Douglas L. Barber

Here's a good article on accuracy vs. precision. Basically, accuracy is a measure of how close you are to the correct answer, and precision is a measure of how close you are to your other answers.

So, for example, if a pollster always guessed 50% Obama, 50% McCain, they would have perfect precision but lousy accuracy. Likewise, if they bias their results towards more of a 50/50 split, they gain precision but probably lose accuracy.

"Guys, most of us DO NOT want to get into the weed...

2008-07-12T15:05:00.000-04:00

"Guys, most of us DO NOT want to get into the weeds about statistics. Leave the explaining to Nate, you guys suck at it."

Unlike Anon at 11:28, I don't speak for "most of us", only for myself. I appreciate very much the statisical discussions here, which I think is what separates this site from the flotsam and jetsam that is the rest of the political blogosphere. As someone who took statistics in college 10 years ago but has forgotten most of it, I say keep the stats discussions coming, even if they are sometimes wrong. There is always someone around to correct a mistaken explanation.

On the issue of poll weighting, while we usually think of Rasmussen as the "Gold Standard" in this regard, or at least the service that seems the most consistent in their weighting formulas, consider their 5/29 and 7/1 Connecticut polls, which went from a three point Obama lead to a 17 point lead. The outlier was probably the three point poll, but is just goes to show they even Rasmussen can come up with some rather screwy results, their commitment to historical weighting notwithstanding.

I don't understand how any given polling decision ...

2008-07-12T15:01:00.000-04:00

I don't understand how any given polling decision could increase the precision of the poll while reducing its accuracy. Thanks in advance if someone would spell this out.

Matt asked "I am not from a swing state. You are f...

2008-07-12T14:13:00.000-04:00

Matt asked

"I am not from a swing state. You are from Pennsylvania. On the ground in Pennsylvania, whats the word on Obama. How do the hard working folks in that state feel about him?

Chris Matthews mentions every day he thinks theres a "Bradley Effect" where white folks say to pollsters that they will vote for a black man and then don't."

First, for me personally, I live just outside of Philadelphia in a community that votes consistently 75-80% Democratic, and that will happen again this year. As for the five counties that make up the Pennsylvania portion of the Philadelphia metro area, they have been trending bluer for the past decade, and a tipping point has been reached in three of the four suburban counties, with more registered Democrats than Republicans for the first time ever.

In 2004, Kerry carried this region by almost 400,000 votes, which gave him enough to carry Pennsylvania by a decent margin. In 2008, the Philadelphia metropolitan area should provide at least that margin to Obama. There are some regions in the area that may have a "Bradley effect," but they will be drowned out by the rising Democratic tide.

As for the source of this tidbit, Chris Matthew, we know where his allegiances lie. There was little evidence of a Bradley effect in Pennsylvania during the primaries. Obama polled down around 6% according to the RCP average, and lost by around 9%, but that most likely was the undecided vote breaking to Clinton.

Obama will win big margins around Philadelphia and Pittsburgh, and will likely lose the "T" of Pennsylvania.

Matthews is trying to raise a perception that Pennsylvania is more competitive than the polls indicate, but that narrative has no merit. Any potential Bradley effect was minimal in the primary, and should not be a factor in the general.

It doesn't really matter if Obama has Flip Flopped...

2008-07-12T14:11:00.000-04:00

It doesn't really matter if Obama has Flip Flopped or not, he has been perceived as doing so to the credit of the McCain campaign. As much as we would like a respectful campaign on the issues, thats not what the cable news networks are going to cover, so we're left with the media spin game which the cable news networks do like to cover.(Apparently being full of shit gets better ratings. Having two surrogates from each side promote the party line is more important than covering issues)

John Kerry is a legitimate war hero and he was crucified in the media because the Swift boat attacks were excellent for ratings. Much like Jeremiah Right. If the GOP releases viscous anti Obama ads, whether its true or not, the MSM will love it and it will be played repeatedly for days. I don't like this stuff but if Obama has to go nuclear to win, do it. Screw the new politics, worry about that later.

And I must say its time for the left to shut up. Ariana Huffington is exactly what middle america hates about the democratic party. Every time I see her advocating for the Dems I cringe. She may be the nicest person in the world but she comes off as a latte drinking San Francisco liberal. Where did she get that accent. Every time she proclaims Obama doesn't need the center and needs to come home to the democratic party, Obama loses votes in Michigan, Ohio and Pennsylvania.

Nate you might want to update your post on the New...

2008-07-12T13:39:00.000-04:00

Nate you might want to update your post on the Newsweek poll. The first poll was NOT weighed by paritsan id (this put Dems with a 10+ advantage in terms of their share of the poll respondents). The poll Newsweek released today IS WEIGHED by partisan id (so even if 10+ of the respondents in the present poll were calling themselves Dems, Newsweek tuned them out and came up with its own partisan id sample size for the poll, substantially less than the unweighted poll from last month).

Seems that if a pollster is weighting a poll based...

2008-07-12T13:31:00.000-04:00

Seems that if a pollster is weighting a poll based upon the pollster's belief of the make-up of the electorate, the poll is not actually random.

Pollsters seem to be more concerned with being accurate for the moment and not really letting the science of probability work in their polling.

Truly random polling would be better at picking up actual trends or movements of support, versus the weighted polls.

Just my thought.

Bruce

Obama still up by four in Gallup. We'll see next w...

2008-07-12T13:12:00.000-04:00

Obama still up by four in Gallup. We'll see next week if Obama bounce back up in th Rasmussen tracking.

Lyesmith, Matt - The flip-flopper/moving to the ce...

2008-07-12T13:04:00.000-04:00

Lyesmith, Matt - The flip-flopper/moving to the center is a complete invention of the media. Obama has generally held the same position for the past years.

http://www.dailykos.com/story/2008/7/5/13146/00395/160/546980

It's people like you who are at fault for spreading this meme.

It will be interesting to see how this poll plays ...

2008-07-12T12:52:00.000-04:00

It will be interesting to see how this poll plays in the media. I mean, even Newsweek seems to question it's validity.

"Obama's overall decline from the last NEWSWEEK Poll, published June 20, is hard to explain. Many critics questioned whether the Democrat's advantage over McCain was actually as great as the poll suggested, even though a survey taken during a similar time frame by the Los Angeles Times and Bloomberg showed a similarly large margin. Princeton Survey Research Associates, which conducted the poll for NEWSWEEK, says some of the discrepancy between the two most recent polls may be explained by sampling error."

I am with the crowd here that sees Obama losing a point or two in the last week which is not surprising since the previous week saw the high water mark of Obama as flippity floppity with the "refine" news cycle.

Curious to see if we see a shift next week with McCain's "whinner" week.

A somewhat off-topic comment/question: How do the ...

2008-07-12T12:33:00.000-04:00

A somewhat off-topic comment/question: How do the daily tracking polls take into account weekend/weekday differences? (for example, it might be easier to reach working people during the weekend; similarly, the pollsters may sample a higher proportion of unemployed people during the weekdays).

I have noticed that there seems to be some evidence of weekly fluctuation in the Gallup data. For the last three weeks or so the race seems to tighten around the 2nd point following the weekly bars. On their longer term trend there seems to be some evidence of oscillations with period of ~1 week during and after May. It seems that Gallup has been using a three day average since June 8th, which presumably wouldn't average out fluctuations during the week. Granted, these fluctuations could be just coincidence/noise, but it may be interesting to see if this continues.

Guys, most of us DO NOT want to get into the weeds...

2008-07-12T12:28:00.000-04:00

Guys, most of us DO NOT want to get into the weeds about statistics. Leave the explaining to Nate, you guys suck at it.

We don't care that pie squared has an inverse relationship to Barack's lead in New Hampshire. Thats going too far into the weeds.

To add one more thing in an attempt to make this s...

2008-07-12T12:21:00.000-04:00

To add one more thing in an attempt to make this sort of thing intuitive, like Clark did... with the first example, you can think about it like this:

You can take a full sampling of the marbles by dividing them up into pairs. Those pairs are the possible ways you can draw two marbles. You either get B/B (blue and blue) or B/R (blue and red). No matter how you divide the marbles up into pairs, one of the pairs is going to be B/R and the other 5 are going to be B/B. Try it. It's impossible to get another outcome. There in front of your eyes is every possible result of dividing 12 marbles of that composition up into 6 groups of two. Now you see, 1 in 6 of those pairs must contain a red marble.

@Clark Miller:I'm glad that you're trying to simpl...

2008-07-12T12:13:00.000-04:00

@Clark Miller:

I'm glad that you're trying to simplify things for people, but your understanding of statistics is giving you all the wrong numbers.

In your example with 11 red marbles and 1 blue marble...

You say the odds of getting 1 red and 1 blue is 1/11. Not true. The real odds are 1/6. The only condition that needs to be met for a 50/50 split in this model is the drawing of 1 red marble out of the two that you draw from the bag (since the other is guaranteed to be blue and not drawing a red marble gives you two blues, a failure). You draw two marbles, each of which has a 1 in 12 chance of being the red one, so the odds of either of them being red is 2 in 12 (1/6).

When you ask what the odds of drawing 4 straight red balls is...

You say the odds of drawing 4 straight red balls out of a bag of 6 blue and 6 red balls is 1.7%. Well, the real answer is about 3%. The odds of the first balls being red is 1/2 (6/12), the 2nd is 5/11, the third is 2/5 (4/10), and the 4th is 1/3 (3/9). Since each of these events must happen in order to draw 4 straight red balls, the odds are multiplied to give you 3%.

I think most of your conclusions are good and I like the idea of simplifying statistics, but you cannot lose sight of how to actually solve these sorts of problems, or you'll never be able to draw reasonable conclusions when presented them (just one man's overly-righteous view about statistics in America today).

Obama better get his ass in gear. All the maneuver...

2008-07-12T12:03:00.000-04:00

Obama better get his ass in gear. All the maneuvering on the ground isn't going to matter a row of beans if Obama keeps getting hammered in the media every day. He better go hire some of Hillary's people and learn how to fight that battle.

Even if Barack is employing a "ground" strategy, you need to fight McCain every day in the media. The Flip Flop mantra has stuck now making any more shifts in position very risky. John Kerry let himself be swift boated and Obama is letting himself be painted. Imagine being called a flip flopper from McCain. Mr. Flip Flopper. But McCain waged a media battle and the press ran with it. So now Obama has to deal with a problem that should never have occurred.

He's given up a lot with this shift to the center, the left is pissed, he's lost maneuvering on future issues and he's being defined. Not to mention the small loss in poll support.

16 Hard McCainHas McCain been hanging out with Bob...

2008-07-12T11:58:00.000-04:00

16 Hard McCain

Has McCain been hanging out with Bob Dole?

A back-of-the-envelope calculation weighting the t...

2008-07-12T11:52:00.000-04:00

A back-of-the-envelope calculation weighting the two polls for party ID (based on the assumption that Dems have about a 10 point party edge) would have had the first poll with around a 10-11 point Obama edge, and this current poll with a 6-7 point Obama edge.

My gut says this is closer to reality. Obama obviously did drop, but that drop wasn't nearly that large. If his drop in Independent support is real, this is culprit for his drop overall.

Nate, are there enough polls out there that release what the party ID differential is in their polls (both nationally and statewide) to get a good representation. It'd be interesting to know exactly what percentage of independents Obama has to win in each state. It'd also be interesting to try to get a 538 evaluation of where the parties stand.

In a state where Obama only has to win 30% of the Independents, I'd be a lot more willing to call it a safe, or perhaps stable, Obama state, even if he's only up 5. Essentially if there's some way to gauge how hard/soft each candidate's support is in each state, I think it should factor in to whether we think that state will flip later on. If Obama has 70%+ hard support and a 10 point lead, it would take an epic event to knock him out. 30% hard support with a 10 point lead, that state can still be swung.

Matt, you said: "I must say I find the comments ge...

2008-07-12T11:47:00.000-04:00

Matt, you said: "I must say I find the comments getting more interesting by the day. A lot of McCain supporters just dropping in to create noise. Does McCain actually have people going around posting spin?"

The answer is, yes. McCain is encouraging his supporters to spread propaganda on the Internet. In turn they get "McCain points". So don't be surprised if you read McCain talking points everywhere. Also, be on the lookout for Republicans posing as Democrats.

http://www.johnmccain.com/ActionCenter/BlogInteract/BlogInteract.aspx

There are, just FYI, those of us who are life-long...

2008-07-12T11:42:00.000-04:00

There are, just FYI, those of us who are life-long Republicans, and who thought McCain was definitely the best of the choices on offer in our party. But still would be classified overall for November as leaning towards Obama. Make of that what you will.