Comments on FiveThirtyEight: Politics Done Right: Coleman Leads Franken By 136 Votes After Thursday ...

酒店經紀人, 菲梵酒店經紀, 酒店經紀, 禮服酒店上班, 酒店小姐兼職, 便服酒店經紀, ...

2009-10-02T05:30:54.625-04:00

酒店經紀人,
菲梵酒店經紀,
酒店經紀,
禮服酒店上班,
酒店小姐兼職,
便服酒店經紀,
酒店打工經紀,
制服酒店工作,
專業酒店經紀,
合法酒店經紀,
酒店暑假打工,
酒店寒假打工,
酒店經紀人,
菲梵酒店經紀,
酒店經紀,
禮服酒店上班,
酒店經紀人,
菲梵酒店經紀,
酒店經紀,
禮服酒店上班,
酒店小姐兼職,
便服酒店工作,
酒店打工經紀,
制服酒店經紀,
專業酒店經紀,
合法酒店經紀,
酒店暑假打工,
酒店寒假打工,
酒店經紀人,
菲梵酒店經紀,
酒店經紀,
禮服酒店上班,
酒店小姐兼職,
便服酒店工作,
酒店打工經紀,
制服酒店經紀,
酒店經紀,
菲
梵,

艾葳酒店經紀公司提供專業的酒店經紀, 飯局小姐,領檯人員,領台,傳播妹,或者想要到台北酒店、林森北路...

2009-08-28T02:49:59.752-04:00

艾葳酒店經紀公司提供專業的酒店經紀, 飯局小姐,領檯人員,領台,傳播妹,或者想要到台北酒店、林森北路酒店,私人招待所,或者八大行業當酒店PT,酒店公關,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禮服酒店,鋼琴酒吧當酒店領檯,酒店小姐,公關小姐??,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、學生兼差、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,想找打工、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打工,酒店上班,酒店打工環境、上班條件給水水們。

水水們妳有缺現金、有卡債、缺錢卡奴的煩腦嗎?想到日本留學日本打工嗎?妳是工讀生找工作??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,兼職工作日領假日打工的機會嗎??想實現夢想卻又缺錢沒錢嗎!??整天還在煩腦如何賺錢有什麼賺錢方法,和賺錢最快方法!?,想要打工,日領工作,短期打工,兼差工作,打工兼差工作嗎!?,
請加入我們艾葳酒店經紀公司工作單純輕鬆”高時薪”又可日領徵想要當傳播妹,上班小姐,酒店兼差,酒店兼職,歡迎學生打工,!!!
加入我們實現夢想就從現在開始^__^

外遇外遇外遇外遇外遇外遇外遇外遇外遇外遇外遇外遇外遇外遇外遇外遇外遇外遇 ...

2009-04-16T04:29:00.000-04:00

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,...

2009-02-08T21:23:00.000-05:00

The Star Tribune keeps making drastic revisions to...

2008-11-21T16:17:00.000-05:00

The Star Tribune keeps making drastic revisions to their posted results throughout the day. The most recent posting, and the assumptions listed in my previous posts, lead me to predict Coleman by 70 votes (after all of the challenges have been ruled on).

WTF happened in Renville County? Franken challeng...

2008-11-21T13:37:00.000-05:00

WTF happened in Renville County? Franken challenged 476 votes in a county where there were only around 6200 Coleman/Franken votes cast in total!

Reposting my recent comment with updated data:My l...

2008-11-21T12:59:00.000-05:00

Reposting my recent comment with updated data:

My latest simple projection: Franken to win by 16.

This assumes that:
- 75% of all challenges are against ballots counted for your opponent (requesting that they be thrown out), while 25% are against ballots that were not previously counted (requesting that they be counted for you).
- 75% of all challenges, of either kind and by either side, will be denied.
- Candidates will continue to gain votes in counties where the recount is partially completed at the same rate they have previously been gaining votes in that county
- Candidates will gain votes in counties where the recount has not started at the same rate that they have previously been gaining votes statewide

Note: Changing these assumptions, or differentiating them between the two candidates or types of challenge can be done easily. (Differentiating among counties can also be done, although less easily)

Note: Holding constant all other assumptions:
- The winner is dependent on what proportion of challenges are of one type vs the other. (ex: Coleman wins if more than 86% of challenges are of the kind challenging a ballot cast for the opponent)
- Franken wins regardless of the rate at which challenges are upheld - again assuming that this rate is similar for both candidates and both types of challenge. (however, if EVERY last challenge were dismissed, I would predict an exact tie)
- Franken wins by 62 if there is no net gain for either candidate in any counties where the recount has not yet begun.
- Coleman wins by 142 if there is no net gain for either candidate in any county from this point on

- My prediction assumes a net gain of 147 and 105 votes by Franken in St. Louis and Ramsey counties, and a net loss of 62 votes by Franken in Hennepin County - based on the results that have come out of those counties so far.

Here is something to consider:Challenges, for the ...

2008-11-21T12:45:00.000-05:00

Here is something to consider:

Challenges, for the most part, can be of two types. Define a Type I challenge as a claim that what has been judged a vote for your opponent should be a non vote. Define a Type II challenge as a claim that what has been judged a non vote should be a vote for you.

Let us accept, as seems likely, that most challenges are denied. Then when Type I challenges are resolved, they will tend to add to the total of the opposing candidate (since challgned ballots are considered non votes until they are resolved). Type II challenges will tend not to affect the total since they will stay non votes.

If the mix of the types of challenges are different between the two candidates, this would affect the expectation of how the resolving of challenges will impact the net total.

Suppose 1000 challenges each with 95 percent of challenges failing. Suppose Coleman's challenges are 52% Type I and Franken's are 48% Type I. By my calculations, just this modest difference would net Franken 36 votes, which could easily be determinative.

The question then is, do we have any reason to think that the candidates' challenges could be of a different mix? I believe that it is in fact plausible that more of Coleman's challeneges are likely to be of Type I than Franken's. This could arise if Franken supporters are even slightly more likely to mark ballots in a questionable way. It could also, arguably, arise from the philosophical tendencies of conservatives to tend to be more wary of counting a vote that shouldn't count and the liberal tendency to be more wary of failing to count one that should. (In genereal, it seems to me, conservatives fear voter fraud and liberals fear disenfrancihsement).

Just speculating.

@tmess2Thanks for that very helpful clarification....

2008-11-21T12:35:00.000-05:00

@tmess2

Thanks for that very helpful clarification. Types A and B are pretty much equivalent, since it's hard to imagine many cases when the board will confidently switch a ballot from Fr to Co or vice versa. I'll lump them as type A, and call your type C type B (freeing up the letter C).

If the total number of challenges remains about equal between F and C, then the candidate with more type B challenges is in much better shape. Any data on that distribution?

It can be inferred roughly, with poor statistics, from a precinct-level comparison of the hand audit and the recount, now available on about 100 precincts.

/mbw

My latest simple projection: Franken to win by 129...

2008-11-21T11:57:00.000-05:00

My latest simple projection: Franken to win by 129.

This assumes that:
- 75% of all challenges are against ballots counted for your opponent (requesting that they be thrown out), while 25% are against ballots that were not previously counted (requesting that they be counted for you).
- 75% of all challenges by either side will be denied.
- Candidates will continue to gain votes in counties where the recount is partially completed at the same rate they have previously been gaining votes in that county
- Candidates will gain votes in counties where the recount has not started at the same rate that they have previously been gaining votes statewide

Note: Holding constant all other assumptions:
- Franken wins regardless of what proportion of challenges are one type vs the other. (ex: Franken wins by 88 if every challenge made so far was against a ballot cast for the opponent)
- Franken wins regardless of the rate at which challenges are upheld - again assuming that this rate is similar for both candidates. (ex: Franken wins by 124 if all challenges are dismissed)
- Franken wins by 67 if there is no net gain for either candidate in any counties where the recount has not yet begun.
- Coleman wins by 142 if there is no net gain for either candidate in any county from this point on

Just a small tip for any other would-be-amateur-nu...

2008-11-21T10:10:00.000-05:00

Just a small tip for any other would-be-amateur-number-cruncher out there:
there is a nice way to import the data published by Star Tribune into your "favorite" spreadsheet.

-- Go to the site
-- View HTML source
-- find the part with the Table,
Select and copy
-- Create a new file named something.htm
-- Fire Micro$oft Excel
-- Hit "Open" and point to the file you just created.
-- Begin crunching to your heart's content.

Not at my "real" machine, so I can't tell
WV: wisset. say no more.

The problem with talking about challenges and how ...

2008-11-21T09:49:00.000-05:00

The problem with talking about challenges and how they will be resolved is that none of us really know what we are talking about (even after looking at the MPR ballots).

First, there are three categories of challenges (and multiple subcategories within each of these categories): Category A -- votes for the other candidate which the challenger believes to be a vote for neither conteder; Category B -- votes for the other candidate which the challenger believes to be a vote for his/her candidate; and Category C -- ballots in the "other" pile which the challenger believes to be a vote for his/her candidate.

We don't know how many of either candidates challenges fall within each category. If one candidate's challenges are mostly Category C and the challenges are frivolous, they do not effect the vote totals. However, if the challenges are frivoloys and belong to Category A or Category B, when those challenges are denied the other candidate's vote totals will be increased.

Second, we really can't tell how many of the challenges are frivolous even by looking at the ballots. While I think I understand the Minnesota statute, I have not read any of the cases interpreting it. There are going to be some counties where the attorneys representing one side or the other has a better understanding of the standard than the local counting team.

If you looked at the MPR site, some of the ballots had close votes. My personal belief is that the majority vote is legally wrong on two or three of the ballots. However, the interpretation that is going to matter will be the canvassing board. Some of the challenges that folks here think will fail will actually succeed.

For a not-so-serious idea of what the recount is l...

2008-11-21T09:34:00.000-05:00

For a not-so-serious idea of what the recount is like, my friend the Minnesota election judge describes it thusly (I'm paraphrasing): imagine sorting through several 55-gallon drums filled with M&Ms. You're looking to sort them by color, and looking for ones that are smashed, chipped, or otherwise no representative of M&Ms as a candy.

So it's tedious, but that's not the problem, really. The problem is that every now and then, people from the E&E company come by and insist you've actually got one of their candies instead.

Here is a serious description of the recount from ...

2008-11-21T09:22:00.000-05:00

Here is a serious description of the recount from a Minnesota perspective.

Features explanations of what is considered an acceptable or unacceptable vote.

Science and math be damned, the O'Reilly factor is...

2008-11-21T08:19:00.000-05:00

Science and math be damned, the O'Reilly factor is all that matters. Think about falafel boy, his blood boiling, completely unraveling ala Major Frank Burns (movie Burns)on national TV because Franken is senator.

@JonathanHow is counting all of the votes a frivol...

2008-11-21T08:17:00.000-05:00

@Jonathan

How is counting all of the votes a frivolous way to determine the outcome? It strikes me that this is how it ought to be.

One interesting thing jumped out at me once I sort...

2008-11-21T07:11:00.000-05:00

One interesting thing jumped out at me once I sorted through the data. Franken has consistently managed to gain recount votes in Coleman majority counties, while Coleman has gained only 8 votes from Franken counties - 3 from Big Stone and 5 from Kittson.

If you project this onto the (heavily red) counties that haven't started, you take them almost entirely out of the picture, leaving just the in progress counts and the challenges.

If you do a linear projection of the gained votes (votes gained/% of voters counted), Franken picks up over 232 votes total in those counties not including any additional votes from Dakota, which looks like an outlier. Subtract the net 70 votes he has been credited, and that's a pickup of around 160 votes not including any additional votes from Dakota. However it seems more reasonable to assume that Dakota will regress to the mean of in process Coleman majority counties with positive results for Franken, which would give him another 24 votes.

The result is a Franken win by around 40 votes ignoring challenges.

I'm looking into challenges. One thing I have noticed is that there are 2 major sorts of challenges - those which challenge a counted vote and those which attempt to add or block a new vote. One would expect these classes to have different probabilities of success. There are 823 challenges, reducing the candidates' totals by a combined 285 votes. Ignoring double challenges (yes, there are some) that implies 538 (!) challenges that either attempt to add a new vote or to block the opponents new vote. Maybe Nate has some numbers on what the success rates of different sorts of challenges are.

We shouldn't forget that there is also a challenge to the exclusion of certain absentee ballots that is taking place outside the recount proper. I have no idea how many or on what grounds.

Note that Franken is gaining nicely in the first t...

2008-11-21T06:33:00.000-05:00

Note that Franken is gaining nicely in the first two days of the recount even though, overall, the recount so far has been (cumulatively) in areas that favor Coleman...as shown by the overall counts released at 8 p.m. tonight by the MN Secy of State:

http://electionresults.sos.state.mn.us/20081104/SenateRecount.asp

Recounted Data

Totals

Percent

RECOUNT Number of Ballots for Coleman (as recounted)

534475

43.28

RECOUNT Number of Ballots for Franken (as recounted)

494804

40.07

RECOUNT Number of All Other Ballots (as recounted)

204971

16.60

This is a frivolous way to decide an important ele...

2008-11-21T06:19:00.000-05:00

This is a frivolous way to decide an important election. There's no way we'll ever know which of these two worthies actually received more votes on election day.

I don't generally believe in a runoff election, but this is the situation where it would make sense.

Despite all the endless speculation it's all going...

2008-11-21T05:59:00.000-05:00

Despite all the endless speculation it's all going to come down to a few simple problems voters have with their ballots.

The side with the most undervotes will win. And we know that there were more undervotes from Franken counties than Coleman counties.

And virtually ALL the ballots are going to be counted.

So, once all the challenges are thrown out, the winning side will simply be the one with the most undervotes. And that's Franken most likely since his voters had more ballot problems.

His precincts were more urban and their voting machines had more problems and his voters screwed up more by doing things like filling in the wrong oval, then crossing it out and filling in the other one. Machines can't deal with that, but election judges will count all those ballots.

You people don't need to "extrapolate" and make an...

2008-11-21T05:52:00.000-05:00

You people don't need to "extrapolate" and make an analysis based on shaky presumptions and a lack of evidence.

STEP 1: The ballot challenges are IRRELEVANT! Very few challenges will be upheld. Just as I suspected most of them come from 2 or 3 obvious voter problems: like the voter tapping her pen on the ballot while considering her vote and leaving a tiny stray mark in one oval before filling in the other.

Franken is challenging more because they've counted 3% more Coleman leaning precincts than Franken ones. I imagine Coleman will surge back into the challenge lead tomorrow. But, both camps will be disappointed in their challenges.

From the Minn. Star:

"Challenges rising

The number of challenged ballots continued to increase Thursday, reaching 823. Mansky -- who worked from 1984 to 1999 in the secretary of state's office and is widely considered the state's foremost elections expert -- said that people shouldn't expect many of those challenges to bear fruit.
"I can only remember two ballot challenges in all those years that were sustained," he said, meaning that the campaign lawyers' views prevailed over the opinion of election judges.

He said he wouldn't be surprised if campaign lawyers negotiate a reduction in the number of challenged ballots before the Canvassing Board meets next month to go through them.

The same issues come up regularly, Mansky said -- one filled-in oval with a faint dot in another (typically voters tapping their pens on the ballot as they read it), two ovals filled in with one crossed out, which "comes up every election."

The third most common challenge, Mansky said, involves stray marks on the ballot. State law prohibits identifying marks, such as initials, because they can signal vote fraud. "If they don't jump out as an attempt to identify the ballot, they should be counted," Mansky said."

So all those ballots will count. Nearly a wash.

STEP 2: More Coleman precincts have already been counted by 3%. Yet Franken is closing the gap. One would expect that the Franken leaning precincts would yield even more votes for him, but we'll have to wait and see. But, his camp is clearly anticipating victory. http://www.huffingtonpost.com/2008/11/20/franken-camp-were-doing-e_n_145268.html

"Marc Elias, the Franken campaign's chief counsel, said that the early recount results (which decreased the margin separating the Democratic challenger and Coleman by 43 votes) actually underestimated the ground gained.

"We do in fact feel very good about how the first day of the recount went... We believe that through last night, 26.5 percent of the ballots were hand counted. And that represents slightly three percent more of the Coleman vote or Republican vote than was true during the election. And nevertheless we picked up a significant chunk of votes," said Elias. "In other words, the ballots counted yesterday were more Republican than the total ballot pool will end up being. It was a slightly redder pile versus what it will finally be. And not withstanding that slightly redder view, we picked up votes yesterday. In fact, we believe that number is higher than the 43 votes reflected on the Secretary of State's official margin."

Elias declined to reveal just how many recount votes the campaign has internally calculated, saying it would lead to endless speculation and inspection. He added that the campaign did feel "very good about how the first day of the recount went." And while it's possible that Coleman will expand the margin as the recount proceeds, the Franken team was heartened that their position had improved even in traditionally Republican areas."

Now I'm not sure exactly what purpose would be served by counsel spinning this "we're doing very well" stuff. Both camps seem to be doing it, but the numbers are going to be what they're going to be. And nothing the campaigns say will change the vote count at all.

I can understand his optimism if they've counted 3% more Coleman ballots than Franken ones in the precincts completed, but Franken continues to gain (43 votes today).

I think by saying the number is "slightly higher" than the official vote totals he's counting on winning some challenges.

Because very few challenges will be upheld, the party that makes the MOST challenges actually will lose ground once those ballots are inevitably counted.

Perhaps he's thinking about the challenged Franken votes that will ultimately be added to his total, but most all of the Coleman ones will be too.

Here are links provided by MN sec of state, which ...

2008-11-21T04:55:00.000-05:00

Here are links provided by MN sec of state, which update the recount numbers at 8pm (local time) everyday during the recount. The link shows all the detail numbers; county by county, precinct by precinct, and how many votes have “flipped” by each county from the Nov 4 number.
http://electionresults.sos.state.mn.us/20081104/SenateRecount.asp
http://electionresults.sos.state.mn.us/20081104/

This link shows how each county voted on Nov 4, whether it went blue or red.
http://www.cnn.com/ELECTION/2008/results/individual/#mapPMN

From studied of these two links, 42.33% have recounted as of Nov 20, while the heavily populated and Obama won counties such as Ramsey, Hennepin, Dakota, Washington and St. Louis are only about 30% reporting (except Washington, 62.37% reporting). What the mean is; counties that went for McCain make up most of the 42.33% and counties that went for Obama are still in their early stage of recounting. With that in mind, Franken already cut the lead down to 136 votes form the originally 215 votes (538 dot com). Though the final recount number may be razor thin either way as everyone predicted, it thus favorite Franken.
I am absolutely not cherry picking numbers, study the links for yourself.

from reading these posts my prediction is that nob...

2008-11-21T04:26:00.000-05:00

from reading these posts my prediction is that nobody will know the answer until well into December, because the gap is likely to be so narrow that everything will come down to what happens to the challenged ballots, and that will probably end up in court.
Trying to extrapolate the final result from the bitty data available is impossible. Nobody knows enough about why challenges are being made.
Incidentally, you cannot expect either side to be absolutely pure in what they challenge. If there was a vote in the Franken pile that shouldn't be there, would you really expect a Franken supporter to point it out? Or vice versa for Coleman? Nice to think they would, but in reality in a close race, no way would it happen. It would be like demanding a recount when you're in front.

Oh, and on the subject of thread purity: some relevance to topic is good, but *just* statistics and maths with no leavening would make a terribly dry thread for most people. The site is evolving post-election - it will take time to get the balance right.
As for trolls, RWC should be on a starvation diet. The real Mule Rider can sometimes post things worthy of discussion.

with 30% of votes recounted in Ramsey county, Fran...

2008-11-21T04:07:00.000-05:00

with 30% of votes recounted in Ramsey county, Franken has gained ground by 39 votes. unlike other counties with big swings, there have been few challenges here, only 10 total (8 Franken & 2 Coleman). if this county kept up this pace Franken could make up almost half the difference there alone, even if the majority of challenged ballots ended up going to Coleman.

@redhawk04Nice of you to post a link to the Glenro...

2008-11-21T03:32:00.000-05:00

@redhawk04

Nice of you to post a link to the Glenrothes victory over the SNP which was only successful because of nothing more than fear mongering tactics; no different from what the Republicans did on this side of the pond. Too bad the Scottish people in Glenrothes bought into that B.S helped on by a labour controlled media. The difference between America and Scotland is America saw the opportunity to reboot, and went down a fresh direction with choosing Obama; Scotland still lives in the dark ages with following idiots like Brown, the man responsible for the current economic conditions due to lack of regulation, and that prat Blair with his 5 houses and 12 million dollars that he made last year having the balls to say he represented the working class man with his new labour tripe. Frankly, I'm praying for a Tory victory in the next election to get rid of those pricks because anyone that voted for them have sold Scotland down the river.