Comments on FiveThirtyEight: Politics Done Right: Return on Investment

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, ...

2009-10-08T08:08:16.789-04:00

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, 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、...

2009-09-11T01:50:06.602-04:00

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~!!

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

2009-03-11T05:50:00.001-04:00

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

2009-03-11T05:50:00.000-04:00

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,...

2009-02-12T22:56:00.000-05:00

徵信社, 感情挽回, 挽回感情, 徵信, 徵信社, 徵信, 捉姦, 徵信公司, ...

2009-02-08T23:07:00.000-05:00

Time to reexamine the Tipping States and ROI.1. I...

2008-09-08T11:40:00.000-04:00

Time to reexamine the Tipping States and ROI.

1. Its obviously asymmetric, so there needs to be two charts one for O and one for M.

2. So many states are colored in on the Tipper and ROI that's its becoming increasingly impossible to separated noise from information.

3. ROI, really does take into consideration the paths revealed by State Similarity Scores. E.g., the ROI for investing in IN is not just to win IN, but also affects MI and OH because if O is really close in IN, even if he doesn't win that predicts that he will win OH and MI.

Now that the EV distribution chart has started sho...

2008-07-24T15:53:00.000-04:00

Now that the EV distribution chart has started showing a bi-modal 7/24 (pretty funky), I think it is even more critical that the Tippers for
Obama be separated from the Tippers for McCain.

@chariots July 6, 2008 3:33 PM You indicated that...

2008-07-12T15:38:00.000-04:00

@chariots July 6, 2008 3:33 PM

You indicated that "On the issue of tipping states there is no difference between an O-tipper and an M-tipper."

I think you were wrong then and you are certainly incorrect with the tippers V2.0.

As to ROI, I'm not sure that this metric is saying anything important or actually providing information. Part of the problem is the breaks and coloring (there is no wheat from chafe separation). Essentially the map is just Obama likely or "within reach" states minus safe Democratic and safe Republican states.

Part of the problem may be that the tipper metric which needs to be first calculated is not separating O Tippers from M Tippers.

@Randall:This is similar to the approach used at h...

2008-07-09T14:41:00.000-04:00

@Randall:

This is similar to the approach used at http://election-projection.net. Election-Projection.net computes the partial derivative of the probability of each candidate winning verses a one percentage point shift in the state's opinion polls.

Original message:
"Another way of phrasing the relative value metric, which may produce different answers from your 'tipping point' approach, is 'what is the relative probability that a single vote (or a hundred votes) flipped in this state would flip the election's outcome?' That seems like the most unambiguous definition of the value of a vote in a state."

@jdkOn the issue of tipping states there is no dif...

2008-07-06T16:33:00.000-04:00

@jdk

On the issue of tipping states there is no difference between an O-tipper and an M-tipper. In each simulation rank the states from most Obama to most McCain and identify the state that surpasses 270 EV. Repeat that ordering from McCain to Obama. You get the same "tipping" state.

Furthermore the distribution of rankings will be independent (or close to it) of the popular vote. Thus the probability of a state tipping an election will be approximately independent of whether it was won for McCain or Obama.

I agree with you on ROI. The price of a vote may vary by candidate even in tossup states. (Keep in mind Nate's ROI is not based on infra-marginal returns).

Secondly a message may work in multiple states i.e a bundling effect thus reducing the relevance of the notion of a singular tipping state.

The criticism (far above in this thread) of the ne...

2008-07-06T05:37:00.000-04:00

The criticism (far above in this thread) of the news cycle for honing in on Obama's comment -- that he might refine his Iraq policy depending on what the military commanders tell him -- is not phantom news. The journalists are focusing on a real political point at issue:
Obama had been saying for many months now that his Iraq policy was withdrawal of most U.S. troops in 16 months from inauguration. Now, it appears, he is saying that he would refine that (i.e. most troops might not be withdrawn in 16 months) if the military advice was that such a withdrawal could not be accomplished safely, and leave a stable Iraq behind [WTF? we can't create that]. This shift in emphasis is politically important, because it brings Obama perilously close to a Bush position: we'll leave, except for some Al Qaeda-fighting special teams, when safety and stability permit.
Obama now has a real problem in this area. The situation on the ground is better than it was, but the improvement is fragile. That means, looking ahead, that Obama could inherit a somewhat better situation in Iraq than we would have expected six months ago. But his policy of beginning immediate troop withdrawal could well show equally immediate decrease in stability on the ground in Iraq -- a black eye for him early in his term of office. Possibly gets tagged as 'the guy who lost Iraq.'
The correct answer to this may not be politically feasible or persuasive: Yes, things are now better in Iraq (level of violence down, so forth). So, Repubs, why do you need to keep so many troops in Iraq if things are going so well? [The answer is that only with persistent U.S. forces efforts do we keep the lid on the simmering civil strife.] But the voters are likely sympathetic to the view 'if things are going better, probably we should keep doing what we are doing.' And I think Obama is sensitive to this dynamic, and is trying to do something about it. The cost, however, is the accurate perception that he has shifted his policy position.

I think the derivatived(overall percentage of winn...

2008-07-05T16:19:00.000-04:00

I think the derivative
d(overall percentage of winning)/d(expected number of votes in state Y)
would be a much better ROI metric - but I suppose that can't easily be measured without extra simluations.

This ROI metric is pretty volatile or maybe not a ...

2008-07-05T15:03:00.000-04:00

This ROI metric is pretty volatile or maybe not a measure of anything In the intitial 7/3 post Indiana is 3.3 (ninth), but today (7/5) the ROI puts IN somewhere below 2.0 and IN does even make the top 15.

How? change in method?

Also, I'm a broken record but O tippers are not M tippers (there should be two maps), so an ROI would need to be related to whose doing the investing. E.g., if NM is more of an O tipper than a McCain tipper, then maybe O really gets more bang for his buck than McCain in NM. On the hand NH may give McCain more bang for his buck than Obama, because Kerry - NH + IA makes NH less relevant in terms of ROI for O.

As I've just noticed this, I post here. NJ now just makes the tipper list. Consider that it is more likely than NH to go Democratic, but then look at the most recent polls. O breaks 50% in NH 2 out of 3 times, but can't break 50% in NJ is recent polls. Why.

If NJ is really a M possible tipper not an O possible tipper (which two maps would reveal) then we'd (especially the campaigns) would have something to consider.

This is really brilliant analysis. It makes so muc...

2008-07-04T16:49:00.000-04:00

This is really brilliant analysis.

It makes so much sense - I wonder if the campaigns have folks on their staff thinking of biz like this. I mean really... I hadn't heard of this sort of idea before - but when you think about it - comparing the state to the national level makes perfect sense. Really the only time the electoral math matters is if the election is close. So it makes most sense to put money into those states that are most likely to be close if the election is close.

The only reason to play well outside the battle gr...

2008-07-04T16:32:00.000-04:00

The only reason to play well outside the battle ground states is to send a message of power.

Hardly. While that's one reason, there are two other big ones. One is the real chance of winning big: the wider and deeper Obama's win, the more effective power he'll have, not with the voters, but with Congresscritters in 2009.

Another is to expand the set of states that might tip his way even in a close election - to expand the number of paths to 270. In 2004, Kerry's campaign essentially turned into a battle for Ohio, which he lost. The more formerly red states where Obama's support is roughly the same as his national support, the more paths to 270. (Or 269, for that matter. I don't see Obama being handicapped by a H of R win, any more than Bush was handicapped by his 5-4 win in 2000.) And the more options for Obama if it comes down to the wire.

Want to reinforce the last anonymous' comment: Ads...

2008-07-04T15:36:00.000-04:00

Want to reinforce the last anonymous' comment: Ads seem to be placed by Google, based on browser history. Over the last weeks, several people had been mocking about McCain ads on the page, while I was always getting a "Visit Lisboa" ad in German language ..

@ Nate: I feel it is getting time for some update on the regression. Ideally, you would couple it with checks on polls' cross-tabs as to the extent to which your regression is mirroring trends among key demographics. You could then also assess the volatility of demographic groups, as shown in regression / cross tab changes. That would yield further information about not only which states, but also which demographics offer the highest ROI. For example, when going after CO, NM and NV, should you rather target them state-by-state, or through a concerted effort towards key demographics? And which demographics would that be: Latinos, well-educated professionals younger than 45 years, seniors, women? Would a focus on any of the a/m demographics offer spill-over effects to other tipping point states? I suppose, the campaigns are already doing such kind of analysis, but it would be nice to also have your (and the readers) pick on such strategies.

From his reaction, he sure didn't sound like someo...

2008-07-04T15:26:00.000-04:00

From his reaction, he sure didn't sound like someone who likes porn =P My guess is the malware suggestion from above sounds the most likely

re porn addsRecently I've noticed weird adds make ...

2008-07-04T14:43:00.000-04:00

re porn adds
Recently I've noticed weird adds make it to mainstream websites, e.g. of relatively obscure online stores I sometimes browse. I think the browsers started to track that stuff and to place banners appropriate for specific users. I guess if you browse porn, you'd see porn adds :)

People talking about media markets may be overlook...

2008-07-04T14:26:00.000-04:00

People talking about media markets may be overlooking the fact that a large proportionof voters get their TV through cable systems. (I must admit I have no knowledge of how satellite advertising works.)

Such systems permit paying for advertising over only a portion of the overall market coverage. That is, if you wanted to reach Delaware but didn't want to pay for virtually all of SE PA, you would buy time through the cable companies in Delaware. Even on national programs some advertising slots are reserved for local buys, so it is possible to target closely even on national TC shows.

(I just realized it's been close to 30 years since I was invoilved in media buys - too much tempus fidgitted away.)

Nitpick: NM is 5 EVs, so in Cugel's last scenario ...

2008-07-04T14:09:00.000-04:00

Nitpick: NM is 5 EVs, so in Cugel's last scenario you have 273 EVs (although that's pretty easy to infer from context).

But of course, if this only told me things I alrea...

2008-07-04T13:52:00.000-04:00

But of course, if this only told me things I already knew it wouldn't be worthwhile, but there are some nice "hmms" in there like Delaware and Maine.

Delaware, however, is not actually a good return on your investment because to contest it you have to use the Philadelphia media market, which is very expensive. Of course, they're going to be using the Philadelphia media market anyway to contest Pennsylvania, so maybe that evens out.

P.S. I forgot to mention the most plausible winnin...

2008-07-04T11:57:00.000-04:00

P.S. I forgot to mention the most plausible winning scenario from holding New Hampshire would be that then Colorado (9) + New Mexico (4) + Iowa (7) gets you to 272, all of which are very doable, gets you to 272 rather than 269 and a horrible mess in the House of Representatives, after which President Obama starts out crippled instead of with a mandate.

This is an interesting analysis for outsiders to p...

2008-07-04T11:51:00.000-04:00

This is an interesting analysis for outsiders to ponder, but rather like fantasy baseball, the teams aren't doing things that way.

I imagine that Obama has four or five scenarios ranked in order of preference to get him to 270 EV. Based on some core assumptions:

"If we can hold all the Kerry states that gets us to 252, then we just need 18 more EV to win. How can we assure ourselves of getting at least that much?"

That would mean they have to win Michigan and New Hampshire and are currently leading comfortably in both as well as all the other Kerry states, but will need to do some more work to nail down Michigan.

Then the scenarios arise:

Plan A: Ohio 20 EV
Plan B: Pick up Iowa, Colorado and New Mexico +21 EV
Plan C: Any combination leading to 18 EV from): Virginia (13), Nevada (5), Indiana (11)
Plan D: Throw in any of North Carolina (15), North Dakota (3), Montana (3), Florida (27), Missouri (11) as needed to fill in to 270+.

The ROI index might suggest that when executing plan C, you place more emphasis on Indiana than Nevada but New Mexico, Colorado, Iowa, and Ohio were always going to be key investments anyway, because the figure in the most obvious winning scenarios, and New Hampshire's 4 EV are surprisingly important in that holding them enables you to have SEVERAL more plausible winning scenarios.

Of course, none of this will matter if Obama continues to cruise along at +5%. In that case he'll win every one of the states on that list with the possible exception of North Dakota and Montana.

Its not really a 50 state strategy. No one really ...

2008-07-04T11:48:00.000-04:00

Its not really a 50 state strategy. No one really cares if he spends in Utah. If he did he would be called incompetent. At the end of the day, if you win they say your strategy was good. If you lose it was bad. It can be argued that John Kerry's strategy was bad on 04. He has admitted as much. Barrack is taking a lot of heat right now for his "shift" to the center. It will be interesting to see if it was worth the political heat in the long run.