This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page. |
Archive 1 |
I want to thank you all for your suggestions and work on this article. You have contributed greatly to the success and professional appearance of the site. Note that I consider every suggestion with as much care as you took in writing them. Please continue to make suggestions and comments about the article. Your friend PollShark ( talk) 06:32, 22 April 2008 (UTC)
This is a response to JamesMLane's edit summary content about the need for this section: YES. It's not particularly relevant who is ahead in a single state, as the presidency is decided by an aggregation of the state results through the Electoral College. Sure, individual readers could aggregate the results for themselves if they want to know how the big picture looks (and realize that national opinion polls which do not segregate their results by state are dubious indicators at best), but as PollShark has been so diligent as to keep the aggregation up to date for everyone, no one else has to duplicate the work (unless they want to check PS's work). Sketch051 ( talk) 18:43, 20 March 2008 (UTC)
I've made two substantial changes to the state links, neither of which affects the data reflected: (1) The links in the summary (Predicted Results) now point to the state-specific poll results in the main article (2) The state name headers of the state-specific poll result sections are now links to the state's article.
I did this for a string of reasons: the article is long, and scrolling is a nuisance; if you are clicking on a state's name in the results summary, you're more likely interested in the poll results which put that state in a particular candidate's column than in general information about the state; but if you're already looking at a state-specific set of results, you very likely might want to read more about the state to gain a greater understanding of the significance of the poll results.
Hopefully PollShark will still be able to easily cut & paste state references into the appropriate candidate's column as new data are available without any significant increase in the workload. Sketch051 ( talk) 00:08, 27 March 2008 (UTC)
The latest poll shouldn't be included in the electoral thing because it includes Condoleezza Rice.
The poll does include Secretary Rice, however the particular questions that include the general election matchup does not and is not the information I placed on the website. PollShark
What is the difference between the wikipedia article that we are talking about now and this one http://en.wikipedia.org/wiki/Opinion_polling_by_state_for_the_United_States_presidential_election%2C_2008
68.45.9.206 ( talk) 20:25, 11 April 2008 (UTC)
That one has a color-coded map for the hypothetical matchups, this one has tabular data. I see no particular reason why they need to be kept separate, as long as a merged version has: 1) the state-by-state poll data, 2) the tabular aggregate of electoral votes, and 3) the color-coded map. Sketch051 ( talk) 16:24, 14 April 2008 (UTC)
I agree. Marshie71 ( talk) 21:27, 14 April 2008 (UTC)
It's inaccurate for the state of Michigan, Minnesota, New Jersey, New York, Oregon, and Wisconsin. I propose getting rid of it as we already have a text comparison section, and maintaining the graphic isn't being done. —Preceding unsigned comment added by ChPr ( talk • contribs) 19:37, 15 April 2008 (UTC)
CT should be colored light blue for Clinton vs McCain. Jayavarman1 ( talk) 05:46, 29 April 2008 (UTC)
The most recent electoral map is inaccurate, only reflecting the single most recent poll (ie, Obama only won VA by a large margin once, but McCain won VA the two polls before that). Also, Obama isn't winning SC or WV. —Preceding unsigned comment added by Compsci10 ( talk • contribs) 23:30, 9 June 2008 (UTC)
The page that was merged with this one listed how many people were polled in each poll; I find that information useful when doing statistical analysis, and I'd like to see it return. PsyMar ( talk) 00:30, 16 April 2008 (UTC)
I have yet another version of the map. Fairly similar to the one above, except that now we indicate states with very heavy learnings (>10%). Thus the text might look something like this:
What do you think of the new format? Do you like the really dark states and the appearance? If not, we might try a scheme like this one. The Evil Spartan ( talk) 08:11, 29 April 2008 (UTC)
I figure, given that Obama is now presumed to be the nominee, that it's time we focus solely on McCain vs Obama. Also, there has to be time period in which polls have to have been taken to continue being considered valid. It's now May, and some states' latest data goes as far back as February and March. Finally, we have to be sure we update the top of the article along with the bottom. There have been discrepancies between the maps/individual state polling sections and the predicted results section. Fifty7 ( talk) 14:22, 8 May 2008 (UTC)
I don't think so, a lot of polls gave the win to McCain and then one gives a tie between Obama and McCain, I think McCain still wins this state, or at least until two polls in a row place Obama ahead. Someone, please update the states and electoral votes. This is especially considering the fact that a poll made the same day gave McCain a 9 point lead over Obama. So yeah, McCain is winning North Carolina, not too close to call. —Preceding unsigned comment added by 68.243.98.35 ( talk) 19:09, 10 May 2008 (UTC)
Thought I would add some color to our site. Hope you like. Your friend PollShark ( talk) 08:00, 29 May 2008 (UTC)
I have to say that the way this article projects electoral vote totals is flawed. FiveThirtyEight.com's process is much more accurate, as it takes into account everything from the date polls were taken with different weight to the past accuracy of polling organizations. I propose that this article follow the fivethirtyeight.com model rather than simply basing a prediction on the latest poll taken with no regard for how close or how far apart, on average, various state races are. -- Fifty7 ( talk) 16:34, 31 May 2008 (UTC)
What should we do about Hillary Clinton? She has not dropped out, and she's actually relatively close in number of delegates (compared with, say, Ron Paul vs. McCain). I do not believe the content should be removed as this juncture. 64.178.96.168 ( talk) 17:37, 4 June 2008 (UTC)
I was thinking the time has come to separate the polls. From now on, as Survey USA has done recently, most of the polls will be only between McCain and Obama. She's now begun her campaign for Vice President and seems to be resigned to the fact that she's not going to be the nominee. Last night was as close to a concession speech as you'll get from the Clintons. I say lets seperate. PollShark ( talk) 20:31, 4 June 2008 (UTC)
New York Times just reported that Clinton has suspended her campaign effective tommorrow. I guess this moots the point of the poll. PollShark ( talk) 23:24, 4 June 2008 (UTC)
I would just like to know how the presidential seal, white house logo, party logos and campaign symbols are not fair use images. PollShark ( talk) 13:28, 6 June 2008 (UTC)
South Carolina and West Virginia definitely in McCain's hands, but the map shows SC blue. Can someone please change the map and the Electoral Vote count too? Thank you. —Preceding unsigned comment added by Cchow2 ( talk • contribs) 05:26, 10 June 2008 (UTC)
I added the pretty darn credible projections out of Nate Silver's FiveThirtyEight.com, as it is a highly accurate representation of the information presented in the whole article. I'll keep the map updated each day as new polls come in. -- Fifty7 ( talk) 23:36, 12 June 2008 (UTC)
Why is this article called "Statewide..." instead of "State..." or "State by state..."? "Statewide" means "at the state level rather than the local". Either of the others would means what's intended, i.e., "at the state rather than national level". - Rrius ( talk) 08:30, 14 June 2008 (UTC)
I think the date on which maps and final results were last updated should be more visible for all those political junkies who come on the site more than once. —Preceding unsigned comment added by 89.212.193.29 ( talk) 11:43, 14 June 2008 (UTC)
Someone needs to change the color of Florida on the map. A poll was released today that shows Obama leading McCain for the first time ever. RIP Tim Russert. -- Hobie Hunter ( talk) 18:51, 18 June 2008 (UTC)
What is going on with the map? (and I don't mean which state is shaded for which candidate..) It seems like every other day the structure of the map changes or the key changes or the color scheme changes. (i am not referring to the fivethirtyeight map) The map is an incredible resource that I check daily to see updates and general trends, when I do not have the time to digest all of the individual statewide polls. Why can't we just keep it one way and then update it as the polls change accordingly. Why do we need to continuously change the colors and layout of the map? Marshie71 ( talk) 00:47, 20 June 2008 (UTC)
I would like to raise two issues with this map. 1. is what The Evil Spartan raised above. 2. is the fact that this map seems political bias to me. NOT ONE other political website that I have covered down to the detail rates Ohio as "Safe Obama". Not one othere website list Colorado and New Mexico as "Safe Obama", but as "Tossup" or "Leaning Obama". This map is definently not right. I ask that all of you look at Nate Silver's Wikipedia page. It says "Beginning in October 2007, Silver began to publish a diary under the pseudonym "Poblano" on the progressive political blog Daily Kos. Daily Kos is a far left blog. If we are going to put maps on Wikipedia, let's make sure that they are politically correct, and none of the far left, or far right leaning stuff. This just looks like another liberal blog, making liberal maps. America69 ( talk) 01:42, 20 June 2008 (UTC)
The map is getting ridiculous. I'm not here to promote a certain candidate, but there is no way Montana and West Virginia are undecided based on the current polls that are listed - they are for McCain, either safe or solid. Also, North Carolina and Nevada are not undecided either - they are swing McCain. Someone already pointed out Ohio is not solid Obama, but I'd also like to add I wouldn't even consider VA or MI swing Obama - they're closer to undecided. Furthermore, I don't know why several states along the southern border have changed from safe McCain to solid McCain - it is clearly obvious they are safe, not solid. Everytime I come back to this page there's an even more liberal slant than before, and it's reached the point where it's inaccurate. —Preceding unsigned comment added by Compsci10 ( talk • contribs) 04:10, 21 June 2008 (UTC)
I'm the Wikipedia editor who has been managing the FiveThirtyEight map in this article. First, let me say that the worded-assignments in the map (for example, that a margin of 8% or more makes a state "safe") were my own doing, not Silver's, and as we can all agree, that was definitely a mistake. Now, the map currently says Ohio is "Solid Obama", not "Safe", and while the color is correct (the current statistical projection is that Obama would win Ohio by 6.8 points, FiveThirtyEight.com says Ohio is "Lean DEM". Please be assured, I am not here to defend my map as infallible, but only to point out a few things. One is that this map was put together for my own private purposes, and I assigned the titles for margins of victory ("Safe", "Solid", etc.) long before I ever considered putting it up on Wikipedia. In order to put it up on Wikipedia, I obtained permission from 270toWin.com to use their map model here, and then I simply color coded the states according to the day's projections on FiveThirtyEight.com. Now, and this has to be said, to call DailyKos far-left, or at least to assert that by association that Silver has some sort of inherent bias, is just plain uninformed. Silver's a statistician, and the only reason his blog has gotten any attention is that it's been pretty damn accurate so far. I don't consider it to be biased, nor am I trying to push a partisan agenda here. I think I've shown that by admitting the dumb mistake I made, and I'm definitely open to suggestions as to how to improve the map. Note that I'm asking how to improve how to present the reliable, credible information from FiveThirtyEight.com, and not what would be a good substitute for it, as I don't believe such a thing exists. Should the map simply be two colors, red or blue, no matter how small or large the margin is, with the margin of projected victory within each state typed in below the electoral vote total? Or perhaps try to color code it according to the newly-worded designations being used at FiveThirtyEight ("Toss-Up", "Lean", "Likely", "Safe"), which had already led me to consider altering the map this morning? Or maybe some combination (which is where I'm leaning)? Thanks. -- Fifty7 ( talk) 04:48, 21 June 2008 (UTC)
As I type this, based on the average of the last three polls in Michigan which is a six point lead for Obama, that state should be colored medium blue. NOT the light blue that shows as I type this. Steelbeard1 ( talk) 11:48, 28 June 2008 (UTC)
Shouldn't there be a poll listed for the District of Columbia. ⁂†Poison the Well†⁂ ( talk) 21:58, 24 June 2008 (UTC)
why do we have to keep on changing the map ??? adding the different shades of red/blue just make it more confusing and much harder to read. PLEASE PLEASE PLEASE stop making it more complicated and keep things they way they were. —Preceding unsigned comment added by 138.16.29.197 ( talk) 21:46, 28 June 2008 (UTC)
There is a link at the very bottom of the external link section. I was looking at it, and I really don't see why that is there. It is a blog. Why is his blog there. If it stays there, why is the Daily Kos not there, or Politics 1 not there. All of them have polls. I want to know what you think. America69 ( talk) 19:16, 29 June 2008 (UTC)
Using Wikipedia data from this article and Excel, I crunched the numbers in states where the lead in the polls for any one candidate is close or has shown signs of swinging back and forth between Obama and McCain. At this point, I've selected Florida, Pennsylvania, Ohio, Michigan, Virginia, Indiana and Missouri for additional analysis using charts and trendlines. I believe that these seven states are, as it happens, a good representative example, because three of the states are currently leaning toward Obama (Ohio, Michigan and Pennsylvania), three are currently leaning toward McCain (Florida, Indiana and Missouri) and Virginia is virtually tied. Incidentally, between these seven states, there are 120 electoral votes at stake, which represents 44% of the total (270) needed to secure the nomination. In most of the states, there are still at least about 10% of the voters with no opinion yet between the two candidates (or are supporting a third candidate). Your thoughts, suggestions and general input are greatly appreciated. If time is available, I will work to create these charts in the more visually friendly PNG format.-- Robapalooza ( talk) 23:31, 7 July 2008 (UTC)
Zogboy polls should not be included, as you know they are very inaccurate... —Preceding unsigned comment added by 203.84.187.94 ( talk) 03:50, 16 July 2008 (UTC)
Marking the polls with an asterisk is ridiculous. The polls are either scientific and should be included as-is, or they're not and should be removed. It can't be both ways. @Spartan: Why redo the map? Isn't the whole point of using poll averages to prevent a single poll from skewing the results? Rami R 10:20, 20 July 2008 (UTC)
Above all, it's an asterisk. No "huge" change. Smrterthansmrt ( talk) 17:14, 20 July 2008 (UTC)
Anyone have an opinion as to whether War Room Logistics [3] July 7-8 2008 Barack Obama 47.2% John McCain 44.5% is a RS?-- Robapalooza ( talk) 16:48, 17 July 2008 (UTC)
The latest poll for Arizona should be the Rasmussen poll surveyed on June 25 rather than the online Zogby poll. For the Zogby poll, the majority of the data is collected before June 25 since it spams from June 11 to 30. The "median date" is June 20/21 so that's before Rasmussen's June 25. Therefore, I believe it makes more sense to use the Rasmussen poll as the latest poll. (User talk: cchow2|talk) 3:30, 27 July 2008 (UTC)
What is the reasoning behind using Nate's projection map in contrast to any other projection map? Electoral-vote.com also compiles polls and puts an algorithm to them, but they don't have a special section, as do countless other sites. It doesn't seem that his prediction holds any more weight than theirs, so I am just wondering how it fits onto a encyclopedic article, or where the expert consensus is in regards to his research. -Luckydevil713 ( talk) 00:11, 29 July 2008 (UTC)
Almost every state by state poll has a margin of error of +/- 4%. When two candidates are within 4% of each other, they are tied as far as the data can tell. The list of states where each candidate is in the lead under "latest results" should list these states as tied: CO, FL, IN, MI, MT, NV, NH, NC, ND, OH, SD, VA.
I know that there is an urge to declare a winner and to show a definitive result, but things are not that certain and an encyclopedia article should show the facts as they are, with all of the uncertainty. The use of bold to indicate the leader in each poll should likewise be removed when the candidates are within 4% (or within the margin of error) of each other. Bcharles ( talk) 06:51, 3 August 2008 (UTC)
To whoever just fixed the article, THANK YOU. This is the way that it should be. 68.45.9.206 ( talk) 20:14, 3 August 2008 (UTC)
Some great work on this article. But I have a question/problem: can someone explain how this 'Latest results' section is formulated? Taking Missouri as an example, it is in the Republican list. But although the latest polls show a 5% for McCain the previous one shows a 5% for Obama and other recent ones show the state a tie. Strikes me that the overall situation in this state is a tie at the moment or very narrow McCain. Some other states are classed as Rep, Dem or Tie and have similar issues with them. Setwisohi ( talk) 16:07, 3 August 2008 (UTC)
OK, thanks for that. So it's a fairly 'rough' list but does give a general idea of the way the state is tilting? That's good enough for me. Setwisohi ( talk) 19:20, 3 August 2008 (UTC)
Why is Indiana colored light red? If anything, it should be light blue, since Obama is leading in the two most recent polls that we are accepting as credible.
We are not using the Zogby International polls for the "Latest results (using latest polling when available)" section. Therefore, I believe that we shouldn't use the Zogby International polls for the "Maps of most recent polling data" section.
We would then also probably need to change the colors of Arizona and Texas (darker red needed) and any other states that the Zogby polls effected. 68.45.9.206 ( talk) 18:56, 5 August 2008 (UTC)
Polls commissioned from partisan sources are notoriously unreliable; as such, I have recently removed an Alaska poll showing Obama up by 4. Most other sites do not include such polls (e.g., http://fivethirtyeight.com, http://realclearpolitics.com) because they introduce logistical problems into the equation. Pollsters are under pressure to get a favorable result to insure a returning customer, and if the poll doesn't show a positive result, it's usually not released into the public.
The only exception thus far has been some polls commissioned by the dailykos; because they have been done in conjunction with Research2000, an established polling group.
So far, it has been the practice on this page not to introduce such polls (e.g., one commissioned by the Florida Chamber of Commerce, a pro-McCain group); as such we ought to be consistent and leave them out. The Evil Spartan ( talk) 00:04, 13 August 2008 (UTC)
I agree with EvilSpartan. Most major site do not even include that Hays Poll. If you want to include it, fine, but don't use that poll in any calculations. cchow2 ( talk) 12:58, 13 August 2008 (UTC)
It would be interesting to see them on the spredsheets. Or at least "don't know" percentage should be visible. —Preceding unsigned comment added by 213.250.61.151 ( talk) 16:17, 14 August 2008 (UTC)
Good work all those who keep these coming. Amazing how things have changed recently. If it was election day today, it looks like 264 - 262 with just 12 votes undecided. (Montana and Colorado). Could be real cliffhanger. Keep the polls coming! Setwisohi ( talk) 15:18, 21 August 2008 (UTC)
Every news outlet is reporting that Palin will be McCain's running mate. I've added her to the article along with a photo I found on Commons. If someone can find a better one, that would rock. Henrymrx ( talk) 15:15, 29 August 2008 (UTC)
Joe Biden's picture is missing! Who took it? —Preceding unsigned comment added by 89.212.2.86 ( talk) 18:28, 29 August 2008 (UTC)
Where is this information coming from? Where the references?-- Rtphokie ( talk) 03:40, 1 September 2008 (UTC)
I've archived all polls older than 3 months, provided that there are at least three poll without a star next to them for the state. The polls really don't help that much anyway, and the page is getting very large. You can find the archive at Talk:Statewide opinion polling for the United States presidential election, 2008/Archive. The Evil Spartan ( talk) 20:31, 5 September 2008 (UTC)
I have had a look at the Rasmussen Reports balance of power calculator - while I think it's mostly accurate, it isn't based on the same polls as ours, and seems to leave several pretty safe states out. I propose one of two different methods of determining a safe state: 1) A state has never polled for the opponent or as a tie. States with only a few polls (e.g., Delaware) shouldn't be a problem because pollsters (at this point) have only really ignored states that are already obvious. 2) A state has never polled within 5 points for the opponent. This is a little less inclusive, but looks better to me - it would remove states like Minnesota, Washington, and North Carolina which are not really safe.
Thoughts? The Evil Spartan ( talk) 21:35, 9 September 2008 (UTC)
I think we can follow Rasmussen's method and make a category called 'Likely States". Obama is definitely a lot stronger in Deleware than PA, but right now they are in the same category (Lean Dem). McCain is definitely a lot stronger in TX than OH, but right now they are in the same category (Lean GOP).
cchow2 ( talk) 21:00, 11 September 2008 (UTC)
Agree with cchow2 - we need to add a 'Likely' column for the reasons given Setwisohi ( talk) 09:17, 13 September 2008 (UTC)
Just a thought, shouldn't Palin's 'outdoor' profile image be replaced with the same staid 'formal' image as the others? (Or, alternatively, the others be replaced with rather more endearing images?). Setwisohi ( talk) 09:21, 13 September 2008 (UTC)
Indiana should be removed in the battleground graphs... it's not a battleground state anymore... should be replaced with Colorado instead —Preceding unsigned comment added by 125.60.235.231 ( talk) 02:26, 14 September 2008 (UTC)
Is it possible to match the map with poll results? Today Obama is leading on the polls but not on the map: it's confusing. Thanks -- Jaykb ( talk) 09:09, 18 September 2008 (UTC)
According to the latest poll, he is ahead... I think those results need revision! —Preceding unsigned comment added by 99.202.249.245 ( talk) 12:44, 20 September 2008 (UTC)
I find it strange that in this page polls & especially maps are edited by only a single (conservative) contributor. This might explain your question Why isn't Washington state among the states Obama is ahead in?. I think that one of the problem is that maps do not match with the current polls. I fear a manipulation of these maps that overestimate McCain's results on the statewide polls through map colors. -- Jaykb ( talk) 19:33, 21 September 2008 (UTC)
Allright! I am glad to see there are several people interested inthis issue, not just Evil spartan who is doing all the work with maps. I tryed to modify them but it is true it takes me very long and maybe it is not possible to change it every day eve if I'd like to. Therefore I also agree on good faith for Evil Spartan but maybe wie could help him if we can? -- Jaykb ( talk) 21:08, 25 September 2008 (UTC)
I've been fixing the tables recently that are usually messed up by various other candidates. But I was thinking today: Why don't we have their images at the top of the page with Obama and McCain? Additionally, do we need images of Biden and Palin there, as the polling on this page is indirectly about them, and their names only appear once (in the Latest results column) and with that of their running mates.-- Patrick «» 17:55, 25 September 2008 (UTC)
I'm having issues with this disclaimer:
Notes:
With shifting opinion polls including those from Virginia and North Carolina shifting from McCain to Obama as well as Michigan solidifying Obama's support, I feel that the disclaimer should be removed so that it reflects web sites such as http://www.electoral-vote.com and http://www.fivethirtyeight.com which quickly change a state's status as the polling changes. This would not apply to the maps in this article as long as it indicates the date that the map was last revised. Steelbeard1 ( talk) 17:27, 27 September 2008 (UTC)
The colors of the electoral map are much easier to see now. NS Zakeruga ( talk) 20:57, 29 September 2008 (UTC)
Now that there is a new table reflecting the latest state polling results, I think we can kill the straight list which has become misleading because of the disclaimer forcing states such as Michigan which show solid leads which are forced to show thin leads. Steelbeard1 ( talk) 16:55, 2 October 2008 (UTC)
Steelbeard, I am reverting your latest edit. You did this backwards. You should have brought up the deletion here first. Instead, you are expecting others to justify restoring it. You're the one proposing the change, so you need to justify your position first, especially when we're talking about deletion of material that is properly sourced. Henrymrx ( talk) 18:25, 2 October 2008 (UTC)
The Evil Spartan started the new table as shown by the edit in [6]. If you are reading this, Evil Spartan, can you add your input? Steelbeard1 ( talk) 22:14, 2 October 2008 (UTC)
I've marked this as an organization affiliated with a particular group. Democracy Corps is a group very similar to MoveOn.org; this alone wouldn't be enough, because we've included Research 2000/Daily Kos polls before. However, I am particularly worried about the statement on the GQR website:
Greenberg Quinlan Rosner is a global leader in public opinion research and strategic consulting. Whether you want to win your election, lead your country, increase your bottom line, or change the world, Greenberg Quinlan Rosner can help you find the answer.
This reads to me like several of the other polling groups I've seen that try to find a result friendly to the group paying for the result. The Evil Spartan ( talk) 17:49, 3 October 2008 (UTC)
Why is this Colorado poll marked as partisan?
198.185.18.207 ( talk) 20:27, 3 October 2008 (UTC)
Which version should we use? Any preferences? 90.231.2.252 ( talk) 21:34, 3 October 2008 (UTC)
Personally I prefer the one picked by User:JaGa - but reverted their choice in case people felt strongly about it. 90.231.2.252 ( talk) 21:36, 3 October 2008 (UTC)
Choice A
Choice B
Choice C
I am confused by how we are categorizing states as safe, likely, or leaning. What makes New Hampshire a "safe" Obama state but Wisconsin, a state in which McCain has apparently never led, just "likely"? I have observed a lot of movement lately, not so much of swinging leaning states (which are easier to categorize) but "leaning" to "likely" and "likely" to "safe" in Obama's columns. Minutiaman ( talk) 23:25, 6 October 2008 (UTC)
To whoever suggested that multiple polls 'concluded' on the same day be averaged out, that's a good idea to end the confusion. Steelbeard1 ( talk) 15:09, 7 October 2008 (UTC)
In the article, it is implied that "a poll commissioned by an organization affiliated with a specific candidate" should be marked with "**" and it is stated that such a poll should "not [be] used in calculating results below". At Real Clear Politics, certain pollsters are regularly marked with "(D)" and with "(R)". Assuming this stands for party affiliation, shouldn't such polls be marked here as well and be excluded from calculations in general? On the other hand: This would apply to all polls conducted by "Public Policy Polling" (D) and "Strategic Vision" (R), two polling firms which have relatively good marks in the Pollster Ratings at the "FiveThirtyEight" website. I want to raise this issue, because I just posted a poll by "Public Policy Polling" which shows a relatively large lead for Obama in Virginia. FanofPolling ( talk) 15:38, 9 October 2008 (UTC)
Isn't the colored map in the "Maps of recent polling data" section sort of the definition of original research? Presenting recent polls in this article is fine, but then doing a weighted average of the previous 10, selecting which polls to present in the first place, picking a weighting algorithm, and picking thresholds for different colors are all original analysis. If we want to present a map, I think we need to pick some reliable source and use their rolled up state-by-state analysis, from which we could create our own map (that would probably end up looking very similar to their map). -- Rick Block ( talk) 16:25, 12 October 2008 (UTC)
Upon further thought, I am thinking of moving this to Wikinews, with a prominent link from this page. I would do so myself if I wasn't working 8000 hours per week. Original research is not a problem there, from what I can tell. The Evil Spartan ( talk) 19:52, 17 October 2008 (UTC)
The information accompanying the "recent polling" maps states that North Dakota is in the Leans McCain category while its shade of red on the maps suggests it is in the Safe McCain category. One or the other is not up to date. —Preceding unsigned comment added by 74.178.203.248 ( talk • contribs)
Because of the recent edit warring, I had requested semi-protection for this article. It was declined for reasons given at [7]. If any administrators are editing this article, can you check out the IP addresses carrying out the edit warring and block them if necessary? Steelbeard1 ( talk) 17:01, 22 October 2008 (UTC)
I think the regular contributors here should take note of the "Lead Margin" column in the National polling article. This visual appeal may be better than the comparison of all those numbers and boldings. -- Natural RX 16:29, 23 October 2008 (UTC)
For example, Missouri would look like so:
Poll Source | Date administered | Democrat | % | Republican | % | Lead Margin |
---|---|---|---|---|---|---|
Rasmussen Reports/Fox News | October 26 2008 | Barack Obama | 48% | John McCain | 47% | 1 |
Survey USA | October 25- 26 2008 | Barack Obama | 48% | John McCain | 48% | 0 |
Reuters/Zogby | October 23- 26 2008 | Barack Obama | 48.2% | John McCain | 45.7% | 2.5 |
Mason-Dixon/NBC | October 22- 23 2008 | Barack Obama | 45% | John McCain | 46% | 1 |
Research 2000/Post-Dispatch | October 20- 23 2008 | Barack Obama | 48% | John McCain | 47% | 1 |
Zogby Interactive * | October 17- 20 2008 | Barack Obama | 48% | John McCain | 48.3% | 0.3 |
Fox News/Rasmussen Reports | October 19, 2008 | Barack Obama | 49% | John McCain | 44% | 5 |
Suffolk University | October 17- 19 2008 | Barack Obama | 44% | John McCain | 45% | 1 |
Rasmussen Reports | October 15, 2008 | Barack Obama | 52% | John McCain | 46% | 6 |
CNN/Time Magazine/Opinion Research | October 11- 14 2008 | Barack Obama | 48% | John McCain | 49% | 1 |
Zogby Interactive * | October 9- 13 2008 | Barack Obama | 50% | John McCain | 44% | 6 |
Fox News/Rasmussen Reports | October 12, 2008 | Barack Obama | 50% | John McCain | 47% | 3 |
Public Policy Polling | October 11- 12 2008 | Barack Obama | 48% | John McCain | 46% | 2 |
SurveyUSA | October 10- 12 2008 | Barack Obama | 51% | John McCain | 43% | 8 |
American Research Group | October 4- 6 2008 | Barack Obama | 46% | John McCain | 49% | 3 |
Fox News/Rasmussen Reports | October 5 2008 | Barack Obama | 50% | John McCain | 47% | 3 |
CNN/Time Magazine/ Opinion Research |
September 28- 30 2008 | Barack Obama | 49% | John McCain | 48% | 1 |
Survey USA | September 23- 24 2008 | Barack Obama | 46% | John McCain | 48% | 2 |
St Louis Post-Dispatch/Research 2000 | September 22- 24 2008 | Barack Obama | 47% | John McCain | 48% | 1 |
St Louis Post-Dispatch/Research 2000 | September 15- 18 2008 | Barack Obama | 45% | John McCain | 49% | 4 |
American Research Group | September 11- 15 2008 | Barack Obama | 45% | John McCain | 50% | 5 |
Zogby Interactive * | September 9- 12 2008 | Barack Obama | 42.4% | John McCain | 48.5% | 6.1 |
Rasmussen Reports | September 11, 2008 | Barack Obama | 46% | John McCain | 51% | 5 |
Time/CNN/Opinion Research | September 7- 9 2008 | Barack Obama | 44% | John McCain | 48% | 4 |
Public Policy Polling | August 13- 17 2008 | Barack Obama | 40% | John McCain | 50% | 10 |
Rasmussen Reports | August 7 2008 | Barack Obama | 44% | John McCain | 50% | 6 |
Survey USA | July 29- 31 2008 | Barack Obama | 44% | John McCain | 49% | 5 |
Research 2000/ St. Louis Post Dispatch |
July 7- 10 2008 | Barack Obama | 48% | John McCain | 43% | 5 |
Rasmussen Reports | July 7 2008 | Barack Obama | 42% | John McCain | 47% | 5 |
Public Policy Polling | July 2- 5 2008 | Barack Obama | 44% | John McCain | 47% | 3 |
Thoughts? -- Natural RX 19:20, 29 October 2008 (UTC)
I started with the above Missouri example and did Alabama. I encourage others to include the lead margin for the other states, but I have not yet figured out how to do it in states with minor party candidate fields inserted. Steelbeard1 ( talk) 15:45, 30 October 2008 (UTC)
I think the election results should be placed in this article so that the poll accuracy could be described for each state. See this!
This could be done like this:
Poll Source | Date administered | Democrat | % | Republican | % | Lead Margin |
---|---|---|---|---|---|---|
Election result | November 4 2008 | Barack Obama | 38.8% | John McCain | 60.4% | 21,6
|
Survey USA | October 27- 28 2008 | Barack Obama | 36% | John McCain | 61% | 25
|
Capital Survey Research Center | October 15- 21 2008 | Barack Obama | 34% | John McCain | 54% | 20
|
Survey USA | October 8- 9 2008 | Barack Obama | 35% | John McCain | 62% | 27
|
Capital Research Center | October 6- 7, 2008 | Barack Obama | 35% | John McCain | 54.5% | 19.5
|
Rasmussen Reports | September 22, 2008 | Barack Obama | 39% | John McCain | 60% | 21
|
Survey USA | September 16- 17 2008 | Barack Obama | 34% | John McCain | 64% | 30
|
American Research Group | September 13- 16 2008 | Barack Obama | 36% | John McCain | 58% | 22
|
University of South Alabama/ Press-Register |
September 8- 15 2008 | Barack Obama | 25% | John McCain | 52% | 27
|
Capital Survey Research Center/ Alabama Education Association |
September 3- 9 2008 | Barack Obama | 35% | John McCain | 55% | 20
|
Capital Survey Research Center/ Alabama Education Association |
July 29- August 4 2008 | Barack Obama | 34% | John McCain | 47% | 13
|
Rasmussen Reports | July 31, 2008 | Barack Obama | 38% | John McCain | 58% | 20
|
This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page. |
Archive 1 |
I want to thank you all for your suggestions and work on this article. You have contributed greatly to the success and professional appearance of the site. Note that I consider every suggestion with as much care as you took in writing them. Please continue to make suggestions and comments about the article. Your friend PollShark ( talk) 06:32, 22 April 2008 (UTC)
This is a response to JamesMLane's edit summary content about the need for this section: YES. It's not particularly relevant who is ahead in a single state, as the presidency is decided by an aggregation of the state results through the Electoral College. Sure, individual readers could aggregate the results for themselves if they want to know how the big picture looks (and realize that national opinion polls which do not segregate their results by state are dubious indicators at best), but as PollShark has been so diligent as to keep the aggregation up to date for everyone, no one else has to duplicate the work (unless they want to check PS's work). Sketch051 ( talk) 18:43, 20 March 2008 (UTC)
I've made two substantial changes to the state links, neither of which affects the data reflected: (1) The links in the summary (Predicted Results) now point to the state-specific poll results in the main article (2) The state name headers of the state-specific poll result sections are now links to the state's article.
I did this for a string of reasons: the article is long, and scrolling is a nuisance; if you are clicking on a state's name in the results summary, you're more likely interested in the poll results which put that state in a particular candidate's column than in general information about the state; but if you're already looking at a state-specific set of results, you very likely might want to read more about the state to gain a greater understanding of the significance of the poll results.
Hopefully PollShark will still be able to easily cut & paste state references into the appropriate candidate's column as new data are available without any significant increase in the workload. Sketch051 ( talk) 00:08, 27 March 2008 (UTC)
The latest poll shouldn't be included in the electoral thing because it includes Condoleezza Rice.
The poll does include Secretary Rice, however the particular questions that include the general election matchup does not and is not the information I placed on the website. PollShark
What is the difference between the wikipedia article that we are talking about now and this one http://en.wikipedia.org/wiki/Opinion_polling_by_state_for_the_United_States_presidential_election%2C_2008
68.45.9.206 ( talk) 20:25, 11 April 2008 (UTC)
That one has a color-coded map for the hypothetical matchups, this one has tabular data. I see no particular reason why they need to be kept separate, as long as a merged version has: 1) the state-by-state poll data, 2) the tabular aggregate of electoral votes, and 3) the color-coded map. Sketch051 ( talk) 16:24, 14 April 2008 (UTC)
I agree. Marshie71 ( talk) 21:27, 14 April 2008 (UTC)
It's inaccurate for the state of Michigan, Minnesota, New Jersey, New York, Oregon, and Wisconsin. I propose getting rid of it as we already have a text comparison section, and maintaining the graphic isn't being done. —Preceding unsigned comment added by ChPr ( talk • contribs) 19:37, 15 April 2008 (UTC)
CT should be colored light blue for Clinton vs McCain. Jayavarman1 ( talk) 05:46, 29 April 2008 (UTC)
The most recent electoral map is inaccurate, only reflecting the single most recent poll (ie, Obama only won VA by a large margin once, but McCain won VA the two polls before that). Also, Obama isn't winning SC or WV. —Preceding unsigned comment added by Compsci10 ( talk • contribs) 23:30, 9 June 2008 (UTC)
The page that was merged with this one listed how many people were polled in each poll; I find that information useful when doing statistical analysis, and I'd like to see it return. PsyMar ( talk) 00:30, 16 April 2008 (UTC)
I have yet another version of the map. Fairly similar to the one above, except that now we indicate states with very heavy learnings (>10%). Thus the text might look something like this:
What do you think of the new format? Do you like the really dark states and the appearance? If not, we might try a scheme like this one. The Evil Spartan ( talk) 08:11, 29 April 2008 (UTC)
I figure, given that Obama is now presumed to be the nominee, that it's time we focus solely on McCain vs Obama. Also, there has to be time period in which polls have to have been taken to continue being considered valid. It's now May, and some states' latest data goes as far back as February and March. Finally, we have to be sure we update the top of the article along with the bottom. There have been discrepancies between the maps/individual state polling sections and the predicted results section. Fifty7 ( talk) 14:22, 8 May 2008 (UTC)
I don't think so, a lot of polls gave the win to McCain and then one gives a tie between Obama and McCain, I think McCain still wins this state, or at least until two polls in a row place Obama ahead. Someone, please update the states and electoral votes. This is especially considering the fact that a poll made the same day gave McCain a 9 point lead over Obama. So yeah, McCain is winning North Carolina, not too close to call. —Preceding unsigned comment added by 68.243.98.35 ( talk) 19:09, 10 May 2008 (UTC)
Thought I would add some color to our site. Hope you like. Your friend PollShark ( talk) 08:00, 29 May 2008 (UTC)
I have to say that the way this article projects electoral vote totals is flawed. FiveThirtyEight.com's process is much more accurate, as it takes into account everything from the date polls were taken with different weight to the past accuracy of polling organizations. I propose that this article follow the fivethirtyeight.com model rather than simply basing a prediction on the latest poll taken with no regard for how close or how far apart, on average, various state races are. -- Fifty7 ( talk) 16:34, 31 May 2008 (UTC)
What should we do about Hillary Clinton? She has not dropped out, and she's actually relatively close in number of delegates (compared with, say, Ron Paul vs. McCain). I do not believe the content should be removed as this juncture. 64.178.96.168 ( talk) 17:37, 4 June 2008 (UTC)
I was thinking the time has come to separate the polls. From now on, as Survey USA has done recently, most of the polls will be only between McCain and Obama. She's now begun her campaign for Vice President and seems to be resigned to the fact that she's not going to be the nominee. Last night was as close to a concession speech as you'll get from the Clintons. I say lets seperate. PollShark ( talk) 20:31, 4 June 2008 (UTC)
New York Times just reported that Clinton has suspended her campaign effective tommorrow. I guess this moots the point of the poll. PollShark ( talk) 23:24, 4 June 2008 (UTC)
I would just like to know how the presidential seal, white house logo, party logos and campaign symbols are not fair use images. PollShark ( talk) 13:28, 6 June 2008 (UTC)
South Carolina and West Virginia definitely in McCain's hands, but the map shows SC blue. Can someone please change the map and the Electoral Vote count too? Thank you. —Preceding unsigned comment added by Cchow2 ( talk • contribs) 05:26, 10 June 2008 (UTC)
I added the pretty darn credible projections out of Nate Silver's FiveThirtyEight.com, as it is a highly accurate representation of the information presented in the whole article. I'll keep the map updated each day as new polls come in. -- Fifty7 ( talk) 23:36, 12 June 2008 (UTC)
Why is this article called "Statewide..." instead of "State..." or "State by state..."? "Statewide" means "at the state level rather than the local". Either of the others would means what's intended, i.e., "at the state rather than national level". - Rrius ( talk) 08:30, 14 June 2008 (UTC)
I think the date on which maps and final results were last updated should be more visible for all those political junkies who come on the site more than once. —Preceding unsigned comment added by 89.212.193.29 ( talk) 11:43, 14 June 2008 (UTC)
Someone needs to change the color of Florida on the map. A poll was released today that shows Obama leading McCain for the first time ever. RIP Tim Russert. -- Hobie Hunter ( talk) 18:51, 18 June 2008 (UTC)
What is going on with the map? (and I don't mean which state is shaded for which candidate..) It seems like every other day the structure of the map changes or the key changes or the color scheme changes. (i am not referring to the fivethirtyeight map) The map is an incredible resource that I check daily to see updates and general trends, when I do not have the time to digest all of the individual statewide polls. Why can't we just keep it one way and then update it as the polls change accordingly. Why do we need to continuously change the colors and layout of the map? Marshie71 ( talk) 00:47, 20 June 2008 (UTC)
I would like to raise two issues with this map. 1. is what The Evil Spartan raised above. 2. is the fact that this map seems political bias to me. NOT ONE other political website that I have covered down to the detail rates Ohio as "Safe Obama". Not one othere website list Colorado and New Mexico as "Safe Obama", but as "Tossup" or "Leaning Obama". This map is definently not right. I ask that all of you look at Nate Silver's Wikipedia page. It says "Beginning in October 2007, Silver began to publish a diary under the pseudonym "Poblano" on the progressive political blog Daily Kos. Daily Kos is a far left blog. If we are going to put maps on Wikipedia, let's make sure that they are politically correct, and none of the far left, or far right leaning stuff. This just looks like another liberal blog, making liberal maps. America69 ( talk) 01:42, 20 June 2008 (UTC)
The map is getting ridiculous. I'm not here to promote a certain candidate, but there is no way Montana and West Virginia are undecided based on the current polls that are listed - they are for McCain, either safe or solid. Also, North Carolina and Nevada are not undecided either - they are swing McCain. Someone already pointed out Ohio is not solid Obama, but I'd also like to add I wouldn't even consider VA or MI swing Obama - they're closer to undecided. Furthermore, I don't know why several states along the southern border have changed from safe McCain to solid McCain - it is clearly obvious they are safe, not solid. Everytime I come back to this page there's an even more liberal slant than before, and it's reached the point where it's inaccurate. —Preceding unsigned comment added by Compsci10 ( talk • contribs) 04:10, 21 June 2008 (UTC)
I'm the Wikipedia editor who has been managing the FiveThirtyEight map in this article. First, let me say that the worded-assignments in the map (for example, that a margin of 8% or more makes a state "safe") were my own doing, not Silver's, and as we can all agree, that was definitely a mistake. Now, the map currently says Ohio is "Solid Obama", not "Safe", and while the color is correct (the current statistical projection is that Obama would win Ohio by 6.8 points, FiveThirtyEight.com says Ohio is "Lean DEM". Please be assured, I am not here to defend my map as infallible, but only to point out a few things. One is that this map was put together for my own private purposes, and I assigned the titles for margins of victory ("Safe", "Solid", etc.) long before I ever considered putting it up on Wikipedia. In order to put it up on Wikipedia, I obtained permission from 270toWin.com to use their map model here, and then I simply color coded the states according to the day's projections on FiveThirtyEight.com. Now, and this has to be said, to call DailyKos far-left, or at least to assert that by association that Silver has some sort of inherent bias, is just plain uninformed. Silver's a statistician, and the only reason his blog has gotten any attention is that it's been pretty damn accurate so far. I don't consider it to be biased, nor am I trying to push a partisan agenda here. I think I've shown that by admitting the dumb mistake I made, and I'm definitely open to suggestions as to how to improve the map. Note that I'm asking how to improve how to present the reliable, credible information from FiveThirtyEight.com, and not what would be a good substitute for it, as I don't believe such a thing exists. Should the map simply be two colors, red or blue, no matter how small or large the margin is, with the margin of projected victory within each state typed in below the electoral vote total? Or perhaps try to color code it according to the newly-worded designations being used at FiveThirtyEight ("Toss-Up", "Lean", "Likely", "Safe"), which had already led me to consider altering the map this morning? Or maybe some combination (which is where I'm leaning)? Thanks. -- Fifty7 ( talk) 04:48, 21 June 2008 (UTC)
As I type this, based on the average of the last three polls in Michigan which is a six point lead for Obama, that state should be colored medium blue. NOT the light blue that shows as I type this. Steelbeard1 ( talk) 11:48, 28 June 2008 (UTC)
Shouldn't there be a poll listed for the District of Columbia. ⁂†Poison the Well†⁂ ( talk) 21:58, 24 June 2008 (UTC)
why do we have to keep on changing the map ??? adding the different shades of red/blue just make it more confusing and much harder to read. PLEASE PLEASE PLEASE stop making it more complicated and keep things they way they were. —Preceding unsigned comment added by 138.16.29.197 ( talk) 21:46, 28 June 2008 (UTC)
There is a link at the very bottom of the external link section. I was looking at it, and I really don't see why that is there. It is a blog. Why is his blog there. If it stays there, why is the Daily Kos not there, or Politics 1 not there. All of them have polls. I want to know what you think. America69 ( talk) 19:16, 29 June 2008 (UTC)
Using Wikipedia data from this article and Excel, I crunched the numbers in states where the lead in the polls for any one candidate is close or has shown signs of swinging back and forth between Obama and McCain. At this point, I've selected Florida, Pennsylvania, Ohio, Michigan, Virginia, Indiana and Missouri for additional analysis using charts and trendlines. I believe that these seven states are, as it happens, a good representative example, because three of the states are currently leaning toward Obama (Ohio, Michigan and Pennsylvania), three are currently leaning toward McCain (Florida, Indiana and Missouri) and Virginia is virtually tied. Incidentally, between these seven states, there are 120 electoral votes at stake, which represents 44% of the total (270) needed to secure the nomination. In most of the states, there are still at least about 10% of the voters with no opinion yet between the two candidates (or are supporting a third candidate). Your thoughts, suggestions and general input are greatly appreciated. If time is available, I will work to create these charts in the more visually friendly PNG format.-- Robapalooza ( talk) 23:31, 7 July 2008 (UTC)
Zogboy polls should not be included, as you know they are very inaccurate... —Preceding unsigned comment added by 203.84.187.94 ( talk) 03:50, 16 July 2008 (UTC)
Marking the polls with an asterisk is ridiculous. The polls are either scientific and should be included as-is, or they're not and should be removed. It can't be both ways. @Spartan: Why redo the map? Isn't the whole point of using poll averages to prevent a single poll from skewing the results? Rami R 10:20, 20 July 2008 (UTC)
Above all, it's an asterisk. No "huge" change. Smrterthansmrt ( talk) 17:14, 20 July 2008 (UTC)
Anyone have an opinion as to whether War Room Logistics [3] July 7-8 2008 Barack Obama 47.2% John McCain 44.5% is a RS?-- Robapalooza ( talk) 16:48, 17 July 2008 (UTC)
The latest poll for Arizona should be the Rasmussen poll surveyed on June 25 rather than the online Zogby poll. For the Zogby poll, the majority of the data is collected before June 25 since it spams from June 11 to 30. The "median date" is June 20/21 so that's before Rasmussen's June 25. Therefore, I believe it makes more sense to use the Rasmussen poll as the latest poll. (User talk: cchow2|talk) 3:30, 27 July 2008 (UTC)
What is the reasoning behind using Nate's projection map in contrast to any other projection map? Electoral-vote.com also compiles polls and puts an algorithm to them, but they don't have a special section, as do countless other sites. It doesn't seem that his prediction holds any more weight than theirs, so I am just wondering how it fits onto a encyclopedic article, or where the expert consensus is in regards to his research. -Luckydevil713 ( talk) 00:11, 29 July 2008 (UTC)
Almost every state by state poll has a margin of error of +/- 4%. When two candidates are within 4% of each other, they are tied as far as the data can tell. The list of states where each candidate is in the lead under "latest results" should list these states as tied: CO, FL, IN, MI, MT, NV, NH, NC, ND, OH, SD, VA.
I know that there is an urge to declare a winner and to show a definitive result, but things are not that certain and an encyclopedia article should show the facts as they are, with all of the uncertainty. The use of bold to indicate the leader in each poll should likewise be removed when the candidates are within 4% (or within the margin of error) of each other. Bcharles ( talk) 06:51, 3 August 2008 (UTC)
To whoever just fixed the article, THANK YOU. This is the way that it should be. 68.45.9.206 ( talk) 20:14, 3 August 2008 (UTC)
Some great work on this article. But I have a question/problem: can someone explain how this 'Latest results' section is formulated? Taking Missouri as an example, it is in the Republican list. But although the latest polls show a 5% for McCain the previous one shows a 5% for Obama and other recent ones show the state a tie. Strikes me that the overall situation in this state is a tie at the moment or very narrow McCain. Some other states are classed as Rep, Dem or Tie and have similar issues with them. Setwisohi ( talk) 16:07, 3 August 2008 (UTC)
OK, thanks for that. So it's a fairly 'rough' list but does give a general idea of the way the state is tilting? That's good enough for me. Setwisohi ( talk) 19:20, 3 August 2008 (UTC)
Why is Indiana colored light red? If anything, it should be light blue, since Obama is leading in the two most recent polls that we are accepting as credible.
We are not using the Zogby International polls for the "Latest results (using latest polling when available)" section. Therefore, I believe that we shouldn't use the Zogby International polls for the "Maps of most recent polling data" section.
We would then also probably need to change the colors of Arizona and Texas (darker red needed) and any other states that the Zogby polls effected. 68.45.9.206 ( talk) 18:56, 5 August 2008 (UTC)
Polls commissioned from partisan sources are notoriously unreliable; as such, I have recently removed an Alaska poll showing Obama up by 4. Most other sites do not include such polls (e.g., http://fivethirtyeight.com, http://realclearpolitics.com) because they introduce logistical problems into the equation. Pollsters are under pressure to get a favorable result to insure a returning customer, and if the poll doesn't show a positive result, it's usually not released into the public.
The only exception thus far has been some polls commissioned by the dailykos; because they have been done in conjunction with Research2000, an established polling group.
So far, it has been the practice on this page not to introduce such polls (e.g., one commissioned by the Florida Chamber of Commerce, a pro-McCain group); as such we ought to be consistent and leave them out. The Evil Spartan ( talk) 00:04, 13 August 2008 (UTC)
I agree with EvilSpartan. Most major site do not even include that Hays Poll. If you want to include it, fine, but don't use that poll in any calculations. cchow2 ( talk) 12:58, 13 August 2008 (UTC)
It would be interesting to see them on the spredsheets. Or at least "don't know" percentage should be visible. —Preceding unsigned comment added by 213.250.61.151 ( talk) 16:17, 14 August 2008 (UTC)
Good work all those who keep these coming. Amazing how things have changed recently. If it was election day today, it looks like 264 - 262 with just 12 votes undecided. (Montana and Colorado). Could be real cliffhanger. Keep the polls coming! Setwisohi ( talk) 15:18, 21 August 2008 (UTC)
Every news outlet is reporting that Palin will be McCain's running mate. I've added her to the article along with a photo I found on Commons. If someone can find a better one, that would rock. Henrymrx ( talk) 15:15, 29 August 2008 (UTC)
Joe Biden's picture is missing! Who took it? —Preceding unsigned comment added by 89.212.2.86 ( talk) 18:28, 29 August 2008 (UTC)
Where is this information coming from? Where the references?-- Rtphokie ( talk) 03:40, 1 September 2008 (UTC)
I've archived all polls older than 3 months, provided that there are at least three poll without a star next to them for the state. The polls really don't help that much anyway, and the page is getting very large. You can find the archive at Talk:Statewide opinion polling for the United States presidential election, 2008/Archive. The Evil Spartan ( talk) 20:31, 5 September 2008 (UTC)
I have had a look at the Rasmussen Reports balance of power calculator - while I think it's mostly accurate, it isn't based on the same polls as ours, and seems to leave several pretty safe states out. I propose one of two different methods of determining a safe state: 1) A state has never polled for the opponent or as a tie. States with only a few polls (e.g., Delaware) shouldn't be a problem because pollsters (at this point) have only really ignored states that are already obvious. 2) A state has never polled within 5 points for the opponent. This is a little less inclusive, but looks better to me - it would remove states like Minnesota, Washington, and North Carolina which are not really safe.
Thoughts? The Evil Spartan ( talk) 21:35, 9 September 2008 (UTC)
I think we can follow Rasmussen's method and make a category called 'Likely States". Obama is definitely a lot stronger in Deleware than PA, but right now they are in the same category (Lean Dem). McCain is definitely a lot stronger in TX than OH, but right now they are in the same category (Lean GOP).
cchow2 ( talk) 21:00, 11 September 2008 (UTC)
Agree with cchow2 - we need to add a 'Likely' column for the reasons given Setwisohi ( talk) 09:17, 13 September 2008 (UTC)
Just a thought, shouldn't Palin's 'outdoor' profile image be replaced with the same staid 'formal' image as the others? (Or, alternatively, the others be replaced with rather more endearing images?). Setwisohi ( talk) 09:21, 13 September 2008 (UTC)
Indiana should be removed in the battleground graphs... it's not a battleground state anymore... should be replaced with Colorado instead —Preceding unsigned comment added by 125.60.235.231 ( talk) 02:26, 14 September 2008 (UTC)
Is it possible to match the map with poll results? Today Obama is leading on the polls but not on the map: it's confusing. Thanks -- Jaykb ( talk) 09:09, 18 September 2008 (UTC)
According to the latest poll, he is ahead... I think those results need revision! —Preceding unsigned comment added by 99.202.249.245 ( talk) 12:44, 20 September 2008 (UTC)
I find it strange that in this page polls & especially maps are edited by only a single (conservative) contributor. This might explain your question Why isn't Washington state among the states Obama is ahead in?. I think that one of the problem is that maps do not match with the current polls. I fear a manipulation of these maps that overestimate McCain's results on the statewide polls through map colors. -- Jaykb ( talk) 19:33, 21 September 2008 (UTC)
Allright! I am glad to see there are several people interested inthis issue, not just Evil spartan who is doing all the work with maps. I tryed to modify them but it is true it takes me very long and maybe it is not possible to change it every day eve if I'd like to. Therefore I also agree on good faith for Evil Spartan but maybe wie could help him if we can? -- Jaykb ( talk) 21:08, 25 September 2008 (UTC)
I've been fixing the tables recently that are usually messed up by various other candidates. But I was thinking today: Why don't we have their images at the top of the page with Obama and McCain? Additionally, do we need images of Biden and Palin there, as the polling on this page is indirectly about them, and their names only appear once (in the Latest results column) and with that of their running mates.-- Patrick «» 17:55, 25 September 2008 (UTC)
I'm having issues with this disclaimer:
Notes:
With shifting opinion polls including those from Virginia and North Carolina shifting from McCain to Obama as well as Michigan solidifying Obama's support, I feel that the disclaimer should be removed so that it reflects web sites such as http://www.electoral-vote.com and http://www.fivethirtyeight.com which quickly change a state's status as the polling changes. This would not apply to the maps in this article as long as it indicates the date that the map was last revised. Steelbeard1 ( talk) 17:27, 27 September 2008 (UTC)
The colors of the electoral map are much easier to see now. NS Zakeruga ( talk) 20:57, 29 September 2008 (UTC)
Now that there is a new table reflecting the latest state polling results, I think we can kill the straight list which has become misleading because of the disclaimer forcing states such as Michigan which show solid leads which are forced to show thin leads. Steelbeard1 ( talk) 16:55, 2 October 2008 (UTC)
Steelbeard, I am reverting your latest edit. You did this backwards. You should have brought up the deletion here first. Instead, you are expecting others to justify restoring it. You're the one proposing the change, so you need to justify your position first, especially when we're talking about deletion of material that is properly sourced. Henrymrx ( talk) 18:25, 2 October 2008 (UTC)
The Evil Spartan started the new table as shown by the edit in [6]. If you are reading this, Evil Spartan, can you add your input? Steelbeard1 ( talk) 22:14, 2 October 2008 (UTC)
I've marked this as an organization affiliated with a particular group. Democracy Corps is a group very similar to MoveOn.org; this alone wouldn't be enough, because we've included Research 2000/Daily Kos polls before. However, I am particularly worried about the statement on the GQR website:
Greenberg Quinlan Rosner is a global leader in public opinion research and strategic consulting. Whether you want to win your election, lead your country, increase your bottom line, or change the world, Greenberg Quinlan Rosner can help you find the answer.
This reads to me like several of the other polling groups I've seen that try to find a result friendly to the group paying for the result. The Evil Spartan ( talk) 17:49, 3 October 2008 (UTC)
Why is this Colorado poll marked as partisan?
198.185.18.207 ( talk) 20:27, 3 October 2008 (UTC)
Which version should we use? Any preferences? 90.231.2.252 ( talk) 21:34, 3 October 2008 (UTC)
Personally I prefer the one picked by User:JaGa - but reverted their choice in case people felt strongly about it. 90.231.2.252 ( talk) 21:36, 3 October 2008 (UTC)
Choice A
Choice B
Choice C
I am confused by how we are categorizing states as safe, likely, or leaning. What makes New Hampshire a "safe" Obama state but Wisconsin, a state in which McCain has apparently never led, just "likely"? I have observed a lot of movement lately, not so much of swinging leaning states (which are easier to categorize) but "leaning" to "likely" and "likely" to "safe" in Obama's columns. Minutiaman ( talk) 23:25, 6 October 2008 (UTC)
To whoever suggested that multiple polls 'concluded' on the same day be averaged out, that's a good idea to end the confusion. Steelbeard1 ( talk) 15:09, 7 October 2008 (UTC)
In the article, it is implied that "a poll commissioned by an organization affiliated with a specific candidate" should be marked with "**" and it is stated that such a poll should "not [be] used in calculating results below". At Real Clear Politics, certain pollsters are regularly marked with "(D)" and with "(R)". Assuming this stands for party affiliation, shouldn't such polls be marked here as well and be excluded from calculations in general? On the other hand: This would apply to all polls conducted by "Public Policy Polling" (D) and "Strategic Vision" (R), two polling firms which have relatively good marks in the Pollster Ratings at the "FiveThirtyEight" website. I want to raise this issue, because I just posted a poll by "Public Policy Polling" which shows a relatively large lead for Obama in Virginia. FanofPolling ( talk) 15:38, 9 October 2008 (UTC)
Isn't the colored map in the "Maps of recent polling data" section sort of the definition of original research? Presenting recent polls in this article is fine, but then doing a weighted average of the previous 10, selecting which polls to present in the first place, picking a weighting algorithm, and picking thresholds for different colors are all original analysis. If we want to present a map, I think we need to pick some reliable source and use their rolled up state-by-state analysis, from which we could create our own map (that would probably end up looking very similar to their map). -- Rick Block ( talk) 16:25, 12 October 2008 (UTC)
Upon further thought, I am thinking of moving this to Wikinews, with a prominent link from this page. I would do so myself if I wasn't working 8000 hours per week. Original research is not a problem there, from what I can tell. The Evil Spartan ( talk) 19:52, 17 October 2008 (UTC)
The information accompanying the "recent polling" maps states that North Dakota is in the Leans McCain category while its shade of red on the maps suggests it is in the Safe McCain category. One or the other is not up to date. —Preceding unsigned comment added by 74.178.203.248 ( talk • contribs)
Because of the recent edit warring, I had requested semi-protection for this article. It was declined for reasons given at [7]. If any administrators are editing this article, can you check out the IP addresses carrying out the edit warring and block them if necessary? Steelbeard1 ( talk) 17:01, 22 October 2008 (UTC)
I think the regular contributors here should take note of the "Lead Margin" column in the National polling article. This visual appeal may be better than the comparison of all those numbers and boldings. -- Natural RX 16:29, 23 October 2008 (UTC)
For example, Missouri would look like so:
Poll Source | Date administered | Democrat | % | Republican | % | Lead Margin |
---|---|---|---|---|---|---|
Rasmussen Reports/Fox News | October 26 2008 | Barack Obama | 48% | John McCain | 47% | 1 |
Survey USA | October 25- 26 2008 | Barack Obama | 48% | John McCain | 48% | 0 |
Reuters/Zogby | October 23- 26 2008 | Barack Obama | 48.2% | John McCain | 45.7% | 2.5 |
Mason-Dixon/NBC | October 22- 23 2008 | Barack Obama | 45% | John McCain | 46% | 1 |
Research 2000/Post-Dispatch | October 20- 23 2008 | Barack Obama | 48% | John McCain | 47% | 1 |
Zogby Interactive * | October 17- 20 2008 | Barack Obama | 48% | John McCain | 48.3% | 0.3 |
Fox News/Rasmussen Reports | October 19, 2008 | Barack Obama | 49% | John McCain | 44% | 5 |
Suffolk University | October 17- 19 2008 | Barack Obama | 44% | John McCain | 45% | 1 |
Rasmussen Reports | October 15, 2008 | Barack Obama | 52% | John McCain | 46% | 6 |
CNN/Time Magazine/Opinion Research | October 11- 14 2008 | Barack Obama | 48% | John McCain | 49% | 1 |
Zogby Interactive * | October 9- 13 2008 | Barack Obama | 50% | John McCain | 44% | 6 |
Fox News/Rasmussen Reports | October 12, 2008 | Barack Obama | 50% | John McCain | 47% | 3 |
Public Policy Polling | October 11- 12 2008 | Barack Obama | 48% | John McCain | 46% | 2 |
SurveyUSA | October 10- 12 2008 | Barack Obama | 51% | John McCain | 43% | 8 |
American Research Group | October 4- 6 2008 | Barack Obama | 46% | John McCain | 49% | 3 |
Fox News/Rasmussen Reports | October 5 2008 | Barack Obama | 50% | John McCain | 47% | 3 |
CNN/Time Magazine/ Opinion Research |
September 28- 30 2008 | Barack Obama | 49% | John McCain | 48% | 1 |
Survey USA | September 23- 24 2008 | Barack Obama | 46% | John McCain | 48% | 2 |
St Louis Post-Dispatch/Research 2000 | September 22- 24 2008 | Barack Obama | 47% | John McCain | 48% | 1 |
St Louis Post-Dispatch/Research 2000 | September 15- 18 2008 | Barack Obama | 45% | John McCain | 49% | 4 |
American Research Group | September 11- 15 2008 | Barack Obama | 45% | John McCain | 50% | 5 |
Zogby Interactive * | September 9- 12 2008 | Barack Obama | 42.4% | John McCain | 48.5% | 6.1 |
Rasmussen Reports | September 11, 2008 | Barack Obama | 46% | John McCain | 51% | 5 |
Time/CNN/Opinion Research | September 7- 9 2008 | Barack Obama | 44% | John McCain | 48% | 4 |
Public Policy Polling | August 13- 17 2008 | Barack Obama | 40% | John McCain | 50% | 10 |
Rasmussen Reports | August 7 2008 | Barack Obama | 44% | John McCain | 50% | 6 |
Survey USA | July 29- 31 2008 | Barack Obama | 44% | John McCain | 49% | 5 |
Research 2000/ St. Louis Post Dispatch |
July 7- 10 2008 | Barack Obama | 48% | John McCain | 43% | 5 |
Rasmussen Reports | July 7 2008 | Barack Obama | 42% | John McCain | 47% | 5 |
Public Policy Polling | July 2- 5 2008 | Barack Obama | 44% | John McCain | 47% | 3 |
Thoughts? -- Natural RX 19:20, 29 October 2008 (UTC)
I started with the above Missouri example and did Alabama. I encourage others to include the lead margin for the other states, but I have not yet figured out how to do it in states with minor party candidate fields inserted. Steelbeard1 ( talk) 15:45, 30 October 2008 (UTC)
I think the election results should be placed in this article so that the poll accuracy could be described for each state. See this!
This could be done like this:
Poll Source | Date administered | Democrat | % | Republican | % | Lead Margin |
---|---|---|---|---|---|---|
Election result | November 4 2008 | Barack Obama | 38.8% | John McCain | 60.4% | 21,6
|
Survey USA | October 27- 28 2008 | Barack Obama | 36% | John McCain | 61% | 25
|
Capital Survey Research Center | October 15- 21 2008 | Barack Obama | 34% | John McCain | 54% | 20
|
Survey USA | October 8- 9 2008 | Barack Obama | 35% | John McCain | 62% | 27
|
Capital Research Center | October 6- 7, 2008 | Barack Obama | 35% | John McCain | 54.5% | 19.5
|
Rasmussen Reports | September 22, 2008 | Barack Obama | 39% | John McCain | 60% | 21
|
Survey USA | September 16- 17 2008 | Barack Obama | 34% | John McCain | 64% | 30
|
American Research Group | September 13- 16 2008 | Barack Obama | 36% | John McCain | 58% | 22
|
University of South Alabama/ Press-Register |
September 8- 15 2008 | Barack Obama | 25% | John McCain | 52% | 27
|
Capital Survey Research Center/ Alabama Education Association |
September 3- 9 2008 | Barack Obama | 35% | John McCain | 55% | 20
|
Capital Survey Research Center/ Alabama Education Association |
July 29- August 4 2008 | Barack Obama | 34% | John McCain | 47% | 13
|
Rasmussen Reports | July 31, 2008 | Barack Obama | 38% | John McCain | 58% | 20
|