Talk:Statewide opinion polling for the 2016 Republican Party presidential primaries

Updating the map
For instructions on how to update the SVG map used on this page, please see File talk:Statewide opinion polling for the Republican Party presidential primaries, 2016.svg

Inclusion of Overtime Politics polls
I've got several concerns regarding the inclusion of Overtime Politics, mainly stemming from the issue that its veracity is questionable.


 * 1) For me, the most concerning element is that there's zero coverage of polls published by them in any media sources, which is an immediate red flag; outside of their own site and an obscure political forum, it isn't used by any other reputable sources.
 * 2) A second concern is mainly that it's impossible to verify whether the polls are legitimate at all: they're essentially just Word documents (exported as PDFs) with a copy-pasted screenshot of tables they made, scaled as an image possibly because the person behind the site wasn't able to fit it in otherwise. While they do now publish breakdowns of gender/income/age, this doesn't make any significant difference.
 * 3) It's literally a WordPress site. (Same with ToS; generic WP ToS.)
 * 4) Unprofessionalism: layman's writing and analysis, unclear origins of the site, and the identity of the individual(s?) behind the site is unknown.

Bottom line is that I could buy a domain name, set up a site and do the same – and I don't see any persuasive reason to believe that it is a legitimate pollster. 108.2.58.56 (talk) 22:12, 6 January 2016 (UTC)


 * It just seems like they are new, they have crosstabs, and they provide information for states which do not get polled very often. The breakdowns absolutely make a difference as well. All pollsters start somewhere, I would say see how well they match up with other polls in the near future. — Preceding unsigned comment added by 173.26.72.74 (talk • contribs) 01:45, 7 January 2016 (UTC)


 * That doesn't at all address my concerns about the veracity of their polls. 108.2.58.56 (talk) 02:14, 7 January 2016 (UTC)


 * I understand your concerns regarding the methodology and veracity of the said pollster. That being said they are a new pollster and deserves a bit of leeway.  They have said they conduct most of their polling through the phone and given they are new they will need time to be quoted by reputable news websites.  They have started using breakdowns of individuals being polled which is a good sign they are making improvements.  They are making an effort to provide information in numerous states, for both parties.  I would consider them suspect if they just concentrate on one party.  Give them time and see how do in future polls.DrFargi (talk) 02:33, 7 January 2016 (UTC)
 * All I could find online was that someone said that they were three or four people working out of someone's house. We don't know ANYTHING about them. Light-jet pilot (talk) 03:37, 7 January 2016 (UTC)
 * WP:CRYSTAL: We're not talking about what they might become; we're not speculating on their eventual fate, but I think it's sufficient to say that most media outlets would be hesitant to cite their polling for their lack of transparency; and it's for this same reason that I don't think that it should be used in the article. 108.2.58.56 (talk) 04:00, 7 January 2016 (UTC)


 * 108, why have you excluded Dave Leip's Atlas of U.S. Presidential Elections from your search results? Reuters reports that that site "has been used a reference for U.S. election and political data by major media outlets", so excluding the Atlas from your above searches doesn't seem to make sense. It seems as though you're writing off the Atlas of US Presidential Elections as an "obscure political forum", but since big media outlets apparently use the Atlas as a reference, why would we disinclude coverage of Overtime Politics from a site considered reputable, and which apparently has been around since 1992?
 * I also visited OvertimePolitics.com, and did not find the Word documents you are talking about; the poll results appear to be published in PDF format, as seems to be standard in poll reporting. 65.129.137.230 (talk) 05:20, 7 January 2016 (UTC)
 * Fluff (just forum discussions), and doesn't constitute any sort of useful "media" coverage; excluding them and the overtimepolitics.com site itself, there appears to have been no usage of the polls outside of there (and outside of Wikipedia!). (Essentially all "coverage" of the site comes from OP themselves, and it's this lack of coverage or willingness to utilize their polls that makes them seem particularly suspicious. (It isn't like they, and others, haven't made an effort to publicize and distribute the results of their polls). 108.2.58.56 (talk) 05:36, 7 January 2016 (UTC)
 * Thanks for the response. Yes, now that you mention it, I do see quite a few forum discussions about OvertimePolitics.com polls on Daily Kos, Twitter, Reddit, etc. It looks as though a broader search term pulls up a little more coverage. And there is also extensive discussion regarding this pollster on DemocraticUnderground.com, which apparently featured their Arkansas poll on their main page. Looking at the DU forums, it seems there's some movement on those forums to discredit OvertimePolitics polls, with some comments like "Landlines? Cell Phones? Two cans connected by a string?". Let's move cautiously here. Who are these "others" you mentioned who are distributing their poll results, and why would actual distribution of poll results be in any way a problem? Is there any reliable source which states their poll methodology is wrong? Also, can you address what you were talking about with the Word documents? As I said above, I didn't find any Word documents. 65.129.137.230 (talk) 06:35, 7 January 2016 (UTC)


 * Can someone clarify the statement that they are 3 or 4 people working out of someone's home? Otherwise it's just someone belittling them or WP:NPOV.  By the way I just happen to find a news portal website that mentioned the Vermont poll that OvertimePolitics.com created.[Link http://www.boston.com/news/local/vermont/2016/01/06/what-donald-trump-doing-burlington-vermont/yEGGsLueEbvTLmIfIRKBcM/story.html]DrFargi (talk) 09:31, 7 January 2016 (UTC)

Significant criticism of Overtime Politics methodology
The discussion on Democratic Underground has gone beyond expressions of suspicion. One of the members who posts frequently about poll methodology looked over what Overtime Politics had posted about the Democratic primaries, found some incongruities, then went back to the Overtime Politics site and found that the numbers had simply been changed to address the problems. The changing of the numbers is itself very suspicious. He further found that one of these changes introduced a whopping new error. The demographic breakdown for Native Americans was Clinton 49%, Sanders 49%, O'Malley 0%, and Unknown 12%, adding up to 110%. Results don't always add exactly to 100%, because of rounding, but it's simply not possible for results for four categories to add to 110%. The DU poster commented: "Oh, and the typo in Native American support equaling out to 110%? That's really damning as well. That means someone was manually inputting values, as opposed to them being formulaic from a respondents tabulation. And someone apparently forgot to carry the one." The thread on Democratic Underground is here, with most of the information in the original post but some elaborations further downthread.

A pseudonymous poster on a message board isn't a reliable source, but the assertions presented in the post are based on what Overtime Politics itself has reported. One doesn't need a RS to confirm that the results in the Democratic race can't possibly add to 110%.

User 65.129.137.230 asks above, "Is there any reliable source which states their poll methodology is wrong?" That's the wrong question. If someone just starts putting out bogus poll reports, it might be that no RS finds these antics important enough to debunk. The right question is whether there's a reliable source that verifies their poll methodology -- or, as a substitute, is it a well-established pollster such that any problems can be presumed to have come to light already. If I call myself Lane Public Opinion Research and post a bunch of made-up numbers on my website, Wikipedia should not report those numbers merely because no RS has bothered to point out that I'm full of it. Our presumption should be for exclusion not inclusion.

On the current state of the evidence, all Overtime Politics results should be removed from Wikipedia. JamesMLane t c 22:06, 7 January 2016 (UTC)


 * Someone on the Atlas said that they received a response from the poll creator who said it was an error and it was corrected. It still seems as though there is no evidence that proves wrongdoing. If they were just releasing polls for the first 4 states, that might be odd, but they are releasing polls for states that vote in March and beyond. Some of these states have not been polled for months. How do we know that the numbers aren't correct? — Preceding unsigned comment added by 173.26.72.74 (talk • contribs) 17:50, 7 January 2016 (UTC)

Strong Agree (should not use) - The burden of proof is on them, not us. We don't need 'evidence that proves wrongdoing' to exclude them, we need evidence that they are legitimate to include them. Read this wiki policy, it tells why. — Preceding unsigned comment added by Light-jet pilot (talk • contribs) 00:23, 8 January 2016 (UTC)


 * If a news portal quotes their poll, what does it say about us and there was one I found recently. That we cut them off because of posts from a forum from the Democratic Underground?  We are neutral here at Wikipedia.DrFargi (talk) 05:36, 8 January 2016 (UTC)


 * Overtime Politics has been mentioned in at least one reliable source i.e. the one poll done in Vermont so we should at least include the Vermont poll. Prcc27 (talk) 06:00, 8 January 2016 (UTC)

Disagree (should continue using OP polls, unless there's real evidence of deception). There have been some strange allegations made above about Overtime Politics, including bizarre "facts" like: they publish their polls in Word format (the polls are actually in standard PDF format, which is what every pollster uses), supposed "zero coverage" in media (DrFargi astutely debunked that with the Boston.com article, but there are others as well), and then some crazy stuff about it being somehow suspicious that OvertimePolitics is "making an effort to publicize and distribute the results of their polls"... which is exactly what polling companies do. What pollster doesn't distribute results? Then the above argument is made that there are problems with the methodology that Overtime uses because "one of the [Democratic Underground] members who posts frequently about poll methodology [...] found some incongruities." For me, this boils down to, "Someone on a forum said". I'm not familiar with who the regular forum-goers are on Democratic Underground, but I'd like to avoid forum groupthink if possible. It also seems clear that some of the allegations on that forum, as well as the comments here, are just kind of silly: the fact that the company uses Wordpress (OMG!) was actually given as a reason to remove the polls, even though Time and other major sites use Wordpress. 65.129.137.230 (talk) 07:51, 8 January 2016 (UTC)
 * Good point. Prcc27 (talk) 08:59, 8 January 2016 (UTC)


 * You can't so blithely dismiss a substantive concern just because it appears in a forum. I gave one notable example, of numbers adding to 110%.  That's strongly supportive of the hypothesis that someone just sat down and made up numbers (and did so carelessly).  Furthermore, that there was one citation in a RS doesn't clinch the argument.  If someone makes up numbers (not polling anyone) and puts out purported results, it's not at all inconceivable that most RS's most of the time take a look, see no indication of reliability, and ignore the "information", but that on one occasion a reporter is a bit too hasty and runs with it.  See Journalistic scandal for a long list of much worse blunders.  (For example Judith Miller, whose coverage in The New York Times "was discovered to have been based on lies....")


 * I still see nothing to distinguish this from a totally made-up operation. If I were to carry out my threat to establish Lane Public Opinion Research and start posting totally fabricated poll results, on what basis would my lies be excluded from Wikipedia?  Especially if, even once, some mainstream publication got careless and published my spurious numbers? JamesMLane t c 06:01, 9 January 2016 (UTC)
 * James, please be careful. You've made no mention of the fact here on Wikipedia that you have solicited forum-goers from the Democratic Underground forum to participate in this discussion. These are not Wikipedia editors, but rather forum users who have formed an opinion based on a forum discussion, and their input here would not constitute a consensus, which is what we're trying to hammer out here. So that invitation does not seem entirely appropriate. You are close to violating at least the spirit of WP:CANVASS: Votestacking is an attempt to sway consensus by selectively notifying editors who have or are thought to have a predetermined point of view or opinion, and thus encouraging them to participate in the discussion.
 * Further, it's clear now, with more than 7,000 posts according to your Democratic Underground profile, that you are a regular on the DU forums, and thus are probably not a great judge of what is and what is not forum groupthink: you have said as much yourself: you value their opinions. But their opinions are not evidence of a scam. I truly appreciate your efforts to fix a situation you believe is a scam, but it's not at all clear to me, and to some others here, that there really is a scam, and the forum messages I have seen on DU have been mostly immature and petty: as someone stated above, it looks like belittling comments like, "are they using a string and two cans?", which rather than convince me of wrongdoing by Overtime Politics, make me instead believe these are mostly young forum posters who don't really understand that polling is an inexact science; I think we all recall that in 2008, Obama won Iowa while supposedly being 10 points behind Clinton in the polls. There's so much discussion in that thread about which polls are "believable" and which are "not believable", when that sort of discussion misses the point about what polls actually are.
 * I hope I won't see a bunch of new editors come in to this conversation and begin "voting" to remove Overtime Politics polls from these articles; it would be a shame to have that happen in what was previously a rational discussion amongst actual Wikipedia editors. I have placed a WP:CANVASS template above, in the hopes of reducing the potential impact from those types of comments. I wish you hadn't done that. 65.129.137.230 (talk) 07:32, 9 January 2016 (UTC)


 * 1) "Canvassing" - This is not an RfC or anything with votes or even !votes. We're discussing a purported pollster about which, everyone must admit, there is very little information available.  Somebody has taken the trouble to examine the published "results" in detail, thereby generating some information that Wikipedians might find useful, so I invited that person to come here and comment.  This is not canvassing.  If s/he chooses to show up, it will quite obviously be a new account, and everyone can take the comments for whatever they're worth on their merits.  I think your notice is superfluous.  Furthermore, given that the DU poster so far hasn't joined in, the whole question may well be moot anyway.  The notice does no harm except to look silly, and on Wikipedia that wouldn't exactly be a first.  (Also, I can't help noting that you're being a bit supercilious to want the discussion restricted to "actual Wikipedia editors" when you yourself have been here less than a month.)


 * 2) "Groupthink" - This is a straw man. I have not made the argument that Overtime Politics should be excluded because most people on DU are suspicious of it.  I made the argument that, for example, the presentation of results totaling 110% was evidence that the purported poll was bogus.  (The 110% is not a false assertion by a DU member, nor is it a product of groupthink.  You can go see it for yourself because as of this writing it's still posted on the Overtime Politics website.)  No such result could have been obtained from any genuine poll.  That theory is also not the product of groupthink; it is the product of mathematics.  As Neil deGrasse Tyson said, "The good thing about science is that it's true whether or not you believe in it," and I'm sure he'd agree with including mathematics.


 * 3) Wikipedia standards - What this really comes down to is the difference between two approaches. You ask, "Is there any reliable source which states their poll methodology is wrong?"  I ask, "Is there any reliable source which states their poll methodology is right, or at least reasonable? or that they actually conducted a poll at all? or is the pollster sufficiently well established that we can reasonably presume that any pervasive fraud would have been unearthed by now?"  If we have a complete absence of reliable information (or, as here, a nearly complete absence), the presumption should be to exclude.  As we see in the policy on self-published sources, "Anyone can create a personal web page or publish their own book, and also claim to be an expert in a certain field."  That's essentially what we have here. JamesMLane t c 08:50, 9 January 2016 (UTC)
 * WP:CANVAS does not only apply to Requests for Comments, as you must know. We are on this talk page to come to a consensus, but as the guideline states, canvassing "compromises the normal consensus decision-making process." That DU thread is in my opinion a great example of forum groupthink, and I think you'd recognize the signs more easily if you were not a member of that community. Indeed, your second sentence states that an accusation was leveled against Overtime Politics by "one of the [DU] members who posts frequently", as if the fact that he posts frequently on the Democratic Underground forum should bear weight on Wikipedia, and as if the fact that an anonymous person on a forum said something about a company should somehow carry any weight here. It seems to me that a typo or small error on a single PDF has been blown out of proportion by an already politically-charged forum, and you, as a member of that forum, have bought into that paranoia. Overtime Politics is obviously a new pollster, but nothing stated above indicates fraud, deception, etc., to me. And the "evidence" for dishonesty, which you listed above (the company used Word docs, the company used Wordpress, the somehow sinister distribution of polls, "zero media reporting", "string and cans", etc.), is either lacking, turned out to be wrong, or is somewhat silly. As for me, I've been editing Wikipedia regularly since 2003, though my IP address changes every couple of months. 65.129.137.230 (talk) 12:58, 9 January 2016 (UTC)
 * No one has come here to comment because of the DU thread. 108.2.58.56 (talk) 21:22, 9 January 2016 (UTC)

That tangent aside, what I said at the top still stands and I'll say it again. The entire argument is moot while Overtime Politics remains an unreliable source. The policy on self-published sources and the policy about questionable sources both apply to Overtime Politics and to Democratic Underground. One reference in one reliable website is not near enough to change that. Light-jet pilot (talk) 07:34, 10 January 2016 (UTC)
 * But LJP, it's not just one source which is using Overtime Politics' polls, despite James' original claim, above, that zero media sites were using their poll data. Besides the Boston.com articles, Overtime Politics polls have been reported on Daily Kos, Conservative Angle, Wopular.com, Daily Pundit, TheNewsCommentor.com, and other media sites as well. There's nothing to the claim that "zero" or even "one website" is relying on, and reporting on, polls from Overtime Politics. 65.129.137.230 (talk) 12:32, 10 January 2016 (UTC)
 * I think it's questionable if any of those sources are indeed 'reliable sources' aside from the Globe-affiliated site; the Daily Kos link you provided is just someone's random forum post, and in general, if there isn't a (Wikipedia) article on a news outlet, it probably should not considered 'reliable'. 108.2.58.56 (talk) 14:59, 10 January 2016 (UTC)

Theory Has anyone considered the considerable possibility that Overtime politics exists merely to promote a single candidate? If you look at their polls, all the Republican polls start with the headline "Trump leads". Now you might dismiss this and say that he is currently leading everywhere. BUT, if you look at the Iowa poll December 8-12, it is the only recent poll from that state to have Trump comfortably ahead. Ted Cruz is within a few points or comfortably ahead in all the recent polling except OP. Then, look at the poll they did in South Carolina. While the polls conducted on either side of it have Trump leading by 7.2 and 8 points respectively, Overtime's "poll" has him leading by 20 points! Light-jet pilot (talk) 19:28, 10 January 2016 (UTC)
 * LJP, your leaning towards WP:OR, in assuming that the poll favours Trump. I hope that is not the case.DrFargi (talk) 03:30, 11 January 2016 (UTC)
 * I said it was a theory. WP:OR only applies if I use it in article space, not a talk page. It looks like your point of view is causing you to attack other editors with the proverbial "rule book" just because you disagree. I hope this is not the case. Light-jet pilot (talk) 03:43, 11 January 2016 (UTC)
 * LJP, I am not attacking you in anyway. Just stating the obvious that your theorizing that the polls favour Trump.  If you say the polls are favouring Trump, then what about the polls favouring Hillary?

This is a two way street here.DrFargi (talk) 09:32, 11 January 2016 (UTC)

June 2016 update: The overtimepolitics.com URL now redirects to bihira.com, a website hosting company. Whatever the "Overtime Politics" scam was, it's evidently over. JamesMLane t c 04:51, 10 June 2016 (UTC)
 * Hate to say I told you so, but...... Light-jet pilot (talk) 14:37, 11 June 2016 (UTC)

RfC on inclusion of Overtime Politics polls
Should Overtime Politics polls be included in statewide opinion polling articles? 108.2.58.56 (talk) 15:17, 10 January 2016 (UTC)


 * COMMENT - The discussion so far has focused on whether the poll is reliable or not... but I don't think that is the right question... what I think we should be asking is whether mentioning it gives it WP:UNDUE weight. The Overtime Politics poll is apparently a new and still fairly obscure poll - rarely (if ever) discussed by notable political commentators or by the broader media. As such, I think there is a legitimate question as to whether it has enough of a reputation (good or bad) to merit being mentioned.  Yes, it is verifiable that the poll says what it says... but simply being verifiable does not guarantee inclusion. Blueboar (talk) 15:54, 10 January 2016 (UTC)


 * Yes, include. I was initially persuaded by Blueboar's comments above that including a pollster which has not been discussed by notable political commentators or in the broader media may be giving said pollster undue weight. However, today on his talk show, Rush Limbaugh discussed Overtime Politics, and their polls showing Trump ahead in Florida, Virginia, and Michigan. The transcript is available on Limbaugh's website. This discussion came about after The Daily Caller included a robust article reporting on Overtime Politics' last few polls. Michigan radio station WHBF, too, reported on Overtime Politics' latest poll numbers on the 11th. These reports, along with the aforementioned Boston.com article, OP polls being included on DemocraticUnderground's front page, and the discussion on quite a few smaller media sites indicate to me that inclusion of Overtime Politics' polls on Wikipedia, alongside the more established pollsters' results, would be acceptable. Edited to add: Overtime Politics has also been cited on major Republican presidential candidate John Kasich's website repeatedly  and in official GOP party websites in Ohio. The pollster's polls have also been reported by the Northwest Florida Daily News, The Daily Progress and the The News Virginian, the Bangor Daily News , as well as being reported by progressive website Truthout , The University Chronicle (a paper founded in 1924), The Toledo Blade, The Weekly Standard (News Corp)  and is also cited in aggregate polls from HowWillAmericaVote.com. Since the mainstream media, official GOP campaign websites, and some poll aggregators are all now reporting on and including Overtime Politics polls in poll discussions, Wikipedia likely should include OP's polls as well. It would be strange to omit a pollster that has been included in the press in both left-leaning and right-leaning publications (from Rush Limbaugh to Democratic Underground), and which is being used on official campaign websites. WP:Undue doesn't really apply here, when any more recent statewide poll would knock out Overtime Politics' results on our articles. But since it looks like Overtime Politics is the only pollster doing polling in several states, some of their polls are potentially quite valuable. 65.129.137.230 (talk)/67.1.128.249 06:04, 13 January 2016 (UTC)


 * No I would want to see it referenced in secondary sources, particularly those that aggregate or analyze polls. Rush Limbaugh trying to make a political point does not lend OP any credence as a legitimate polling organization, but established outlets that work with these such as 538, RCP, or Daily Kos Elections would. Reywas92 Talk 09:52, 16 January 2016 (UTC)


 * No, should be omitted. Anyone can create a website and announce "results" from a poll.  If you put out enough press releases about your supposed poll, you'll get a few mentions in blogs and the like, and even a few mentions in mainstream media, but there's not enough here to convince me that this outfit has been vetted. JamesMLane t c 16:01, 21 January 2016 (UTC)


 * Yes, include. Bloomberg Politics used their information, Donald Trump used their information, John Kasich used their information. They seem to have a solid methodology down and they are putting out polls for a number of lightly polled states, giving us valuable information. FandangoFan (talk) 20:20, 22 January 2016 (UTC) — FandangoFan (talk • contribs) has made few or no other edits outside this topic.


 * No (note: initiator of RfC). Recent polls conducted around the time that Overtime Politics has conducted their own shows that their results are too often absurd anomalies, and that they don't observe the same practices as other legitimate pollsters do. Methodology does not check out; there's nothing to show that they simply aren't making numbers up, as they seem to be a heavily contrarian pollster. That media sources utilize them does not necessarily legitimize them; they aren't used by any polling averages, for one, essentially every other pollster used on the polling articles are. Here's a blog from the Daily Kos noting some further problems which have arisen. The guy who runs the site seems to know zilch about both politics and statistics: I'm convinced they're just making numbers up, and doing a rather sloppy job of it. (Essentially just rehashing similar comments above.) 108.2.58.56 (talk) 18:08, 25 January 2016 (UTC)
 * As it turns out, Overtime Politics was the closest pollster in Iowa, and their poll was the most accurate, with the Selzer being second-closest. The allegation that they are "just making numbers up" doesn't hold up under the results. 67.1.128.249 (talk) 08:01, 4 February 2016 (UTC)
 * You're wrong on that front – it was the Emerson poll just before the caucuses was the closest for the Republican race and Overtime Politics was far off. Just because one poll was close to the actual results doesn't convince me of their veracity. It's implausible that polls released prior to criticisms expressed on DU showed results adding to 100 in every column – and not so immediately thereafter. 108.2.58.56 (talk) 17:51, 4 February 2016 (UTC)
 * It's not me that's saying it: it's the same group of bloggers on DailyKos that you linked to immediately above. Overtime Politics had the most accurate poll out of all pollsters for the Iowa Democratic Caucus, more accurate than Selzer, who has been considered the "gold standard" of pollsters in Iowa. OP's Iowa Republican poll was about as accurate as any of the major pollsters. Yes, let's remove the pollster with the most accurate results. Are you actually still arguing to remove them after their Iowa Democratic poll turned out to be more accurate than any other poll, even the gold standard, and their Iowa Republican poll was about as accurate as the rest? Above, you've attacked their methodology ("Methodology does not check out; there's nothing to show that they simply aren't making numbers up, as they seem to be a heavily contrarian pollster"), but whatever they're doing seems to be working. 67.1.128.249 (talk) 07:49, 5 February 2016 (UTC)
 * Alternatively, they're just lucky. 108.2.58.56 (talk) 21:50, 5 February 2016 (UTC)


 * Yes. Let's put this discussion to bed. Regarding the use of polling averages, we don't even post the state polling averages in this article.  The only time we post and update the polling averages are the national figures for both parties- Nationwide opinion polling for the Republican Party 2016 presidential primaries and Nationwide opinion polling for the Democratic Party 2016 presidential primaries- HuffPost, RealClear and 270 to Win.  Overtime Politics has only been posting state polls all this while, so that mutes that argument on polling averages not quoting Overtime.  Now there are many news and analyst websites quoting Overtime Politics' polls and utilizing their numbers for their own readers. That ends that other argument there. Whether it legitimizes them or not is not up for us to interpret. Finally regarding problems with their methodology, if a blog website like the Daily Kos has an issue with their statistics do we take the blog's words as a reliable source.DrFargi (talk) 08:33, 26 January 2016 (UTC)
 * I think you're misunderstanding what I'm arguing. The polling averages are also offered for individual states. I'm not saying that we should include the polling averages in the article, though. What I'm taking issue with is OP being a reliable source. 108.2.58.56 (talk) 15:52, 26 January 2016 (UTC)


 * Exclude unless specifically referenced by reliable sources. The only top tier reliable source that references OP is the Boston Globe affiliate. As such OP's reliability has not been established. However, specific polls referenced by reliable sources can be included.  Rami  R  10:07, 26 January 2016 (UTC)


 * Note: I've requested closure of this RfC, as it's gone past a month now. 108.2.58.56 (talk) 18:51, 16 February 2016 (UTC)


 * Until Overtime Politics is mentioned in a significant amount of reliable sources like aggregates and what not, I say we exclude it. BTW, Prcc27💋 (talk) 08:40, 17 February 2016 (UTC)

Chronological
Doesn't it make more sense for the list to be in chronological order? I propose the polls are. They are for past election pages, and it makes sense for it to be here. Primaries are not all held on the same day. Makes sense to put them in order. Easier to read and reference. Manful0103 (talk) 14:08, 1 May 2016 (UTC)


 * Can we please have this discussion in one place?  Rami  R  14:14, 1 May 2016 (UTC)


 * That's a different page, so discussion is here. We can look to other pages for guidance, but each page is its own page and has it's own discussion.  Besides that, please give your reason for not having this polls in chronological order.  As I said, in the past presidential polling pages they are in chronological order, not alphabetical.  Yes, every page is different, but the set up in those can serve as a model and it looks better and reads better.  People look for polls in order of time, not name.  Please give a reason for us not to set up this page as it has for the last few presidential primary pages.  Manful0103 (talk) 14:20, 1 May 2016 (UTC)


 * I gave my reasons there. The fact that it's a separate article is immaterial: it's the same election process, and will have the exact same arguments. We should continue this discussion where it started.  Rami  R  14:58, 1 May 2016 (UTC)

Page size
This article has 427,057 bytes of markup - it is far too big. What is the best way to split it? Andy Mabbett ( Pigsonthewing ); Talk to Andy; Andy's edits 11:55, 14 September 2019 (UTC)