Talk:Opinion polling for the 2017 German federal election

CDU or Union?
I think the numbers shown here for the CDU are the combined numbers of the Union (CDU+CSU). This should be fixed in the table header. 217.82.131.99 (talk) 19:30, 21 March 2017 (UTC)

Preferred Chancellor Polls
So I've been seeing some head-to-head polls between Merkel and Schulz, and I think they would be a good addition to the page. Yes, I understand that German voters don't vote directly for Chancellor and thus the polls are hypothetical - but the movement between these polls still say a lot about the parties' relative positions. The polling page for the next Canadian election has this too. Chuborno (talk) 02:24, 8 February 2017 (UTC)
 * Yeah, this is definitely a good idea. I'll see what I can find. Mélencron (talk) 02:33, 8 February 2017 (UTC)
 * Doesn't seem that they're easily collated anywhere, so it might take rather a bit of effort to find all the individual data points. Mélencron (talk) 02:50, 8 February 2017 (UTC)

Splitting the table
I can't tell what you're referring to when you're saying it's a "standard practice all across Wikipedia" to split up polling tables – check out this category and see how many split tables you can find. Not many, aside from the French election articles. This is standard practice. Weren't you going on earlier about how we shouldn't care about standard practice? What happened to that? Mélencron (talk) 03:57, 11 February 2017 (UTC)
 * I don't recall ever saying we shouldn't care about standard practice, but my preference is always for readability. A giant wall of tables that goes back to 2013 is simply not readable. The tables are still sortable, the graphic tables easily show trends going back to January 2014, so literally nothing is sacrificed by splitting it up. But it's a lot easier to read. These pages are not your fiefdoms.  Nevermore27  (talk) 04:03, 11 February 2017 (UTC)
 * As for examples, Germany appears to experience a volume of polling quite unlike any other European country other than the United Kingdom. If you see here and here, you will see other examples of splitting up tables by year to avoid the "wall of text" problems. If you want to split it up in the style of those pages instead of the French style I tried to emulate, that is perfectly fine. But I do not think a vast impenetrable table should be the preferred method.  Nevermore27  (talk) 04:09, 11 February 2017 (UTC)
 * A long table is no less readable than a short one. How is this and more "readable" than this? Riddle me that. I'll revert any edits I view as inane, and this is one of them. The tables are going to be long. Collapsing individual tables simply makes it harder to navigate – I wouldn't want to uncollapse tables individually. This doesn't make it any more convenient – if anything, it's significantly less convenient; if I'd want to find the lowest/highest figures for a party, I'd now have to check though 8 different tables. That, in my view, isn't desirable. Mélencron (talk) 04:09, 11 February 2017 (UTC)
 * "Perceived inanity" is not a valid reason to revert a good faith edit, which my edits clearly are. If you have a problem with that, consult with an administrator and see how far that gets you.  Nevermore27  (talk) 04:13, 11 February 2017 (UTC)
 * I'll revert any changes that have a detrimental impact to accessibility, tyvm. How about you ask an admin if you're not allowed to revert a "good faith edit"? If that were the case, then the entire site would be pretty terrible, since there are a whole lot of edits made in good faith which cause problems in themselves – e.g., breaking references, adding unreferenced or unverifiable information, details irrelevant to the article, the return of trivia sections, etc. There's absolutely no policy which dictates that one isn't allowed to revert "good faith edits" – stop making up excuses. Mélencron (talk) 04:16, 11 February 2017 (UTC)
 * To quote from the user who implemented the year-by-year section breakdown in the 2013 German polling article: "There is no benefit to be derived from dividing by year, and the only reason I did it was because some years were top posted, while others were bottom posted... Adding section headings and heading rows would be more disruptive for them than helpful." Mélencron (talk) 04:25, 11 February 2017 (UTC)
 * P.S. I'm fine with adding grey year rows to denote new years, by the way. Not difficult to implement. Mélencron (talk) 04:26, 11 February 2017 (UTC)

Would the current page be an adequate compromise? I cannot stress enough how opposed I am to keeping it in a single table.  Nevermore27  (talk) 04:36, 11 February 2017 (UTC)

Mélencron (talk) 04:41, 11 February 2017 (UTC)
 * So I take it that's a no.  Nevermore27  (talk) 04:43, 11 February 2017 (UTC)
 * I feel like the table speaks for itself. Mélencron (talk) 04:44, 11 February 2017 (UTC)
 * Would adding year markers – which, in my view, is effectively the same as the current solution, aside from the fact that it then also retains the ability to sort the table – be an agreeable solution? Mélencron (talk) 04:52, 11 February 2017 (UTC)
 * I think you greatly overestimate the usefulness of a sortable table with over 500 polls.  Nevermore27  (talk) 04:55, 11 February 2017 (UTC)
 * Have you ever observed the CDU/CSU's levels in INSA polls? Head of the institute has links with AfD. There's a use to it, I assure, and you might as well make each year unsortable if you think it's not useful at all. Mélencron (talk) 04:58, 11 February 2017 (UTC)
 * I never said it wasn't useful at all, I just think the returns diminish rapidly when you have a table that large.  Nevermore27  (talk) 05:00, 11 February 2017 (UTC)
 * Poll economies of scale, . House effects become more apparent with more polls, not fewer. There's only one effective difference between our two solutions – and that's I value the ability to sort among all polls. I really don't see how the ability to sort only among a single year's polls is better; given the relative stability of polls up until September 2015, it isn't much use to merely be able to sort by year – much more value in the ability to sort an entire table, and I really don't see why that's more an issue than with split tables. Mélencron (talk) 05:06, 11 February 2017 (UTC)
 * Your point about house effects applies to the UK and the US too, the only other countries with a comparable number of polls. They both split up polls by year. Why is it right for them and wrong for Germany?  Nevermore27  (talk) 05:15, 11 February 2017 (UTC)
 * Your definition of "many" is suspect, because with only one or two exceptions, the number of polls for Germany are easily double, triple or even quadruple that of all the pages you cite. And for the exceptions, I believe splitting them up would be the right way to go.  Nevermore27  (talk) 04:47, 11 February 2017 (UTC)
 * User:Mélencron, I agree that splitting them into separate tables brings no apparent benefit, but the whole table is long. If we added the year markers which you proposed, which are rather elegant, is it then possible to collapse older years (within the same table)? Jdcooper (talk) 07:54, 11 February 2017 (UTC)
 * Comment Nevermore27 I don't have an special opposition to splitting such an enormous table, but I'd like to point out several facts. Firstly, you come here using the very same argument I used against you (and that you yourself dismissed) back in Talk:Opinion polling for the French presidential election, 2017, that it's "standard practice all across Wikipedia" to split tables. Aside from the fact that you've been proven otherwise, if you dismissed this argument back then it just doesn't seem reasonable for you to pretend to use it here and go away unmolested. If you found opposition to your edits (which you did), you should've immediately come here to the talk to discuss the issue instead of keeping pressing your edits forward. "Readability problems" is something very subjective, and you should understand that not everyone may have the same issues. If you find opposition to your edits, discuss them instead of engaging in edit warring yourself. Everyone has the authority to revert good faith edits if they don't agree with them, so don't act as if your edits had more value than others' and should, somehow, be kept from any revertion. I myself could have easily supported such a change, and not much for "readability" (sincerely, you seem to be having readability issues everywhere as of lately) but for practical reasons (it's much easier to edit smaller tables than to make edits in a single, enourmous table). Cheers. Impru20 (talk) 10:09, 11 February 2017 (UTC)
 * In my experience (chiefly with American polling articles), it was standard practice to split up large tables. I was mistaken. I apologize. I agree it's more an argument of practicality than readability.  Nevermore27  (talk) 23:07, 11 February 2017 (UTC)

Fieldwork date
The article lists the opinion polls according to a "fieldwork date" and not according to publication date. This problematic given that (most) opinion polls are not conducted during a single day. Either we should change for publication dates or we need to instead use something like fieldwork period. The current way of presenting the data is misleading. --Glentamara (talk) 19:50, 19 February 2017 (UTC)
 * The fieldwork date shown is the last day of pulling. It's modelled on the Canadian opinion polling articles, though as of recently I've personally grown in favour of showing the entire fieldwork period and would support it.
 * The only issue is finding the the brave one that will apply the change to these enormous tables, yeah. xD Impru20 (talk) 20:05, 19 February 2017 (UTC)
 * But then it should be "Last fieldwork date", not "Fieldwork date", because there is nothing like a fieldwork date for an opinion poll. --Glentamara (talk) 20:09, 19 February 2017 (UTC)
 * Would using "date" instead and just listing an end date be amenable? I see your point (and agree), but I honestly think that I'd be too lazy to figure out how to fix this for 750 polls. Mélencron (talk) 00:11, 20 February 2017 (UTC)
 * Now it looks better. Thanks. --Glentamara (talk) 06:09, 21 February 2017 (UTC)

Use of unspaced en dashes between two dates with at least one space
MOS:DASH specifies that two dates with at least one space in them should be separated by a spaced, not unspaced, en dash. However, opinion polling articles have often not complied by MOS:DASH in the past – whether intentionally or simply because users aren't aware (i.e., that hyphens or em dashes shouldn't be used to separate dates). I also don't believe there isn't a good reason not to comply by this policy on opinion polling articles, especially as it doesn't cause any particular disruption (only in a few cases do columns have to be adjusted, say, 5px at most).


 * 1) Should we comply by MOS:DASH (i.e., use spaced en dashes) between two dates with at least one space?
 * 2) Should we not use spaced en dashes between two dates with at least one space?

Impru and I have already articulated our positions on this issue – I'd be more interested in hearing the viewpoints of others rather than us two continuing to litigate this between ourselves. Mélencron (talk) 14:54, 25 February 2017 (UTC)
 * Comment It's nice that finally, after a heated conflict in the middle of the night and an (unsuccessful) attempt to report me on the grounds of a rule that wasn't even broken, you actually come to discuss the issue. Congratulations. Several points (I'll not be litigating on this&mdash;which I'm not necessarily opposed to&mdash;but rather giving some general advices):
 * 1. I don't have any special opposition to strictly comply with MOS:DASH if there's agreement for it. However, in such a case, two conditions should be met:
 * 1a. That it's applied to all opinion polling articles for consistency reasons (this would require extensive editing for some).
 * 1b. That it's noted somewhere that it should be done this way, because if your statement that "this is done like this because people don't know it" is true, then some kind of guide should be introduced for newcomers to adhere to the new formatting and not going all the way backwards towards non-spacing dates.
 * 2. I commend your attempt at finally opening a serious discussion on the issue, but I think this would be more appropiate here, since this is an issue that affects a wide range of articles and not just this one. Furthermore, there you'll get a wider consensus since more people do read that page.
 * 3. I also think that if you wish to post links to other articles, you should make them distinguishable so that everyone may know what those actually link to and/or the purpose for their addition. This is:
 * 3a: This: Opinion polling for the next Turkish general election, aren't, aware, Opinion polling for the next Slovenian parliamentary election, Opinion polling for the Norwegian parliamentary election, 2017, Opinion polling for the Czech legislative election, 2017 Opinion polling for the Czech presidential election, 2018, Opinion polling for the next Croatian parliamentary election, Opinion polling for the Czech legislative election, 2017
 * 3b: Or something like this: ''(...) a series of opinion polling articles you may check here, here, here, here, here, here, here, here, here, etc.
 * 3c: But not this: "users aren't aware (i.e., that hyphens or em dashes shouldn't be used to separate dates)", which brings no reason or purpose as to why those links are placed that way and may bring confusion.
 * Nothing else. Cheers. Impru20 (talk) 15:56, 25 February 2017 (UTC)
 * I need to leave in a couple minutes – mind if you crosspost a notice to the WikiProject talk page for further input? Thanks – Mélencron (talk) 16:34, 25 February 2017 (UTC)
 * Sure. Impru20 (talk) 16:51, 25 February 2017 (UTC)


 * Drop it — MOS:DASH is a guideline, not a policy. One of the reasons it's not a policy is that it doesn't matter that much. Edit warring over these low-stakes issues is disruptive, and WP:POINTy. Particularly in articles with much more noticeable style problems (WP:EL, MOS:BOLD, WP:EASTEREGG, etc. Find something productive to work on, and if one or more of these articles reaches a state of such perfection that you are ready to nominate it for WP:GA, wait to see if the reviewer notices or cares if there are spaces around the dashes. --Dennis Bratland (talk) 19:35, 25 February 2017 (UTC)
 * Comment I'm not really sure who is arguing for what here, but if the choice is between "31 Jan – 1 Feb" and "31 Jan–1 Feb", I would prefer the latter as in many cases we are stretched for space in these tables and the spaces are unnecessary. Number   5  7  21:57, 25 February 2017 (UTC)
 * Comment - I agree with other editors above that it doesn't matter very much either way. It doesn't look any different to me. The only condition that it would be worth arguing about is if there is some technical implication, for example regarding space in the tables. Jdcooper (talk) 13:19, 27 February 2017 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified one external link on Opinion polling for the German federal election, 2017. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20170317224352/http://www.spiegel.de/politik/updateumfragen/spd-konstant-ueber-30-prozent-afd-und-linke-im-osten-stark-spon-wahltrend-a-1138683.html to http://www.spiegel.de/politik/updateumfragen/spd-konstant-ueber-30-prozent-afd-und-linke-im-osten-stark-spon-wahltrend-a-1138683.html

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 15:58, 20 November 2017 (UTC)

Missing samples for 2015
The sample sizes are missing for 2015. It seems odd that they would only be missing for this year. Does anyone have them lying around or do I need to start adding them manually? Larssonjohan (talk) 11:09, 14 December 2022 (UTC)