Wikipedia talk:WikiProject Elections and Referendums

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Elections and Referendums (Rated Project-class)
WikiProject icon This page is within the scope of WikiProject Elections and Referendums, an ongoing effort to improve the quality of, expand upon and create new articles relating to elections, electoral reform and other aspects of democratic decision-making. For more information, visit our project page.
 Project  This page does not require a rating on the project's quality scale.

Wiki Loves Pride![edit]

You are invited to participate in Wiki Loves Pride!

  • What? Wiki Loves Pride, a campaign to document and photograph LGBT culture and history, including pride events
  • When? June 2015
  • How can you help?
    1.) Create or improve LGBT-related articles and showcase the results of your work here
    2.) Upload photographs or other media related to LGBT culture and history, including pride events, and add images to relevant Wikipedia articles; feel free to create a subpage with a gallery of your images (see examples from last year)
    3.) Contribute to an LGBT-related task force at another Wikimedia project (Wikidata, Wikimedia Commons, Wikivoyage, etc.)

Or, view or update the current list of Tasks. This campaign is supported by the Wikimedia LGBT+ User Group, an officially recognized affiliate of the Wikimedia Foundation. Visit the group's page at Meta-Wiki for more information, or follow Wikimedia LGBT+ on Facebook. Remember, Wiki Loves Pride is about creating and improving LGBT-related content at Wikimedia projects, and content should have a neutral point of view. One does not need to identify as LGBT or any other gender or sexual minority to participate. This campaign is about adding accurate, reliable information to Wikipedia, plain and simple, and all are welcome!

If you have any questions, please leave a message on the campaign's main talk page.

Thanks, and happy editing!

User:Another Believer and User:OR drohowa

More issues with Polling tables[edit]

Why is it that pretty much every single polling table on elections articles on Wikipedia violates WP:SALORDER by being organized in reverse-chronological order?! As far as I can tell, there is no compelling, policy-based reason for this, and it seems to be a case "they've been done this way since the beginning, so why change them?..." kind of thing. Does anyone else see a problem with polling tables violating WP:SALORDER for no particularly good reason, esp. in articles for past elections where these polling lists are now stable? (I'm wondering if this might be a job for a bot, if consensus can be established that these lists should follow SALORDER like every other chronological list on the project...) --IJBall (contribstalk) 22:23, 14 May 2017 (UTC)

I think showing the polls in reverse order is beneficial to the article; it shows the most recent poll first and therefore is not confusing to readers. MB298 (talk) 22:56, 14 May 2017 (UTC)
Considering nearly every one of those tables is sortable, there is no compelling policy-based reason to ignore WP:SALORDER in their standard configurations. IOW, they should default to forward-chronological order like pretty much every other chronological table on the project. --IJBall (contribstalk) 23:06, 14 May 2017 (UTC)
I strongly agree with User:IJBall that polls should be in forward-chronological order. They are only "useful" if they were perhaps intended as news articles. However, as WP is an encyclopedia and therefore historical in scope, they should be forward-chronological order.—GoldRingChip 18:30, 29 October 2017 (UTC)
Is there a way to propose a formal discussion to seek consensus and a Manual-of-Style-like rule?—GoldRingChip 18:31, 29 October 2017 (UTC)
@GoldRingChip: The obvious way forward on that would be a formal WP:RfC. I imagine that such an RfC should be hosted here, with RfC notices for this at the obvious places: Wikipedia talk:Stand-alone lists, and probably WP:VPP (and possibly elsewhere, such at WP:CENT). My one concern on this is that such an RfC might not garner enough "feedback" to be useful... But if we want to "officially" change the way polling tables are handled at WP:WPE&R, an RfC is probably the only way to do it. --IJBall (contribstalk) 19:33, 29 October 2017 (UTC)
Agreed. If you and I are wrong (which is fine, frankly), it should be decided that way. Still, there are so many articles that have these polls in reversed order, that a consensus should be developed. Would you please make the RfC? There should links back to it from lots of talk pages of active elections (i.e. current/recent) such as Talk:United States Senate special election in Alabama, 2017.—GoldRingChip 21:37, 29 October 2017 (UTC)
I'm on the busy side now, too busy to start an RfC (never started one myself yet). Maybe next month... --IJBall (contribstalk) 22:17, 29 October 2017 (UTC)
No worries. I'll tackle it.—GoldRingChip 01:47, 30 October 2017 (UTC)

Turnout figures[edit]

Hi, I'm doing a bit of work on some local elections (in the United Kingdom), and have been pointed over here by the teahouse regarding this query, and was wondering what the best way to show rejected/spoilt ballots in the turnout figures was inside the election boxes - should I just add them to the total turnout figure and leave it at that or create and extra row for rejected ballots? — Preceding unsigned comment added by Ballotboxworm (talkcontribs) 12:39, 29 May 2017 (UTC)

RfC on voting method[edit]

I've started an RfC on Talk:voting method about the naming of that article. I apologize for reopening this issue, but I feel that the clarity of the evidence warrants it in this case. Homunq () 15:17, 29 May 2017 (UTC)

Infobox problem[edit]

I'm having problems with Template:Infobox Election at United Nations Security Council election, 2017. I can't get it to display the newly elected members of the council properly, and I don't know what I'm doing wrong. Could someone here please have a look. Thanks. Sir Sputnik (talk) 22:06, 2 June 2017 (UTC)

Set | ongoing = no to show results. PrimeHunter (talk) 01:38, 3 June 2017 (UTC)

How much prose should election articles have?[edit]

It seems like in some cases, people want to remove almost all the prose from election articles and just have a bunch of data about endorsements, polling, fundraising, and results. Biographical information about the candidates, and information about the political controversies in the races, along with significant campaign events such as debates, gets eliminated.

Also, in articles about constituencies (e.g. districts), people sometimes want to eliminate almost all the prose and just have data about election results. Let's say there's a constituency that has had elections for the past 200 years. If you include information about the issues, controversies, and events of the most recent election, that gets removed as recentism. But of course, if people would keep adding this kind of information with each successive election, then 100 years from now, we would have 100 years worth of information about the constituency's political history. If that information gets removed every time, then 100 years from now, we still have nothing but results data. Compy book (talk) 00:51, 20 June 2017 (UTC)

All I can tell you is that I'm going through the WP Elections and Referendums "popular pages" report right now, and it looks to me like too many election articles have too many tables, and not enough prose. FWIW. --IJBall (contribstalk) 03:47, 20 June 2017 (UTC)
I haven't seen so much removal of prose, but many articles absolutely need more of it. So many need to actually cover the campaign and not just results. Reywas92Talk 06:17, 20 June 2017 (UTC)
@Compy book: Is there a specific article that has made you say this? Number 57 12:23, 20 June 2017 (UTC)
I too support the call for more prose. It seems to me more often a case that prose never gets written in the first place rather than that it gets removed. Bondegezou (talk) 12:27, 20 June 2017 (UTC)
"About the candidates" sections routinely get scrubbed from articles. Mister Ernest Thayer (talk) 14:24, 23 June 2017 (UTC)
Interesting. Thank you for the example. That seems to represent practice on US election articles: I've not seen the same on UK or many European election articles. Possibly it's something, therefore, better taken up on a more specific US project page? Bondegezou (talk) 14:53, 23 June 2017 (UTC)

Treatment of percentage change on previous election[edit]

I have been looking at Parliamentary election results in the UK, following the recent General Election (and occasionally filling in some details). There seems to be no consensus on how the '±' column should be completed if the relevant party (or individual independent) did not stand in the previous election. I would say, as a matter of common sense, that, if, say, Party A stood in 2017 and got 2.5% but did not stand in 2015, that marks a +2.5 on the previous figure, but many editors seem to put 'N/A' in the column instead. This seems less logical, but is there some general psephological convention being followed here? There really seems to be no consistency of approach.Ntmr (talk) 15:41, 21 June 2017 (UTC)

"N/A" or "–" would seem to be correct to me in this case, as you can't really say there was a "2.5% swing" to that party from the previous election – the party didn't exist before!! --IJBall (contribstalk) 16:01, 21 June 2017 (UTC)
@Ntmr: The most common way of dealing with it across all election articles is to put "New" in the cell in the table. See e.g. South African general election, 2014 or São Toméan legislative election, 2014. Number 57 16:32, 21 June 2017 (UTC)
We could also say that it's an infinity percent increase. Mister Ernest Thayer (talk) 14:25, 23 June 2017 (UTC)
I like putting "new" (or "N/A" as a second choice). Bondegezou (talk) 14:53, 23 June 2017 (UTC)

Results of French legislative elections before 2002[edit]

I'm beginning a project to clean up and correct old French election articles and have been having some trouble locating results of old elections. Right now I'm attempting to locate results of the 1973 legislative election; has 473 out of 490 constituencies, with the source being the CDSP. These numbers also correspond to the totals on the france-politique archive (hobbyists). The National Assembly also has published results for the 1973 legislatives which differ from the above, but I can't find the original source it cites. What is more, neither the French nor English articles on the topic seem to cite sources for the numbers they use (inserted into their articles in 2006) – which differ from both of the previous. Any help locating complete results (by nuance, with seat numbers, etc.) would be greatly appreciated. Mélencron 15:39, 23 June 2017 (UTC)

IP vandalism – made-up election results[edit]

Just a heads up, I've seen a couple of IPs adding apparently made-up result figures to several articles on 19th and early-20th century election articles in the last couple of days. I think I have most national election articles on my watchlist (exceptions being the UK and US), but just something others should probably be aware of. The IPs I've seen so far are (talk · contribs) and (talk · contribs). Cheers, Number 57 20:06, 6 July 2017 (UTC)

Belgian electoral data[edit]

The 1971 election for 212 parliamentary seats is missing 11 seats in the table: 7 from PSB(BRUX) and 4 from PVV/PLP.,_1971

Idea for SVG maps with no exact county results[edit]

I've been recently turning .png files into .svg ones for election box maps in Minnesota - gubernational, senate, presidential, etc. Have about 50 done so far. However, as I'm getting into very old results without exact county data, I have been using 2 colors to denote who won the country, instead of using different shades to denote the margin they won by. Here are 2 articles to explain what I mean.

When we have exact data - Minnesota gubernatorial election, 2002

Idea for no exact result maps - Minnesota gubernatorial election, 1990

The current maps for elections with no specific county data is Minnesota gubernatorial election, 1986, and I am trying to clean these maps up to our current standard. I bring this up because as I expand into other types of maps, I'd like to create a standard color scheme and format for all US maps without exact county data. What are your thoughts on this? I'd like to hear others opinions before I venture too far down this road. — Preceding unsigned comment added by PeterMGrund (talkcontribs) 04:36, 27 August 2017 (UTC)

  • Personally, I think a simple two-color "winner" scheme for maps like this is fine – it's certainly better than no county-level image at all. --IJBall (contribstalk) 05:36, 27 August 2017 (UTC)
  • This kind of map is inherently misleading, and next to useless. It places enormous weight on the physical size of the county, which is irrelevant, while failing to even hint at how many votes were cast, which is paramount. Winning a county with 3 million votes by three points is probably decisive. Winning a county with 50 votes by an 80% margin is rarely of any consequence at all, even if that county is thousands of square miles in size. This type of map gets that exactly backwards. In two-party elections with a third party spoiler, like 1996, the effect of the third candidate is obscured in almost every case. And then we have the fact that there is no such thing as "winning" a county. It's a fiction used for convenience, but when you codify it this way, it creates the misleading impression that these elections are a contest to win the most counties, and by the largest margin. Both are irrelevant. The candidate with the most votes statewide wins. Breaking it down by county is no more essential than breaking the votes down by gender or age or religion. It's just one kind of analysis, not the premier one.

    One of the prime arguments I've heard in favor of keeping these inadequate, misleading maps is that they are "standard". They're everywhere. But one has do ask, how did they become standard? It's because of decisions like this right here. I would strongly discourage putting time into making this type of map, and if another editor wants to put a better map on an article, it's harmful to tell them they can't because 100 other articles use these easily made but useless filled area maps.

    I really like the features of SVG, but I've also noticed that the existence of an SVG graphic biases editors towards using that graphic over a raster graphic image, even if the jpg or png is more accurate or more relevant. Converting an inferior map into SVG has a indicate to set it in stone as the "standard" and makes it harder to improve. SVG, sadly and ironically, becomes a hindrance to making the encyclopedia incrementally better. --Dennis Bratland (talk) 20:28, 3 September 2017 (UTC)

NH voter fraud discussion[edit]

Discussion of New Hampshire fraud allegations over at Talk:United States presidential election, 2016#More allegations of vote tampering. --Dennis Bratland (talk) 22:41, 7 September 2017 (UTC)

Project on local election information[edit]

I have spoken with others interested in promoting the visibility of local and state elections. Citizens would be well served by an easily linked and browsable project that provides information on elected offices along with their filing deadlines and requirements. Some current pages on these municipalities exist, but they are not entirely consistent, general election information is lacking, and specific information on how to run for an office is nowhere to be found (in my admittedly limited search thus far). I am currently researching formats and templates and would greatly appreciate feedback from this community on how to approach this.

My initial thought is to create an entirely new set of pages specifically dedicated to providing this sort of information, organized by state and then county, city, township, or other local government forms. After creating this content, relevant links and summaries can be added to existing pages on the municipalities. I'm looking at the U.S. Counties template as a place to start the organization of the project. Tgillet1 (talk) 05:44, 11 September 2017 (UTC)

@Tgillet1: I would bear in mind WP:NOTDIR. It may be worth you sharing an example here rather than creating dozens of articles that will be quickly deleted. Number 57 07:34, 11 September 2017 (UTC)
@Number 57: I appreciate your feedback. I've been going back and forth on the question of the format and whether it would conform to the standards or not (and between my own schedule and the others I've been trying to work with we've made little progress). Given the likelihood that many of the entries will be of mixed quality or at least of limited length at first (we're interested in rapid dissemination over completeness and high-quality content in the short term), I will start the project on another platform and then migrate to Wikipedia should the completeness and quality be sufficient. --Tgillet1 (talk) 01:37, 23 October 2017 (UTC)

Infobox again[edit]

There's a discussion around the infobox for the recent New Zealand election and what parties to include here: Talk:New_Zealand_general_election,_2017#Sixth_Party. Input from more people would be valuable. Bondegezou (talk) 14:21, 27 September 2017 (UTC)

Splitting general election articles[edit]

Is there a particular reason that for elections in some countries, the "general election" article lists both the presidential and legislative election results on the same article if they are held at the same time? I can understand the argument for it, but it would seem much more consistent – and clearer – with other presidential/legislative election articles if each had its own article, since each is its own unique topic, after all; e.g., this is the case in Turkey, Argentina, Ecuador, and quite a few other countries. Notably, when it comes to the U.S., the legislative (House of Representatives) and presidential election articles are separate, though they coincide on a 4-year beat. Mélencron (talk) 19:01, 28 October 2017 (UTC)

They should only be split if the article has exceeded the recommended size (which I guess is why the US articles are split). Otherwise having separate articles is not really worth it IMO – in many cases you'll have two threadbare articles instead of one with a reasonable amount of content. Number 57 20:14, 28 October 2017 (UTC)

Request for comment (RFC): Chronological order of election polling[edit]

Should election (or other opinion polls, for that matter), be listed in forward-chronological (oldest at top to most recent at bottom) or reverse-chronological order?—GoldRingChip 01:52, 30 October 2017 (UTC)


  • Support, per WP:SALORDER. WP is an encyclopedia and therefore historical in scope. It is not a newspaper or a blog.—GoldRingChip 01:52, 30 October 2017 (UTC)
  • Strong support, as per WP:SALORDER. In addition, most of these tables either are sortable, or can be made sortable, so there is no justification for violating WP:SALORDER in the pristine versions of the tables. --IJBall (contribstalk) 03:36, 30 October 2017 (UTC)
  • @GoldRingChip and IJBall: The RfC (as of this timestamp, anyway) is not asking a yes/no question but an either/or one.  — SMcCandlish ¢ >ʌⱷ҅ʌ<  04:29, 3 November 2017 (UTC)
    • Does this matter? --IJBall (contribstalk) 04:31, 3 November 2017 (UTC)
      Obviously "yes", or I wouldn't have mentioned it.  — SMcCandlish ¢ >ʌⱷ҅ʌ<  04:34, 3 November 2017 (UTC)
    • Thanks for letting us know. I don't think it's a problem, though.—GoldRingChip 12:38, 3 November 2017 (UTC)
  • Forward, per WP:SALORDER. While exceptions can be made, they're rare, temporary, and based on maintainability concerns at very long pages.  — SMcCandlish ¢ >ʌⱷ҅ʌ<  04:34, 3 November 2017 (UTC)
  • Reverse-chronological order (or case-by-case basis at most), as per Mélencron and other reasons discussed below. Impru20 (talk) 14:13, 3 November 2017 (UTC)
  • Reverse-chronological order, as it allows readers to see the most recent information first, and for most casual readers that's what they will be seeking. FriendlyDataNerdV2 (talk) 17:45, 3 November 2017 (UTC)
  • Oppose any decision that would determine ordering of opinion polls for the reasons detailed in the discussion below this survey. Mélencron (talk) 01:56, 4 November 2017 (UTC)

Threaded discussion[edit]

  • No particular view, but it seems something that should approached and decided upon on a case-by-case basis. The usefulness of listing opinion polls in reverse order is twofold: one, for in the months before an election for people to be able to find the most recent polls more easily. The second reason, however, is because opinion polls closer to an election are simply of much greater greater relative importance than those longer from an election, something which WP:SALORDER can't account for; i.e., comparing the relative accuracy of the final polls, rather than the polls for an election that were conducted four or five years out. I'd also suggest that readers are perfectly capable of accessing the end of the table and going in reverse order if they so desire, but to do so is significantly less common, even when looking at historical polling (at least in my case – though again, I'm the kind of person who's looked up opinion polls enough times on Wikipedia to think so). There's an obvious solution that accommodates both approaches, which is to make polling tables sortable, but unfortunately that rarely actually happens, as most editors don't know about "data-sort-value" (which was only implemented relatively recently, as far as I know) and as such it's rarely used where appropriate. Mélencron (talk) 02:11, 30 October 2017 (UTC)
  • Re IJBall: most of these tables either are sortable, or can be made sortable – the prior isn't the case (most, especially in U.S. articles, aren't) and the latter can't be down with considerable effort (that's talking setting "data-sort-value" for polls across hundreds of U.S. articles alone of various obscurity); it's made more difficult when the tables use date ranges in particular, which often aren't in the same format or organized in the same way (compare Greece, Israel, Finland, Turkey, and Ukraine, for instance – wildly inconsistent in formatting so it isn't something that can be done overnight.) There's 179 separate opinion polling articles and hundreds of articles including polling tables, and I don't see why a consensus on this should be determined on a project-wide basis as as opposed to a case-by-case basis where appropriate.
Also, maybe it's just me, but I'm used to reading polls in antichronological order – it's the way it's done everywhere outside of Wikipedia wherever you find poll aggregation or lists of polls: HuffPost Pollster (sadly now mostly inactive due to staff cuts), RealClearPolitics, and FiveThirtyEight; on (Germany) or (Austria), so I'm not sure why Wikipedia should be different in this regard. Is it unreasonable to think that more recent data is more pertinent to particular elections (say, data from 2017 for an election in 2017 as opposed to seeing data from 2012 first)? I lean against a change to chronological order, but I'm not voting because I think that a project-wide RfC is inappropriate since such a change would be a considerable logistical challenge (to either reorder tables or also make them sortable). Mélencron (talk) 03:45, 30 October 2017 (UTC)
  • Well, SALORDER is not something that should be ignored on a WP:LOCALCONSENSUS basis. More importantly, this is really more about establishing the proper way to handle this going forward. I don't think GoldRingChip or I expect people to go back and "fix" many old articles that don't follow SALORDER, though if this RfC passes, nobody should oppose that happening either. --IJBall (contribstalk) 03:57, 30 October 2017 (UTC)
  • SALORDER is a guideline, not a policy – one can make reasonable exceptions regardless of it and doesn't necessarily have to strictly adhere to it in certain cases, and I believe this is one such case. I'm of the opinion that it's reasonable to display opinion polling in antichronological order; a similar exception was made on the French Wikipedia earlier this year with a similar reasoning to what I gave (recency is pertinent to elections, so the polls closer to the date of an election are of greater relative importance). Even when viewing polling data well after an election, antichronological order makes sense since it shows the results closest to that election, not the previous one – and to me, I don't see the point of seeing polling data in chronological order from 2012 to 2017 on an article on the 2017 election. That's my take. Mélencron (talk) 04:04, 30 October 2017 (UTC)
  • See GoldRingChip's !vote – it perfect encapsulates my feelings on the issue as well. If people feel there is such value in seeing these in reverse chronological order, then there is an option for that: sortable tables. But, failing that, there is not a compelling reason to not follow SALORDER. --IJBall (contribstalk) 04:18, 30 October 2017 (UTC)
Same, but insert four letters before "chronological". Of the 28 national elections that have taken place so far this year for which opinion polls have been conducted and listed on Wikipedia, 23 have listed in antichronological order compared to only 5 in chronological order. Furthermore, all 27 of 27 opinion polling articles for future elections are ordered antichronologically. At least in the case of election polling, it's clearly more logical to list them antichronologically. SALORDER isn't a particularly challenge to this for any reason other than "it's a Wikipedia guideline" (which, in any case, is neither binding or universal). Mélencron (talk) 04:38, 30 October 2017 (UTC)
Mélencron talks sense, as usual! Key points I agree on: (a) there are problems with making these tables sortable; (b) the reality is that they are nearly all antichronological at present; (c) that is what reliable sources do; (d) SALORDER is a guideline and exceptions are allowed. That said, I am wary of LOCALCONSENSUS and we should try to follow guidelines, as per IJBall.
I note that SALORDER says, "Special cases which specifically require frequent daily additions, such as Deaths in 2017, may use reverse chronological order for temporary convenience, although these articles should revert to non-reverse order when the article has stabilized, such as Deaths in 2003." Opinion polling for future elections (at least, elections that are fairly soon) would appear to fall under that clause. I am tempted to suggest that we stick with antichronological for forthcoming elections, but try to switch to chronological order once an election is in the past. That seems to me to stick to SALORDER closely enough. I realise, however, that the reality is that it seems unlikely that editors are going to go through past articles to 'fix' them. Bondegezou (talk) 10:24, 30 October 2017 (UTC)
Would anti-chronological before the election and chronological after make sense? Too much work? After the election the polls would be historical not looked at as news.--BoogaLouie (talk) 01:40, 31 October 2017 (UTC)
I'd be OK with this as a compromise, though I still am unclear on why sortability for these tables (going forward) is such an issue – template {{dts}} was created for exactly this (apparent) issue (that, or use of the aforementioned "data-sort-value" in tables...). --IJBall (contribstalk) 04:36, 3 November 2017 (UTC)
  • Comment I have no particular interest in this, but users like @Impru20, Timeshift9, Clesam11, Ron 1987, Nub Cake, and FriendlyDataNerdV2: etc might want to input as they seem to do a fair amount on opinion polling. One thing I would say is that if antichronolgical order is the standard method of presenting opinion polls in the real world, that that's what we should be doing here, regardless of a policy that has been developed without this in mind. Number 57 10:21, 3 November 2017 (UTC)
  • I essentially agree with Mélencron's points. I actually find the SALORDER policy a bit confounding, given that it doesn't give an explanation as to why lists should aim for chronological ordering. Clesam11 (talk) 11:26, 3 November 2017 (UTC)
  • Agreeing with Mélencron's points. SALORDER is not an actual policy but more of a guideline, plus the fact that the only reasoning being brought forward for its enforcement is that "it's a policy and it must be applied". Look, WP:IGNORE is an actual policy. In the unlikely scenario where we actually acknowledged SALORDER as an actual policy, reasons for listing opinion polls anti-chronologically have been explained and detailed, and they're all reasonable and logical. The claim for applying SALORDER seems to be just that "it must be done like it because it says it". So, I not only agree with all the reasons for using anti-chronoliogical lists for opinion polling articles–which do clearly improve reading and help readers in that most recent polls (which are arguably the most important ones as those are the nearest to the election date) are placed first–but I also add a new one, which is that "if a rule prevents you from improving or maintaining Wikipedia, ignore it". Impru20 (talk) 13:06, 3 November 2017 (UTC)
Except not one of you has dealt with the issue of sortable tables, which would allow SALORDER to be followed, while still giving people access to reverse-chrono order listings... --IJBall (contribstalk) 13:31, 3 November 2017 (UTC)
I thought someone had already explained that making these tables sortable would mean a titanic effort (creating some other issues along the way, as I've seen in the few opinion polling articles using sortable tables), as well as the fact that these tables are not usually made sortable in most cases. We have also the fact of (reliable) sources using reverse-chrono order listings for opinion polling lists, so it's not that it's us alone that think it should be done this way. Maybe you could explain why is the application of SALORDER so badly needed (given that it's not a policy, that it also speaks of "should" and not "must", that it provides exceptions may exist and that no reason other than applying it for the sake of it is argued, against many reasons argued against it) so as to require any of us to take on such an enormous task, which involves having to manually apply "data-sort-value" or whatever other template individually over hundreds (if not thousands in some cases) of opinion poll entries per article, over hundreds of such articles in Wikipedia. Impru20 (talk) 13:53, 3 November 2017 (UTC)
@Impru20 and Clesam11: Perhaps worth noting that if you are against it, there is a !vote section above (this is an RfC). Number 57 14:04, 3 November 2017 (UTC)
And, again, this is more about how these tables should be done, going forward, and much less about "fixing" old tables (though that should be allowed too, should some editors want to undertake that...). I have yet to hear of anything that would make formatting these table as sortable to be too great a technical challenge from this point on. --IJBall (contribstalk) 14:15, 3 November 2017 (UTC)
Thanks for the tag, @Number 57:. I think we should order tables with most recent polls first, quite simply because most readers come to Wikipedia seeking the most recent information, especially on current elections. It's how such tables are presented in the real world because it's the most recent information. It's the most common sense thing to do in my view and I'd be strongly against changing it. FriendlyDataNerdV2 (talk) 17:44, 3 November 2017 (UTC)
WP:NOTNEWS. Wikipedia is not a "news" wesbite – it's a historical record of events (which is why SALORDER exists). User:BoogaLouie's suggestion up-thread would be acceptable as a compromise, by GoldRingChip's overall point above is valid, especially for organizing these polling pages, post-elections. --IJBall (contribstalk) 17:56, 3 November 2017 (UTC)
The historical order of opinion polls does tend to be in reverse order, however – as I've mentioned above, outside of Wikipedia, other places where opinion polls are aggregated list them in reverse order. SALORDER isn't a compelling argument here other than for the apparent need to comply by a guideline, which, in any case, can be reasonably ignored. Mélencron (talk) 18:23, 3 November 2017 (UTC)
And, yet, I still don't see how sortable tables doesn't just solve both problems – then tables can default to SALORDER, but can easily be sorted in reverse-chrono. I have yet to see anyone explain why we shouldn't just do this, as it solves all the issues. I'm really not trying to be difficult here – making these sortable tables, from henceforth, really is the best answer. --IJBall (contribstalk) 18:29, 3 November 2017 (UTC)
It's not if the default set-up is chronological. If they were sortable with the default being reverse-chronological, very well. I've not seen any other suggestions that this is an issue, I'm surprised it's suddenly become one. FriendlyDataNerdV2 (talk) 19:44, 3 November 2017 (UTC)
It's not an issue to the small set of WP:WPE&R "regular" editors. But it is an issue to the wider set of editors who aren't WP:WPE&R "regulars" or election junkies. This is a common issue – a specific wikiproject develops a set of "standard practices" that are at odds with the wider En Wikipedia community. This isn't the first example of this, and it won't be the last. (P.S. Going to a standard practice of sortable tables, defaulted to reverse-chrono order, would be better than nothing IMO, but it still would not be my preferred outcome here, as I'd rather we stick to SALORDER, esp. post-elections...) --IJBall (contribstalk) 20:44, 3 November 2017 (UTC)
How has WP:NOTNEWS any relation to the fact that placing most recent opinion polls first is more useful to readers? Opinion polls are pollsters' predictions of what the next election results are going to be. An opinion poll released 10 days ahead the election is going to be much more relevant than one conducted three years previously. Always, in every country you may think of. For the simple reason that opinion polls are meant to predict an election result, and predictions released closer to such an election have much more media prominence and relevance than those conducted farther away in time. The historical order of opinion polls is shown reverse-chronologically by reliable sources; what's the issue of us doing exactly that? Impru20 (talk) 19:54, 3 November 2017 (UTC)
If your only metric is the ultimate election result, then you may have a point. If, however, you're using opinion polls to gauge party support over time (which is more likely to be one of the aims of a "historically-minded" encyclopedia like this one), then putting the polling results in forward chronological order has more historical value. I mean, otherwise, why don't we put the polling figures/graphs at these articles in reverse chronological order as well?! --IJBall (contribstalk) 20:44, 3 November 2017 (UTC)
I mean, otherwise, why don't we put the polling figures/graphs at these articles in reverse chronological order as well?!
Charts/graphs are horizontal. Tables are vertical. Gauging party support over time is done equally with both a forward-chronological order and a reverse-chronological order when put vertically, but it may indeed cause some issues when done horizontally. Nonetheless, horizontal charts/graphs have nothing to do with SALORDER, so I don't know what connection does it have to this.
This is not a matter of people preferring anti-chronological order for opinion polling tables for the sake of it, but because it serves a purpose, and because it's what reliable sources usually do as well. I still have yet to see what is the purpose of trying to enforce a non-policy such as SALORDER for opinion polling tables other than doing it because, somehow, it "must be done so". Impru20 (talk) 20:55, 3 November 2017 (UTC)
And I'm still back to why can't we just make the tables sortable, so we can follow SALORDER and allow for reverse-chrono. listing for anyone that wants it as well? It's the best of both worlds. Bottom line: Even if reverse-chrono. ordering is kept, these tables should be made sortable from now on, for anyone who wants to check chronological ordering in either direction. --IJBall (contribstalk) 21:00, 3 November 2017 (UTC)
You were already replied on this several times by several users.
such a change would be a considerable logistical challenge (Mélencron)
the reality is that it seems unlikely that editors are going to go through past articles to 'fix' them (Bondegezou)
making these tables sortable would mean a titanic effort (Myself)
And also this: There's 179 separate opinion polling articles and hundreds of articles including polling tables, and I don't see why a consensus on this should be determined on a project-wide basis as as opposed to a case-by-case basis where appropriate (Mélencron)
I'll try to explain it again in a detailed way. In order to make these tables sortable, you'll be required to
  1. Manually apply "data-sort-value" or whatever other template individually over hundreds (if not thousands, in some cases) of opinion poll entries per article
  2. Do that over hundreds of such articles throughout Wikipedia
  3. Then making sure such a change is done properly without causing serious issues to the way opinion polling tables work, because otherwise you would require to re-edit all of these again
  4. And I don't see you offering yourself to try to accomplish such an enormous task, so I fear this will end up being tasked to those who–like me–do frequently edit and keep these pages up to date. Specially if this is intended to be imposed through a RfC instead of done in a case-by-case basis
And if there was any urgent, relevant and/or useful reason to merit such a titanic effort, then I'd say "well, go ahead". But from the beginning I'm seeing no reason other than arguing that SALORDER must be applied to those for the sake of it, when it is not even a Wikipedia policy or a compulsory and unavoidable guideline. Impru20 (talk) 21:25, 3 November 2017 (UTC)
You keep ignoring the fact that on multiple occasions in this thread, I have made it clear that I am not advocating that "old" articles be fixed. What I am saying is that all such tables from this point forward should be made sortable. Yes? --IJBall (contribstalk) 00:16, 4 November 2017 (UTC)
  • Two thoughts... If tables can be made sortable on date, then that would be a good thing. I realise that is difficult and there are many articles to cover, but if someone wants to do it, then let them do it. As for what reliable sources do, most reliable sources covering current politics use reverse chronological, but chronological is used, and I think is more common, for academic, historical papers. Bondegezou (talk) 10:11, 6 November 2017 (UTC)

Request for Comment (RFC): Consistency in United States Electoral Maps[edit]

As I've been slowly converting the thousands of United States electoral maps currently in .jpeg/.png format to .svg, I've been coming across many inconsistencies, even with maps made in 2016. For example, when comparing 2016 presidential election maps to 2016 senate election maps, an entirely different color scheme is being used, as well as a lack of color keys.

Should we streamline our electoral map making template for the United States to include all relevant elections (President, Senate, and any other state-wide election)? If a different map is better suited, then of course use that, this is just to streamline the process.

Peter M. Grund (talk) 02:16, 6 November 2017 (UTC)PeterMGrund


I'm not entirely sure what you're asking, but I want to thank you for your efforts on this, and if you are continuing with the conversion project I'd want you to use whatever makes it easiest for you and results in the most consistency across articles. Of the two above I prefer the presidential map because it has better contrast between the colors whereas it's hard to tell some of the senate map's colors apart. Reywas92Talk 02:46, 6 November 2017 (UTC)

  • Everyone is welcome to edit, but I'm opposed. SVG format seems like the best thing since the three letter acronym, but in practice it becomes an obstacle to change. Once anything gets converted to SVG, then it becomes mired in inertia when another editor wants to change it in any way, unless the new file is also SVG. So graphs with flaws, errors, or simply with less information, are preferred on the grounds that they are in SVG. Almost anybody has access to any number of spreadsheet, word processor, graphic editing, or data visualization software tools that will produce files in png, gif or jpg format. Often raster file tools are already installed on a Windows or Mac PC. SVG tools are fewer and far between, and represent an additional cost or installation obstacle. This problem would evaporate if there were wide consensus that a fancy format like SVG is an argument to avoid, but until then the resistance to basic editing policy exists. Keep in mind that we have virtually any raster map or charts you can find here is in such high resolution that the scaling advantage is moot on any normal display our readers have.

    The other major problem here is the use of county-level choropleth maps, whose only advantage is that county data is very easy to come by. We are often stuck with it, but standardizing on any map of this type is extremely harmful because US county filled maps are very misleading. The visual weight is determined by the size of the county, which is meaningless. We want to know how many votes were cast for each candidate, and blowing that up in relation to how many square miles a given county happens to have is false and misleading. These are graphics we are talking about, and visual impact is everything. If you expect readers to mentally correct for the way county size misleads them, why not just use a crosstab? We should not standardize on a poor graphic out of mere convenience when it is so harmful. We have a good alternative: pie File:Minard-carte-viande-1858.png. It allows us to scale the visual impact to match the number of votes cast, and to give an accurate magnitude of a candidate's votes. These single color or color tone filled maps eliminate everything except the county's winning candidate, while pies show us two leading candidates, and easily accommodate three-way races, such as in File:Utah 2016 presidential results by county.png.

    Any editor who wishes to make maps, be they in SVG format or any other, and in filled county or other types of graphic, is welcome to edit. But I strongly oppose standardization on such a sub-optimal type of map. I would encourage making maps for articles that have nothing at all, rather than replacing existing ones with the goal of standardization. --Dennis Bratland (talk) 04:45, 6 November 2017 (UTC)

I'm not as familiar with the format debate but that is true that this is not the type of image that needs to be blown up (or would cause detail problems being in the wrong type). I love your Utah map! Is there a program you use that makes this type of map easy to make, or is every chart pasted on manually? Skimming the discussion in the Utah 2016 article I generally agree with you against the silly argument that chloropleths are commonly used and therefore should be the standard, though they both could be used. I don't care for the inclusion of the less-than-minor candidates though, not even being visible in the pies. County names are too small for legibility on the WA map but at least they're there! Reywas92Talk 07:38, 6 November 2017 (UTC)
I mostly use Tableau, but I think there are other tools with similar features. They're at c:Category:Pie chart maps of the United States presidential election, 2016 (results by county). On some of the maps I combined all the minor candidates into "other", such as File:Michigan 2016 presidential results by county.png. On Washington map I had some reason (I don't remember) for smaller county name fonts, but you can read the county names when you click on it. The details about minor candidates or font size don't need to be standardized anyway. Future editors can make incremental improvements with any raster editing tools, even if they don't have Tableau or whatever special program. Working with county-level data is creates the false impression that it even matters who "wins" a county, when in fact the race is over who wins the state. Breaking it apart by county or precinct or census block is arbitrary, and county is one of the most misleading ways, since the population of counties varies by a factor of 1,000 or more, and land area is often inversely proportional to number of voters. But in most elections that's all we have so we have to make the best of it. --Dennis Bratland (talk) 19:26, 6 November 2017 (UTC)

RfC Should articles say elections are decided based on preliminary returns?[edit]

Are elections described as decided before the result is certified? Should we say in Wikipedia's voice that an election outcome is decided, and say in infoboxes such as {{Infobox election}}, {{Infobox officeholder}} that a candidate is the winner, and is now an incoming officeholder (with a future start date), calling them <office>-elect, based on major media calling the race using projections from preliminary returns, before the final vote tally and certification of the result? Do we base this only on major media annoucing a winner, or is this article content influenced by whether the presumed losing candidate has conceded or not? And is this article content influenced by our own calculations such as the mathematical probability of overcoming a deficit in the uncounted ballots?

Alternatively, do we only say, in prose using in-text attribution, in the article body not in infoboxes: "CNN, the NYT, and WGBH have called the race for candidate X, and they predict that the uncounted ballots are insufficient to overcome their lead over candidate Y" and "Candidate Y has/has not conceded"?

Does this reasoning also apply to sports contests or other outcomes where it is widely agreed by reliable sources that it has been decided, but there is some time delay before an official result is verified and announced, or does this apply only to elections? --Dennis Bratland (talk) 04:57, 9 November 2017 (UTC)


  • Wait for the result to be certified before calling anyone -elect or putting any outcome in an infobox. Describe opinions and predictions calling the race with in-text attribution only, in the article text only. Per WP:CRYSTAL, WP:WikiVoice, and WP:NOTNEWS. --Dennis Bratland (talk) 05:19, 9 November 2017 (UTC)
  • No need to wait for certification. Certification is a long process, and is not needed when multiple RSes clearly call the race. We do not need to wait 3 weeks in a 60-40 race if it has been called, without qualification, by multiple RSes. The same goes for other events with a potentially long certification process.Icewhiz (talk) 07:17, 9 November 2017 (UTC)
  • Don't wait – doesn't really make sense, anyway; the winner is basically called with the exit poll/estimates based on counted ballots in France (which has never gotten an election wrong) for example, while counting takes much longer. Certification of results in many European countries can take more than a week after the election, and I expect it's probably the same in the United States, so there's really no need to wait if the winner is clear after the election and a call has already been made by major media or candidates have conceded defeat. Mélencron (talk) 12:36, 9 November 2017 (UTC)
  • No need to wait I don't know how things work in Seattle, the case at issue here. In most U.S. jurisdictions, the responsible official (registrar, secretary of state, etc.) reports the raw vote count immediately after the election, as fast as it can be counted. That vote count is technically considered "preliminary" and will be "certified" at some time in the future - usually set by law, sometimes depending on how long it takes them. Our practice here has been to report the election results ("so-and-so won") based on that raw vote count without waiting for the formality of certification. The rare exception would be when it is actually in doubt - too close to call, or pending a recount, or pending a challenge, or sometime like that, in which case we report the situation. Others here have noted other ways of recognizing the result - Reliable Source reporting, concession by one candidate, etc. - which are also based on a raw / incomplete / uncertified vote count. Bottom line, we usually are able to report who won and who lost on election night or the next day. Of course, the person will not actually take office until some later date set by law. So we need to guard against the overly enthusiastic editors who try to say that the subject already IS the mayor, governor, or whatever, or to inappropriately change the term of office in the infobox. But as far as putting the election results into the article, we virtually always do that based on the immediate raw vote count. (Obviously, this applies to elections and has nothing to do with sporting events or other contests, where the result is not reported until the contest is over.) --MelanieN (talk) 14:34, 9 November 2017 (UTC)
  • Don't wait In several countries results are only certified several weeks after the vote and we shouldn't be waiting if the preliminary results (often released by the electoral authority) are clear and widely reported (there is usually little change); however, the results section should probably be titled "Preliminary/Provisional result" until that point (one issue I have noticed in a few places is the preliminary results being left in articles as the definitive figures despite the certifying bodies making a few amendments). Also, some countries do not have a centralised certifiction process. Number 57 14:53, 9 November 2017 (UTC)
  • Don't wait if it is a two-person race and one has conceded but wait if there is a serious challenge (especially if there is a court case). In other cases I think we have to apply good judgment, not use WP:FRINGE sources etc. Sorry for voting both ways but I don't think this is something that can be done with one formula that works every time. Concrete examples, in Seattle the mayoral candidate has conceded so that's a done deal. But in the King County Sheriff's race, one candidate is ahead in the preliminary counts, there has been no concession, and it still could go either way; we should still report it that way. It is an especially complex situation in Seattle at times because there are no polling places in King County anymore at which to conduct exit polls; it is 100% vote by mail. ☆ Bri (talk) 18:02, 9 November 2017 (UTC)
  • Don't wait In most countries, including the United States, the preliminary return of election results likely highly correlate with final, certified results while the preliminary returns are posted on the electoral administrations' official websites or trustworthy third-party sources (e.g. AP). It has been common practice to list preliminary/unofficial results and declare winners as-is, since the results rarely change. However, wait if there is an active court challenge or recount, like the case with United States Senate election in Minnesota, 2008. In the case of Seattle, the preliminary return is clear showing a Durkan victory, and Moon has also conceded, so I don't see the point of waiting. Ueutyi (talk) 19:39, 9 November 2017 (UTC)
  • Don't wait For goodness's sake, yes we can say in the infobox that Durkan won the election and that she is mayor-elect in relevant articles. The media has called it and her opponent conceded; certification is a formality that takes weeks. The beauty of Wikipedia is that we can change things if some sort of challenge were to come up, which can be addressed case-by-case. Reywas92Talk 21:53, 9 November 2017 (UTC)
  • Wait - NOTNEWS and factually untrue as not yet officially 'won'. Besides, according to early mentions in the U.S., Gore and Clinton had it in the bag -- and that turned out to be untrue. Markbassett (talk) 00:00, 10 November 2017 (UTC)

Threaded discussion.[edit]

  • My reasoning is that it is enough for us to describe in text that an important source or sources has called the race. We can describe what their rationale is for presuming it is decided, but not treat it as fact or change any infoboxes until certification. Unless election law officially recognizes a concession as relevant to the final outcome (is this the case in any jurisdiction?), concessions shouldn't influence what our infoboxes or what Wikipedia's voice says as facts. If a secondary source has said a result mathematically certain due to the number of uncounted ballots being less than the margin, we can describe what they said. We should not say this based only on our calculations using the raw, primary source data. That would violate WP:NOR and WP:SYNTH. See also: WP:TRUTH.

    Predictions based on likely vote outcomes are not facts, they are opinions. Articles should always clearly distinguish facts and opinions. I like having opinions in articles, but they must be framed honestly, as opinions. I am surprised this issue doesn't already have clear guidelines in WikiProject Elections, and I'm surprised that many editors think we should jump the gun this way. I have always favored a long timetable, and conservatively biding our time until the dust has settled before creating new articles or describing news events. We should treat sports events or anything the same way: it's not over until the officials have announced it, and we have all the time in the world.

    It is the role of a newspaper to give the public up-to-the-minute advice and guidance, and give them hints about election outcomes rather than leave them guessing. It is not the role of an encyclopedia to be in a rush to declare contest decided as long as alternative outcomes are possible. The Wikipedia is not a crystal ball policy is explicit about this. NPOV policy says we do not state opinions as facts. WP:NOTNEWS says we take the long view, and feel no pressure to track the latest developments. --Dennis Bratland (talk) 05:28, 9 November 2017 (UTC)

If we follow this procedure then this statement would not be allowed: "On November 9, 2016, at 3:00 AM Eastern Time, Trump secured over 270 electoral votes, the majority of the 538 electors in the Electoral College, enough to make him the president-elect of the United States" (found at United States presidential election, 2016) – in fact no one could be described as president elect until mid December after the Electoral College meets. This seems less than ideal and badly out of sync with the real world. Or am I missing a nuance? ☆ Bri (talk) 06:47, 9 November 2017 (UTC)
The electoral college meeting is a special case, and it's generally understood as a formality, since faithless electors changing an outcome is a phantom. News media calling a state prematurely is no phantom; it happens all the time. What I'm talking about is those final vote tallies: it's fine to say CNN has called a state, but if that state only counted 30% of the ballots they have on hand, it's not a fact. It's an opinion based on the assumption that the reaming 70% of the ballots will not be much different than the first 30%. Premature calls by the media are not a problem if we frame them and attribute them as opinions.

Referring to the current example, this article lead says plainly that the election was called. We aren't keeping that information from readers. This version adds that Moon has conceded, and then says, in Wikipedia's voice, "Durkan will take office on November 28, after the results are certified." That is not a fact. That is an opinion based on assumptions about the remaining votes. This kind of statement is what you'd see in a newspaper, intended to keep citizens abreast of developments and upcoming events. Saying "it's going to rain tomorrow" is useful newspaper-type information, even if it is unscientific and not precise. An encyclopedia exists to provide insight, to educate. So an encyclopedia article should be precise in explaining that the predicted outcome is an opinion based on reasonable assumptions. The reader might still end up confident that Durkan will be sworn in on November 28, but aware that is an educated guess. --Dennis Bratland (talk) 07:27, 9 November 2017 (UTC)

  • @Number 57:, @MelanieN: Can you clarify again what you mean by "in the article"? I'm sorry to badger you it's just that this is kind of a fine point.

    Nobody has objected to the text in the lead here saying "Durkan had over 60% of the vote in the preliminary count on the last day of the election, and The Seattle Times called the race for Durkan, predicting that Moon was unlikely to overcome Durkan's lead when the uncounted ballots were added leading up to the final count on November 28, 2017." It is a verifiable fact that the Times called the election. The question here is whether we also want the infobox on the right to say "Taking office November 28, 2017 Succeeding Tim Burgess". Or to add a name to List of Mayors of Seattle after only half the ballots have been counted.

    Your comments don't make it completely clear if you are talking about a result after all the ballots have been counted, or only a portion of them. The case here is one in which all the votes have not been counted. This isn't a matter of re-counting votes before certification. We're talking about a total of about 200,000 ballots cast, and only about 100,000 of them were counted on the first day. Because the result was about 60% to 40%, the news media declared a winner, even though the remaining ballots had not been seen by anyone. I am saying this is the same as declaring a winner before we know the final score. It's not a matter of overturning a result after a recount. The first count hasn't even been finished. --Dennis Bratland (talk) 17:52, 9 November 2017 (UTC)

I am surprised the count is going that slowly. Is that usual for your area? I was talking about the more usual situation where by midnight on election day, or early the next morning, 95% of the votes have been tallied and somebody is leading 55% to 45% - in other words, an obvious result even though it hasn't been "certified". The sentence you have in the lede sums up your situation nicely, although probably in unnecessary detail; I think Durkan had over 60% of the vote in the preliminary count on the last day of the election, and The Seattle Times called the race for Durkan would be sufficient. But I would change it to simply "won" as soon as the raw vote tally makes it clear that we know who won; I wouldn't wait weeks for the "certified" result which is a bureaucratic formality. Also, when the raw vote tally shows a winner I think mayor-elect or similar can be added to the person's infobox. I don't think they should be added to "list of mayors of Seattle" until they actually take office. --MelanieN (talk) 21:14, 9 November 2017 (UTC)
Yes, because all elections in WA are mail-in, it takes a while for ballots sent on election day to be processed and counted. But because they are evenly distributed the margin changes very little after the first couple days and calls can be made. There is no reason to ignore RSs' calls of the race for us to put the winner in the infobox. I agree that the original statement was needlessly long as it would have to be changed later anyway (and can still be changed in any other event). Reywas92Talk 21:58, 9 November 2017 (UTC)
Agreed with above – winners can be reasonably called in most cases before all ballots are counted, even if some provisional ones are missing. There are some exceptions, as with the first run of the second round of the 2016 Austrian presidential election, several districts in the Virginia House of Delegates, and other tight races, but Dennis is calling for an unreasonable standard here. Mélencron (talk) 23:05, 9 November 2017 (UTC)
In King County, WA, it takes 7-10 days to count 100% of the ballots, and official certification is typically scheduled for 3 weeks after the election day. We are not talking about some trivial number of "provisional" ballots, or spoiled ballots. We're talking about 40% to 50% of the total: regular, normal ballots with no signature challenges or anything like that.

It turns out that in this case we have a copious amount of detail on how long a count takes -- Seattle mayoral election, 2017#Results has a graph that even shows the exact rate of counting. Possibly more detail than we need but at the time it seemed relevant. On the first day, about 90,000 out of 180,000 ballots was counted. It took 9 days, until 8/9, to finish counting them all. This matches the milestone we usually have about 24 hours after of a US Presidential polls close, when the first count is done and the result of a given state is rarely in doubt. Here it wasn't until 8/15 that the primary result was officially certified.

If you like the text of the lede, but not declaring a winner in the infobox, then we agree. I wouldn't really mind waiting until 100% of the vote has been counted, without waiting until it is "certified". What I definitely object to is the infobox treating the outcome as fact when only 50% or 60% of the votes have been counted. This is analogous to ten minutes after the polls have closed on election night in a US Presidential race and there's as many uncounted ballots as counted. Even if a candidate concedes, those boxes of uncounted ballots are still Schrodinger's Cat. A concession doesn't change what is in those boxes. To me that is declaring a winner in game at the end of the third quarter.

@Reywas92: nobody suggested "ignoring" when the major media calls the race. It appears I'm having a difficult time avoiding being misinterpreted and dealing with straw man arguments. Yes, articles should say that the race has been called. Agreed! The question is, how and where do we present that? When you say that the ballots are evenly distributed and not especially skewed, you make a good argument, but it so happens that our sources don't state that argument. Many sources have the opinion that later-counted ballots skew left in Washington State. The Times didn't say it is mathematically impossible for the uncounted ballots to overcome the margin. The Seattle Times and KING5 called it without explaining why. When our sources don't show their work, should we treat it as fact? --Dennis Bratland (talk) 23:16, 9 November 2017 (UTC)

We are not agreed, because I did NOT like the text of the lede. I think it is too long by half, with a bunch of unnecessary hedging. And I am ok with adding "elect" to the info box as soon as the result appears clear. I would also like to suggest that your proposal here - "wait for the results to be certified" - may apply only to the state of Washington. Most jurisdictions come out with a raw count of most of the ballots within 24 hours. If you look at the discussion here, most people are saying we can go with that raw count, supported by media calls and/or concessions; there is no need to wait for that raw count to be certified. MelanieN alt (talk) 16:21, 11 November 2017 (UTC)