Jump to content

Talk:2018 Swedish general election/Archive 1

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 1

Bloc results

I added two columns to show the aggregate poll results for the two major party blocs (S+Mp+V and M+C+Fp+Kd). I felt that this was missing in previous versions as that is an important metric in deciding the balance of power. It is far more important than the lead the largest party has over the second largest party, which is basically irrelevant in a Swedish context. I checked the talk pages of previous elections and found that it hadn't been discussed before, so I decided to be bold a go ahead. Väsk (talk) 15:14, 1 November 2014 (UTC)

Undid it as per Talk:Swedish general election, 2014#Fake_party_names. The issue with this is that there is a difficult definition of 'blocs'. The only defined one would be probably the Alliance one, and not even it would be relevant now, with a minority government willing to reach agreements with some of those parties if needed. Furthermore, you include the Left Party in the Red bloc, even when it is NOT in the government coalition and it explicitly refused to enter it. There is just no 'blocs' in that sense of the word, really. Impru20 (talk) 16:00, 1 November 2014 (UTC)
EDIT: Re-added them, because opinion pollsters still show bloc results. Separated them from main table, though (given them a table of their own), with their own lead column. Impru20 (talk) 16:19, 1 November 2014 (UTC)
I modelled it on the tables for Danish polling results which have similar columns even though there aren't formal alliances between the parties there either. The bloc formations may not be formal alliances, but they are still very much relevant to Swedish politics. Swedish media and pollsters have been reporting polling results using these blocs since the Greens aligned themselves with the left bloc, as these bloc formations decide who gets to form a government. After all, Löfven proclaimed himself victor on election night because S+Mp+V got more votes than M+C+Fp+Kd, not because the Social Democrats got more votes than the Moderates.
Swedish polling results are hard to decipher unless you also include the bloc dimension. As I said, it is much more relevant than the difference between S and M, which is meaningless and should ideally be removed from the table. I would have preferred to have the bloc results included in the same table as the party results, but this is obviously better than not having them there at all. Väsk (talk) 17:24, 1 November 2014 (UTC)

"Next"

Given that Sweden has fixed dates for elections, I propose we move this article to Swedish general election, 2018. It follows an established precedent for Swedish election articles that have previously not had "next" in their titles. "Next" should only be used for countries without fixed election dates as it adds unnecessary confusion. Per Swedish election law, an election will be held on 9 September 2018, whatever happens. If an "extra election" was to be called before then, a dedicated article can be created for it. Väsk (talk) 20:31, 1 November 2014 (UTC)

"Snap Election" vs. "Special Election"

The Wikipedia article on Snap elections differentiates them from a By-election based on the fact that a new term of office begins after the snap election. By this definition, the constitution of Sweden does not allow for snap elections, and it should instead be described as a by-election, special election or extra election (literal translation of the Swedish term). The next election will still go ahead as scheduled in 2018.

195.24.173.88 (talk) 13:56, 4 December 2014 (UTC) Signature

Seems right to me. I also believe it will be right to consistently refer to the election in the singular instead of plural. Iselilja (talk) 17:38, 4 December 2014 (UTC)
The common way of talking about elections is to use the plural. Number 57 18:10, 4 December 2014 (UTC)
OK. I notice it's different in the title where we use singular, but you are more experienced than me with this (and English is not my first language). Iselilja (talk) 18:12, 4 December 2014 (UTC)
The proper translation seems to be "extraordinary", not "extra".[1] It seems to me the appropriate thing to do would be to rename the article "Swedish extraordinary general election, 2015", and amend the lead to match.
Incidentally, this is not a by-election (or special election). The major distinction between a by-election and a general election is that by-elections are held to fill casual vacancies in seats in an ongoing assembly. The speakership and other parliamentary offices are unaffected by the election, as are committee memberships (until the body chooses to add the newly elected member to a committee, which for most legislatures it won't be required to do). A general election is held for all seats (or a class of seats where the body has staggered terms). The term "snap election" applies to early general elections, which in almost any parliamentary democracy lead to a full term. Sweden is different, but the upcoming elections are more like an odd sort of snap election than an odd sort of by-election because all of that opening of parliament stuff (appointment of speaker, nomination of prime minister, etc.) will happen just as it did earlier this year. -Rrius (talk) 23:06, 4 December 2014 (UTC)
I agreee with Rrius about renaming the article. Here´s a link to the Riksdag website with a link to the Instrument of Government, just in case anyone doubts Rrius' link as a source. Sjö (talk) 06:21, 5 December 2014 (UTC)

Polling table

About this revert conflict:

  • (diff) (adds "sortable" to the polling table)
  • (diff) What is the point of having a sortable function in a date-ordered poll table?
  • (diff) The point is to give readers a chance to sort polling firms by name. Their methods differ; and mixing their results is unscientific.
  • (diff) Such a change should probably be discussed in talk first. Even if they use different methods, it's not the same a poll of October 2014 than a poll in December 2014. Mixing dates is more unscientific, IMO.)

Actually, if you add a "sortable" option to the table, when you sort the polls by name (of the polling firm), you will get a list in which polls are listed 1) by name 2) by the previous order, that is the date. This is exactly what we need for science. If you understand that, then please do as science wishes. 2A01:E35:2F0C:F510:3104:ED56:CAC1:96F9 (talk) 11:18, 6 December 2014 (UTC)

As I said, there is no point in doing this. Date is the important factor here, not name. With the change you propose, you will only manage to group polls of the same name together, but you are not even guaranteed to have them in chronological order, specially as more and more polls come out. That is not science, that is just weird. By the way, neither Wikipedia is to do "what science wishes", neither science in itself is a live subject able to have feelings such as "wishing". Impru20 (talk) 11:52, 6 December 2014 (UTC)
"you are not even guaranteed to have them in chronological order, specially as more and more polls come out" : YES I am guaranteed, because you will still add polls in the right order. Now let me add these 9 bytes. If you don't want to sort by firm, don't click. 2A01:E35:2F0C:F510:A56D:6579:33DF:BA52 (talk) 01:17, 12 December 2014 (UTC)
No, you are NOT guaranteed, because if instead of sorting them by "Polling firm" you decide to first sort them by "Date", and only then go and sort them by "Polling firm", you will see everything messed up. Your change does not only inappropiately fullfil the function you want it to do, but it is even counterproductive to the table as it makes it wider. Btw, seeing as there is a conflict at this point you should have stopped trying to push forward your view of things in the table before discussion here is even over. Reverting changes. Please don't revert unless consensus is reached here. Impru20 (talk) 08:46, 12 December 2014 (UTC)
I have done the necessary to make it "guaranteed".
1) added hidden dates to each line, for perfect sorting by date
Now, any statistician can look at the polls and sort them by date, it works
Once again, it is necessary to distinguish between polling firms, because their different methods produce different results. For instance, Sentio gives higher results for SD, same for Sifo with MP.
2) shrunk tables by adding "unsortable" to the columns where sorting isn't useful
The tables are not large unless you use a very low resolution.
You've been completely reverting my edits 4 or 5 times, each time with a different reason. I've tried to integrate your reproaches, but you ruined this consensus. IN my humble opinion, you are not being constructive. If there are any Wikicops, I may think of bringing them here. 2A01:E35:2F0C:F510:5D2E:48D5:518:BEF7 (talk) 14:12, 14 December 2014 (UTC)
You should have asked for consensus here first instead of pushing forward your changes with an open discussion on the issue. Adding "hidden" dates to each line is problematic, as people editing the table will surely don't know what are those meant for, and may don't know how to edit those. You talked about "Science", so you should know that opinion polls does not matter individually, but actually what matters are the trends. You want to push forward a change that risks breaking a discussed and well-settled system for a reasons that is not really relevant. People takes for granted that different methods produce different results. There is even a graphical summary showing all polls together, disregarding of whether the pollster is, because they show trends, as I said.
You know, we have many people visiting this Wikipedia, not only high-resolution users. We must be as general and simple as possible so to not hinder people wishing to visit these pages.
I've been completely reverting your edits because there is an ongoing discussion on this, yet you keep, however, pushing forward your changes without seeking consensus. I should note you that consensus does not stand for a "I make my own without consulting others but taking some of their reproaches in the way I see fit" behaviour. That requires discussion, and you are not even considering my opinion when pushing forward your change. You say I'm not constructive. Surely, disrupting established consensus (See WP:CONS) and insulting others (See WP:NPA) by calling them "tyrants" or "despots" in the edit summaries is constructive, indeed. If you keep this behaviour I may be forced to consider reporting you to Wikipedia. Both for disrupting consensus and for being disrespectful. Your call. Impru20 (talk) 14:38, 14 December 2014 (UTC)
I expected this bogus talk. Formal arguments, no content.
- "they won't understand". Sure, there are so many idiots out there, except you.
- "there are low-resolution readers". Most of whom couldn't read the current table either.
- "you don't discuss". I couldn't but demonstrate by example my synthesis of our both POVs.
- "you called me X and Y". If you're not the Wikiprince here, then why do you indulge in wrecking like a spoilt brat?
Now, back to the content. I am a frequent reader of Wikipedia's opinion polling pages, and it makes a while that my eyes hit the horrid mish-mash of a polling table they are. It's not statistics, it's mishmash. What kind of conformism and stubborness has led Wiki editors to defending such a low standard?
Trends matter, but even more so within the same pattern of study. You can indeed get contradictory trends from one pollster to another. It happened during Sweden's previous legislature (2010-2014).
That's not the point anyway. My edits do not compel readers to make those subtle distinctions. It only gives them the option! Why on earth wouldn't you understand this simple measure?
Don't you see that your "standard" isn't readable? In Nordic countries (Ger., Swe., Neth.), the polling is regular. However, look at what you are doing on Opinion polling for the Spanish general election, 2015. There are 20+ pollsters and you are denying the readers' right to look at the most trustworthy trends.
I know it is an important change in terms of habits, and I could help on the Spanish page myself. But my edits are firmly grounded in science. 2A01:E35:2F0C:F510:5D2E:48D5:518:BEF7 (talk) 15:23, 14 December 2014 (UTC)
First thing you say, to disqualify my arguments. I may disagree with you, but I have not insulted you. I remind you again about WP:CONS and WP:NPA. Just in case those also seem "bogus" to you.
"If you're not the Wikiprince here, then why do you indulge in wrecking like a spoilt brat?". I'm not even aswering this. I commend you to call "the Wikipedia cops" and explain them this behaviour of yours. Mostly because if you still keep on this aggressive manner, it'll be me the one who reports you. I'm not a "Wikiprince" (the same that you are not who to talk in the name of others or in the name of science (and I'll elaborate on this later)), but neither I nor anyone else do have to stand for this child-like behaviour of yours. I don't know if my arguments are bogus, but the quality of yours reveals itself when you have to resort to insults to defend your position.
"You can indeed get contradictory trends from one pollster to another". Precisely. That is why you must get the general picture. Taking individual pollsters without considering others is not what we would call scientific. If party A has an upward trend according to one pollster in, let's say, May 2014, but according to all other pollsters it has a downward trend at that same date, what is the point of separating them? I'm not saying it would not be a fun exercise, but it just hampers people to get the general picture, to no avail.
Now, responding to personal attacks:
If it is 'me that am denying the readers' right to look at the most trustworthy trends, how is it that no one actually has discussed this issue before? If this was a so dire of a problem, I'm sure people would have widely complained, and a consensus on the matter would have already arose. The issue is that you try to mask a own desires as if it was a problem for people in Wikipedia in general. No one else has put forward this view you are talking about. Don't speak for "the readers" as if you were a representative of them, as no other reader has complained on this issue. Speak by yourself, as what you intend is to break established consensus just to please your own desires (you do recognize yourself that it is matter of you 'not liking' the "current standard". I surely do understand the issue.
But I should note you that I would have gladly discussed a way to solve this issue with you should you have a made a more polite approach. Frankly, as things are now, with you trying to push forward your changes in the middle of an open discussing and with you now resorting to personal attacks, I won't discuss this issue with you unless it actually becomes a real issue for people. Fortunately, for people more polite than you.
By the way, my "standard" is so unreadable that it has even been used by others in their opinion polling articles, such as Opinion polling for the next Danish general election, Polish parliamentary election, 2015#Opinion polls or Norwegian parliamentary election, 2017#Opinion polling, to name a few. Cheers. Impru20 (talk) 16:13, 14 December 2014 (UTC)
Only undue criticism should be avoided.
Take, for instance, this contradiction of yours. You compelled me to discuss, for what I naively believed was justification, and despite you have nothing to respond to, you close the discussion "until other people come up". I deeply regret to have called you a Wikityrant, because it has become a performative statement: your concept of discussion is truly Soviet-style.
I notice, however, that nobody reverted my edits save you. So it is more of a real issue for you alone. Don't get angry, but the truth is that no other reader complained of this issue.
How outrageous is it that people still fail at statistics.
Anyone could understand it. Here's the explaination. Cheap pollsters, or partisan ones, increase the statistical noise of the general trend. That is, if pollsters ABC and XYZ are low-quality pollsters using a sample of just 600 people, in a country like Spain that's a 4% margin of error in 95% of cases, which is huge in a country with many small parties. Whatever meaning are given to the results, they are less trustworthy than they would be with 2,500 responses.
Same is true if the cheap pollster does not weigh its sample by category and previous vote, the basics of serious polling, that is.
But even if every polling company were independant, serious, and using the same sample size, we can still expect a large structural difference in results because of the methodology. Some pollsters ask: What would you vote if the election were next Sunday?, while others ask Who do you expect to vote for next year?. Some pollsters give a set of responses, some don't, and some suggest the names of minor parties only if the respondant says "other". Not to mention the freaks who add up the "don't knows" and the "others" in their press outlets.
So if you have a set of pollsters, and a large majority has one particular type of questionnaire, then the general trend will be skewed. And it doesn't mean anything to add up pears and apples.
So much for this! 2A01:E35:2F0C:F510:5D2E:48D5:518:BEF7 (talk) 18:05, 14 December 2014 (UTC)
"Only undue criticism should be avoided." And insults, too. Please, see WP:NPA. Third time I link it to you, mind you.
No, you are clearly misinterpreting and manipulating my words. I compelled you to discuss because I was against that change of yours, unless until it was discussed, and a consensus could be reached. Then, after opening the discussion but before it was closed, you kept changing the article despite the discussion still being open, something which I noted you about both here in the talk and in the edit summaries. Then you again changed the page and even went to insult me calling me a "tyrant" and a "despot" just because I was not unconditionally accepting your changes; such behaviour being continued on this same discussion. I'm now saying that I would wait until "until other people come up", because it is obvious that this is a breaking-consensus change fruit of a personal desire that is being brought forward in an unpolite and direspectful manner. No contradiction on this point.
I notice, however, that nobody reverted my edits save you." Notice how no one else reverted your edits because I reverted them before anyone else did it. However, if you want to look at it at this how, on these grounds, notice how nobody reverted my reverts. Neither did anybody push forward or defend your changes. Difference, you'll see, is that the current version of the polling table has been agreed upon/accepted by people for much time in plenty of articles. Your changes are the ones disputed, not the current version of the table.
But you know, now that you want to put the example of Spain, and aswering to your entire explanation of statistics: as a frequent reader of the Spanish opinion polling article as you say, you maybe should know that there are Spanish pollsters that have changed their methodology as much as three times since the 2011 election (i.e. Metroscopia, but others such as Celeste-Tel, Invymark, or the CIS itself (since CIS methodologies change each time a new government takes office) have also done it). What does that mean? That means that there are drastic changes in opinion polls of the same pollster because, at some point, they felt like changing the data treatment methods they used to apply until that point. That means that we are not even assured that opinion polls of the same pollster will even be consistent with other polls of the same pollster. Keeping this in mind, and on the basis of your argument, what is the point of sorting polls by pollster when we are not even assured they will always keep using the same methodology? Please, elaborate.
Opinion polls, at least in Spain, are more of a political weapon rather than a true 'labour of science' (specially now, with Podemos' surge, with all pollsters doing mostly 'wild guessing' because they don't know how to treat respondents' data), to the point there is no legal regulation on opinion polls aside that done for opinion polls during electoral campaign (which spans just the previous 15 days before election day).
Indeed, so much for this. Impru20 (talk) 19:11, 14 December 2014 (UTC)
Granted, noone had enough time to judge my edits. 18 hours in total. So let's test the case. Let my proposal be implemented for a few polls, after which we'll finally decide which one of us must get councelling.
I don't think it will be me, because your argument is weak. What is my edit about? Giving readers the option to sort, not by date first, but by pollster first and date second. Why? Because it increases readability of the particular trends - just to those who want it. And what is your objection? That it is not enough of an enhancement! But then why don't you, constructively, suggest something better? Before you invent it, let everyone see my try.
2A01:E35:2F0C:F510:5D2E:48D5:518:BEF7 (talk) 20:54, 14 December 2014 (UTC)
And once again, your reasoning starts by attacking the alleged quality of my arguments instead of refuting them.
You say my argument is 'weak' when it was you who resorted to insult several times. How peculiar. Also peculiar is that, my argument being 'weak' according to you, you have quickly changed your viewpoint as a result of it. Now "no one had enough time to judge your edits". Before it was that "no one had reverted your edits". You have even proposed a new course of action as a result of it. But yes, I guess that argument of mine was weak.
Also, my argument was 'weak' but you can't propose a solution for the problem of the same pollster using different methodologies. How does fit that in your proposal? I asked you specifically on this and you have not answered me.
Going to the point in question. I don't suggest 'something better' because I don't see a problem with the current edition. It is not me the one who should propose something better, but rather you. If you see that your current proposal does not reach consensus, you can either call for more people to discuss the issue, or to modify your proposal so that others may actually like it. You have done nothing of this.
And the course of action you propose is ridiculous. Articles are not 'test grounds' for experiments. Furthermore, with the current setting it is impossible to implement your changes "just for a few polls" without splitting the table (something even more horrible, if possible). Seeing your behaviour, I fear that may be just a trick to just push forward your changes without consensus. You can publish your table version in this talk page, so that people visiting it does see it and feels motivated to enter the discussion.
As it stands, the current edition is the one to which consensus agrees. There is not consensus for your change, mainly because there are only two people discussing it, and each one defends one different viewpoint. I could have agreed to a compromise, but your manners and behaviour clearly put me strongly against it, so a new consensus would require more people entering the discussion (something which, by the way, would strengthen any possible consensus, so it actually doesn't have to be a bad thing). Until then, it is all but logical that the article should show the edition with the latest consensus. Cheers. Impru20 (talk) 21:52, 14 December 2014 (UTC)

While Mr/Ms 2A01:E35:2F0C:F510:5D2E:48D5:518:BEF7 has not exactly been overly cordial on this talk page, I agree that allowing readers to sort the table by polling firm is a definite improvement for the reasons stated. The technical issue with the dates was a valid concern, but that was addressed in the attempt to reconcile the disagreement. I've read the discussion above and can't see that any other issues of real consequence have been raised, so I suggest the sortable table is reinstated. Väsk (talk) 07:51, 15 December 2014 (UTC)

Title

The word "general" here strictly speaking refers to elections taking place across a variety of representative bodies. So the text of Swedish general election, 2014 correctly reads "General elections were held in Sweden on 14 September 2014 to elect the Riksdag, all 21 county councils and 290 municipal assemblies."

However, I understand this election is solely to elect the Riksdag. Therefore it would be much better to title it "Swedish parliamentary election, 2015". Harfarhs (talk) 19:01, 18 December 2014 (UTC)

Left Party

"The party did not support the Löfven Cabinet as they was rejected to participate in his cabinet following the 2014 general election but supported his budget that was voted down on 3 December 2014."

The English here is so poor as to obscure the meaning completely. Did Löfven reject the Left Party's participation in government, or did the Left Party reject him? Was there an arrangement for it to vote in support of the Löfven Cabinet from outside? Harfarhs (talk) 19:27, 18 December 2014 (UTC)

It's not only bad english, it's incorrect. They did support his cabinet. 213.100.108.117 (talk) 22:26, 23 December 2014 (UTC)

Suggesting to merge the raw polling list and the coalition counts

Hello,

Despite democracy still exists in Sweden, it must be signalled that, after the anti-SD pact (with V also left out), coalition counts are not going to matter for the next 39 months. According to the pact, The Allians will take over in 2018 unless De Rödgröna reaches an absolue majority, which is a very unlikely scenario. The inter-coalition pact is the main reason making the second count of little relevance. But there are aesthetical reasons too; we should not need to read and edit two tables, especially considering the importance of opinion polls in such a struggling democracy.

This is why I suggest a merger, which you will find here.

Parties

Polls can be sorted by date first or by polling firm first (and date second), by clicking on the related column header.

Date Polling firm S M SD MP C V FP KD Fi Others Lead De rödgröna
S + MP + V
Alliansen
M + C + FP + KD
Lead
12–26 Jan 2015 Ipsos 34.3 23.2 13.1 6.3 5.0 5.8 6.2 3.8 1.5 0.8 11.2 46.4 38.2 8.2
12–22 Jan 2015 Sifo 29.7 24.0 13.3 7.4 5.8 6.9 5.1 3.7 3.1 1.1 5.7 44.0 34.9 9.1
22 Dec–18 Jan 2015 Novus 30.0 24.1 16.5 6.4 6.0 5.3 4.1 4.5 2.2 0.9 5.9 41.7 38.7 3.0
7–16 Jan 2015 Inizio 33.0 24.8 13.6 5.5 6.1 5.4 3.5 4.9 3.4 8.2 43.9 35.8 8.1
8–13 Jan 2015 Sentio 29.7 23.7 15.8 5.6 5.4 7.0 4.8 3.9 4.1 6.0 42.3 33.9 8.4
7–8 Jan 2015 Sifo 33.3 23.1 13.4 6.7 6.7 5.6 5.0 3.4 1.7 1.1 10.2 45.6 34.8 10.8
30 Dec–7 Jan 2015 Demoskop 32.3 24.4 14.4 6.6 5.3 5.3 4.8 4.1 2.1 0.7 7.9 44.2 38.6 5.6
19–27 Dec 2014 YouGov 30.6 23.2 16.9 5.7 5.7 6.0 3.7 5.5 2.0 0.7 7.4 42.3 34.4 7.9
8–18 Dec 2014 Sifo 31.0 24.5 12.9 7.4 6.1 5.9 5.7 3.7 2.4 0.4 6.5 44.3 36.3 8.0
11–17 Dec 2014 Inizio 30.5 27.9 13.2 6.1 6.6 4.1 4.2 4.2 3.2 2.6 40.7 42.9 2.2
8–15 Dec 2014 Ipsos 32.0 26.3 12.5 6.7 5.1 5.4 5.6 3.9 1.5 1.0 5.7 44.1 36.9 7.2
3–14 Dec 2014 Novus 32.0 23.7 16.0 6.0 6.6 5.3 4.3 3.9 1.1 1.1 8.3 43.3 34.6 8.7
4–10 Dec 2014 Sentio 29.7 24.5 17.5 7.2 5.5 5.1 3.7 3.9 2.9 5.2 42.0 30.0 12.0
4–10 Dec 2014 Inizio 31.0 26.9 14.0 5.8 7.4 4.2 4.8 4.3 1.6 4.1 41.1 43.4 2.3
4–5 Dec 2014 YouGov 29.6 23.1 17.7 6.9 6.1 4.7 5.0 4.2 2.5 0.7 6.5 41.2 38.0 3.2
25 Nov–3 Dec 2014 Demoskop 27.3 26.1 13.5 7.9 6.1 6.4 5.5 4.0 2.3 0.8 1.2 41.6 41.7 0.1
17 Nov–2 Dec 2014 Inizio 29.8 24.5 12.0 7.5 6.3 6.2 6.1 3.7 2.9 1.0 5.3 43.5 36.9 6.6
25–30 Nov 2014 Sentio 27.4 23.9 16.4 7.5 6.2 7.4 6.2 2.3 2.7 3.5 42.3 36.3 6.0
29 Oct–25 Nov 2014 SCB 32.2 24.2 12.4 7.2 6.0 5.7 5.3 3.7 3.2 8.0 45.1 35.5 9.6
13–24 Nov 2014 Ipsos 28.5 26.5 12.5 7.4 6.2 6.2 6.4 2.8 2.4 1.0 2.0 42.1 39.1 3.0
4–24 Nov 2014 SKOP 30.7 25.3 11.8 6.7 6.8 6.4 5.3 4.2 1.9 0.8 5.4 43.8 41.6 2.2
10–20 Nov 2014 Sifo 28.4 25.6 12.1 8.4 6.1 5.9 5.5 3.9 2.8 1.3 2.8 42.7 37.2 5.5
13 Oct–9 Nov 2014 Novus 28.6 25.8 12.3 7.9 7.1 6.3 5.2 4.0 2.2 0.6 2.8 42.8 42.1 0.7
28 Oct–5 Nov 2014 Demoskop 28.3 25.1 13.8 7.0 6.2 7.1 4.8 4.2 2.6 1.0 3.2 42.4 40.3 2.1
15–30 Oct 2014 Inizio 30.6 24.0 12.9 7.3 5.9 5.9 5.6 4.2 3.7 6.6 43.8 39.7 4.1
9–21 Oct 2014 Ipsos 30.7 24.1 12.5 7.9 5.8 6.2 5.5 3.9 2.5 1.0 6.6 44.8 35.4 9.4
6–16 Oct 2014 Sifo 26.6 24.7 12.4 10.0 6.1 7.1 4.9 4.5 3.0 0.9 1.9 43.7 40.2 3.5
15 Sep–12 Oct 2014 Novus 30.3 24.6 13.2 7.1 5.7 5.7 5.8 4.1 2.9 0.6 5.7 43.1 40.2 2.9
2–8 Oct 2014 Sentio 27.3 25.1 16.7 6.9 5.1 5.2 5.1 2.8 2.9 2.9 2.2 39.4 35.3 4.1
30 Sep–7 Oct 2014 Demoskop 30.3 24.1 13.3 8.1 5.8 6.0 5.5 3.9 2.1 0.8 6.2 44.4 35.4 9.0
17–24 Sep 2014 Ipsos 30.7 24.0 12.4 7.1 6.1 6.2 5.6 4.2 2.9 0.8 6.7 44.0 39.9 4.1
17–21 Sep 2014 Demoskop 31.0 24.6 12.1 6.9 6.8 5.5 5.5 4.0 2.8 0.9 6.4 43.4 40.9 2.5
14 Sep 2014 2014 General Election 31.0 23.3 12.9 6.9 6.1 5.7 5.4 4.6 3.1 1.0 7.7 43.6 39.4 4.2

It's up to you. 82.240.207.81 (talk) 03:57, 3 February 2015 (UTC)

The premise is wrong. The agreement means that the larger of the two blocs will be allowed to form a government. No absolute majority is neccessary. Sjö (talk) 07:29, 3 February 2015 (UTC)
Here is an excerpt from what I've read
Quote: "But the deal also guarantees that, if neither bloc wins a majority in the next scheduled election in 2018, the Alliance will be allowed to lead the next government, even if it comes second to the Social Democrats, who will not vote against its budgets." Reuters (link).
Is it an error from Reuters or did things evolve since then? The article 2014 Swedish Cabinet crisis does not specify anything. 2A01:E35:2F0C:F510:347E:C6E3:3DBC:7C88 (talk) 20:26, 3 February 2015 (UTC)
I'm beginning to despair about U.S. media getting things right about European politics. Yes, Reuters is wrong, very wrong. Here's the text of the agreement and here's a secondary source (both in Swedish). They both say very clearly that the candidate for prime minister representing the largest party constellation will be allowed to pass the vote that's necessary to be confirmed as prime minister. I haven't been able to find any sources in English. Sjö (talk) 21:13, 3 February 2015 (UTC)
I'm against this. The second count is useful because it shows coalitions as they would be represented in parliament (since parties polling less than 4% would be out). Furthermore, what you say doesn't make sense, and I don't think that both coalitions agreed to that. I mean, how could the Alliance rule if the Red-Greens outpoll them by, let's say, 10 points? I read the agreement and, as Sjö says, it only keeps a compromise that the lesser bloc will allow the larger one to rule, even if no absolute majority is obtained. Impru20 (talk) 13:59, 3 February 2015 (UTC)
— 1) on the second count. The big picture is that the 7 parties from the 2 coalitions have been constantly present in Parliament since 1994 (see here). One of the three smaller parties may be just below the threshold, between 3 and 4 percent, but it doesn't mean much. Calculating this borders speculation (see WP:CRYSTAL). I suggest to add an asterisk * when the result differs if we take threshold into account, and ** if the affected coalition would also lose its plurality over the other.
It is still possible to add three new columns anyway, but you can see that they are redundant over half of the time.
— 2) on the uselessness of the table because of coalition pact.
I have stated my source to Sjö. I am be waiting for a clarification.
2A01:E35:2F0C:F510:347E:C6E3:3DBC:7C88 (talk) 20:26, 3 February 2015 (UTC)
What? No. 4% is the legal barrier established by the Swedish electoral system. I don't know how can it somehow be WP:CRYSTAL, when it is currently legally recognized as such. You saying that "The big picture is that the 7 parties from the 2 coalitions have been constantly present in Parliament since 1994" is what would mean nothing, since nothing assures that, just because they have been present since 1994, they can't poll below the 4% barrier and get out. That would be WP:CRYSTAL. It's much easier for the reader to show the table the way it is shown now, because the only thing it shows is the possibility (predicted (or not) by polls, and allowed by the Swedish electoral system) of coalitions becoming smaller because of one of its parties not polling above the 4% threshold, getting 0 seats. So, let's say, if the two coalitions poll 44% and 38%, but the 38% one includes a parties with a 3% level, the 38% would effectively mean nothing, since the real vote percentage entering parliament would be 35%. What it does is only showing what the poll indeed indicate, again, as allowed by the Swedish electoral system.
About the link you posted to Sjö, it is definitely a Reuters error. You are invited to read the actual agreement (you would have to traduce it, but believe me, it is not difficult to understand) and says nothing like that, but just that "the Prime Minister candidate who gathers the support of the largest bloc will be allowed". It says nothing that the Alliance will automatically rule after the 2018 election. Impru20 (talk) 21:14, 3 February 2015 (UTC)
Impru20 summed up what I was about to say about the 4% barrier and WP:CRYSTAL. Sjö (talk) 21:16, 3 February 2015 (UTC)
Thanks for the links. I appreciate that Impru20 eventually accepted to add the option of sorting by polling firm. I may do the same for other countries that need it most, especially Spain and to a lesser extent Italy. I did it for The Netherlands and it is even clearer than on Sweden. Ideally we'd use templates asking for a date format, producing the coalition counts and the first party highlighting.
As for the merger, I understand your arguments. When I'm back in a few weeks time I'll try to add the 3 columns. The least space-consuming way would be to divide each coalition in a raw count column and a legal count column.
2A01:E35:2F0C:F510:347E:C6E3:3DBC:7C88 (talk) 01:27, 4 February 2015 (UTC)
While the OP was a bit mistaken when it came to the premise, I agree that it would be better to have the bloc count along with the parties. I would also remove the "Lead" column that shows how much the Social Democrats leads the Moderates, as that is an essentially meaningless statistic of no consequence other than prestige.
The column on the current page with blocs that only include parties "above threshold" should be removed as it is unscientific and misleading. Swedish pollsters are always including all four Alliance parties in the bloc count, even if one of them have momentarily slipped below the four percent threshold, and they do this for several good reasons. As it currently stands, a partly apparently gets excluded from our bloc count when it gets 3.9 percent, but is included when it gets 4.0 percent. Such changes are too small to be within the margin of error, so to make decisions on representation based on that is unscientific and makes the electorate seem more erratic than it actually is. In addition, it is perfectly normal for some of the smaller to slip below 4 percent in individual polls between the elections when their exposure is limited. When the election comes around right-of-centre voters will align themselves with one of these parties to ensure that they don't slip out of the riksdag, due to a phenomenon known as kamrat 4 procent. If a party consistently slips below 2-3 percent for an extended period of time, then I imagine the reporting would change, but until that happens Wikipedia shouldn't be making up its own standards. Väsk (talk) 07:03, 4 February 2015 (UTC)

Poll graph update

Does anyone know how to update the graph of the polls? Its very overdue. I would do it, but I dont know what program or method was used.--Metallurgist (talk) 16:16, 16 August 2015 (UTC)

How did you generated the graphical summary?

Hey all. Sorry for asking this Question which is not relevant to the Article, but I have been trying to generate a graph for Cypriot Elections 2018 at the Greek version of Wikipedia, and it is a little bit confusion. How can someone create such a graph? Is there a tool within the WP or should someone create it in an extra-WP page and upload the picture afterwards? Thank you. Τζερόνυμο (talk) 09:28, 22 November 2017 (UTC)

Left Party (2)

"Following the budgetary agreement, the Left Party is what tips the left-of-centre minority into a larger minority than the Alliance."

Yes, but further down in the section '2014 budget crisis' we read:

"On 9 October 2015, following the Christian Democrats' departure from the agreement, the December 2014 agreement was dissolved."

Surely "dissolved" can only indicate that its provisions no longer apply. In which case, what is the current situation? Has there been a further budgetary agreement since October 2015? Harfarhs (talk) 02:23, 21 July 2018 (UTC)

Number of parties participating

Perhaps someone can make a mention of the fact that it was possible to cast a vote on almost 80 different parties nationwide.[2] Click on one of the counties in the left navigation bar, and all of the parties and their eventual party-specific ballot papers are shown. Talk/♥фĩłдωəß♥\Work 19:12, 10 September 2018 (UTC)