Talk:Opinion polling for the 2022 Swedish general election

Sentio note

Other parties received 0.3%; unclear about the remaining 1.3% Mélencron (talk) 17:37, 12 October 2018 (UTC)

Sifo

https://twitter.com/StatistiskO/status/1063472516151283712 Mélencron (talk) 20:52, 16 November 2018 (UTC)

Ipsos

https://twitter.com/StatistiskO/status/1065679906762297344 Mélencron (talk) 19:36, 22 November 2018 (UTC)

Rename

This page should be renamed "opinion polling for the next Swedish general election" (replacing "2022" with "next") in line with other similar articles; also the situation in the swedish parliament seems particularly fragile, adding weight to the argument that an election might come sooner, perhaps not within months (although that does seem possible) but say in 2020 or 2021 (who knows). Speed74 (talk) 23:59, 28 November 2018 (UTC)

The title is chosen to match that of the election article. Mélencron (talk) 01:29, 29 November 2018 (UTC)

Further, elections are fixed in Sweden: an election is going to be held in 2022 no matter what. There could be an early election indeed if called, but elected MPs would just serve out what remained of the four-year term starting in 2018, until 2022. Impru20^talk 09:06, 29 November 2018 (UTC)

Ipsos

https://twitter.com/StatistiskO/status/1075090650578722818 Mélencron (talk) 20:51, 18 December 2018 (UTC)

Block politics

Since the last election, many analysts have declared that the "blockpolitik" is probably over.

Löfven himself has stated so. He now seeks a coalition with C and L instead of V, while, on the other hand, the M+KD budget was passed with SD support.

In most of the local councils (see the swedish article), the traditional blocks were unable to find a majority, and have formed cross-alliance coalitions, or minority governments.

This is why I suggest that we should adapt the table to the current situation.

There are now four possible alliances:

• traditional blocks, without SD
  • Red-Green (S, MP, V)
  • Alliance (M, C, KD, L)
• new formations, without V
  • "social-liberal*" (S, C, L, MP)
  • "konservativ*" (SD, M, KD)

(*) for want of a consensual term

Currently, six columns on the right are being used, four for coalition sums, and two for the lead. We could spare ourselves from calculating the sums twice, all the more so since they are very often the same. In cases where the threshold isn't met by all parties, it could be signalled with an efn template(^[a]) or an obelisk (†), that one or two parties don't meet the threshold.

Before

Polling firm	Fieldwork date	Sample size	S	M	SD	C	V	KD	L	MP	Others	Lead		Total		Lead		At/above 4%		Lead
														RG+V	A			RG+V	A
Ipsos[1]	5–16 Dec 2018	1,537	32	18	18	7	8	8	4	4	1	14		44	37	7		44	37	7
Novus[2]	12 Nov–16 Dec 2018	5,414	30.0	18.3	19.8	8.6	7.7	6.5	3.9	4.0	1.2	10.2		41.7	37.3	4.4		41.7	33.4	8.3
Sifo[3]	3–13 Dec 2018	8,658	29.7	19.0	20.0	7.7	7.7	6.6	4.0	4.2	1.1	9.7		41.6	37.3	4.3		41.6	37.3	4.3
2018 election	9 Sep 2018	–	28.3	19.8	17.5	8.6	8.0	6.3	5.5	4.4	1.5	8.4		40.7	40.3	0.4		40.7	40.3	0.4

After (suggestion)

Polling firm	Fieldwork date	Sample size	S	M	SD	C	V	KD	L	MP	Others	Lead		Total
														RG (S+MP+V)	A (M+C+L+KD)	S+C+L+MP	M+KD+SD
Ipsos[1]	5–16 Dec 2018	1,537	32.0	18.0	18.0	7.0	8.0	8.0	4.0	4.0	1.0	14.0		44.0	37.0	47.0	44.0
Novus[2]	12 Nov–16 Dec 2018	5,414	30.0	18.3	19.8	8.6	7.7	6.5	3.9	4.0	1.2	10.2		41.7	37.3	46.5	44.6
Sifo[3]	3–13 Dec 2018	8,658	29.7	19.0	20.0	7.7	7.7	6.6	4.0	4.2	1.1	9.7		41.6	37.3	45.6	45.6
2018 election	9 Sep 2018	–	28.3	19.8	17.5	8.6	8.0	6.3	5.5	4.4	1.5	8.4		40.7	40.2	46.8	43.6

• The coalition cell is shaded when it is predicted to have a majority of seats (none in this example)
• The coalition cell is written in bold font when it can form a minority government on the condition that one or two outside parties abstain from voting no (SD for Allians, V for Socialliberal, C and L for the either a left-wing or a right-wing coalition).

Remarks? Kahlores (talk) 12:32, 24 December 2018 (UTC)

Looks a bit overly complicated to me and possibly what's known as "original research". I would support simply removing the political bloc columns altogether, given that there are no alliances really set in stone at this point. Speed74 (talk) 12:54, 24 December 2018 (UTC)

Agree with Speed74; either we leave the current blocs as they are, or we just remove the bloc columns altogether. Otherwise, we would have to report on every conceivable combination of parties (which is not actually practical) to prevent falling into SYNTH and, ultimately, into original research territory. Impru20^talk 13:37, 24 December 2018 (UTC)

I echo the above. All of the major pollsters continue to show results only for the two traditional blocs, so I don't think there's any reason to indiscriminately list speculative combinations of parties which have yet to prove capable of forming a government. Mélencron (talk) 13:53, 24 December 2018 (UTC)

Thanks for your feedback.

In and of itself, tallies are not always O.R.. For instance, during 45 years, the blocks were crucial to the Swedish political life, even more so than the individual parties, as it was clear that a stronger block ahead would mean a different governing coalition. Tallying up is "routine calculation" (allowed by WP:CALC) for all the Swedish and international media. Which is also why we still do it on the Denmark page.

I admit that my calculations for "social-liberal" and "conservative" coalition are original research . . . for the time being. For what if the January vote brings such a never-heard-of government? and the media start to calculate the tallies? Are we going to have one, two, three, four coalition tallies? Two like on the Italy page?

As a sidenote, there's something we can do now on the Swedish table. We would be wise to gain some space, and replace the "above-4%" recalculation by some kind of footnote (as in my example), especially given the fact that the two traditional blocks are less relevant.

Kahlores (talk) 14:23, 24 December 2018 (UTC)

Indeed, tallies are not always OR, but they are not when sources do back these up and are consistent in doing so. While calculations are allowed by CALC, it is the picking up of which data to aggregate what constitutes the true issue here. For instance, S+SD could also be a mathematically possible coalition, yet it is so implausible that it is not even remotely suggested. As you point out, if the media start to calculate different tallies at some point, we can use these. But until then, better not entering into speculation ourselves. Let's wait and see. Impru20^talk 14:29, 24 December 2018 (UTC)

Of course we'll wait and see, this is 24 December! Kahlores (talk) 16:07, 24 December 2018 (UTC)

EP note

https://twitter.com/StatistiskO/status/1100785279001157633 Mélencron (talk) 18:03, 27 February 2019 (UTC)

Requested move 7 February 2020

The following is a closed discussion of a requested move. Please do not modify it. Subsequent comments should be made in a new section on the talk page. Editors desiring to contest the closing decision should consider a move review after discussing it on the closer's talk page. No further edits should be made to this discussion.

The result of the move request was: Speedy moved (non-admin closure) – Thjarkur (talk) 13:34, 7 February 2020 (UTC)

---

Wikipedia:Opinion polling for the 2022 Swedish general election → Opinion polling for the 2022 Swedish general election – When this page was recently moved, "Wikipedia" was added in the title, using "Wikipedia" in titles should only be done in Project namespaces Lochglasgowstrathyre (talk) 12:34, 7 February 2020 (UTC)

• Support - Move and fast. Primergrey (talk) 13:05, 7 February 2020 (UTC)

---

The above discussion is preserved as an archive of a requested move. Please do not modify it. Subsequent comments should be made in a new section on this talk page or in a move review. No further edits should be made to this section.

Added FI, AFS, MED and PP

So I added four other parties that have been shows in SENTIO polls. I though it would be a good idea to put as much information as possible on this page. — Preceding unsigned comment added by Xenenex1 (talk • contribs) 12:06, 17 February 2020 (UTC)

This probably goes against WP:UNDUE. The parties are hardly relevant as they do not even reach 1% (threshold is at 4%). Further, they only get polling results in Sentio polls, and not even in all of them, thus leaving a vast majority of the columns as "?". Hardly useful or needed, it is mostly a waste of wiki markup size than anything else. Impru20^talk 12:11, 17 February 2020 (UTC)

Other parties.

Should we make a expandable row with FI, AFS, MED and PP so they are still there but not in the way? — Preceding unsigned comment added by Xenenex1 (talk • contribs) 12:26, 17 February 2020 (UTC)

How to make graphical summaries?

Hello everyone. It looks like the user who used to make the graphical summaries (Avopeas) is not active anymore. Could someone explain how to make these graphs? Are they made in R, Python...? Thanks! --Spaastm (talk) 11:49, 10 December 2020 (UTC)

Here is a link that might be useful, I haven't tried it out though. Personally, I just use Excel for polling graphs, but I think they could be better. --Gbuvn (talk) 11:59, 10 December 2020 (UTC)

new editable chart.

So i had an idea where we could have a graph that everyone can update when a new poll comes out.

Swedish polls

Add events to the table

This article would benefit from the adding of events in the table like it is in several other related articles such as Opinion polling for the next United Kingdom general election. Adding events such as major political happenings and resignations and so on would help explain the changes in opinion Lochglasgowstrathyre (talk) 19:22, 26 August 2021 (UTC)

Opinion polling two groups

Since Juni 2021 is the situation (Opinionsundersökningar inför riksdagsvalet i Sverige 2022)

First group Höger-oppositions (M, KD, SD, L) Example Januari 2022 48,18 %

Seconde group Regeringen med stödpartier (S, MP, C, V) Example Januari 50,14 %

How could we "Vote Share" show? Wname1 (talk) 07:22, 16 February 2022 (UTC)

Sentio hoax

I wonder why the so-called "surveys" from Sentio are still listed here. This company has no idea about representative polls, which you have proven several times in Sweden. Remember the 2018 election when Sentio saw the right-wing extremists from SD as the "strongest party" (as the only institute). They were so miles away from the later result that you can not call this company "opinion research company". They only do research online and they have no idea how to make their data representative. Why are the polls still listed here? --Petruz (talk) 23:06, 4 March 2022 (UTC)

Blocs

During the last couple of months two blocs have formed in Sweden. One to the left consisting of S, C, V and MP and one to the right consisting of M, SD, KD and L. They’ve put forward National budgets together and have the same candidate for PM within the blocs. Swedish media has for slightly more than a year mentioned these bloc’s in their reporting and the public recognize these two blocs. Therefore i think it should be in the polling. ShutTheFurgessonUp (talk) 09:55, 24 June 2022 (UTC)

What you describe is the normal parliamentary practice of any parliamentary country, i.e. a government relying on coalition/confidence and supply support from other parties, and the opposition working in tandem (most of the time) to oppose the government. That is not noteworthy. Unlike what happened in Swedish politics between 2004 and 2019 (with the Alliance (Sweden) and the Red-Greens (Sweden)), there is no longer two officially established alliances, but rather two loose coalitions of parties with common interests that, nonetheless, may choose to work in a different way should parliamentary maths see a different result. In fact, it is interesting how you unilaterally cherry-picked which parties were shown together, despite sources themselves giving different results: for example, in the latest Demoskop poll you depicted a "Blue" bloc adding up to 48.9% of the vote, and a "Red" bloc adding up to 50.1%, despite the source itself clearly differentiating between S and V+MP+C (i.e. the current government and government external support, as opposed to the whole opposition united). That's outright WP:SYNTH and is not allowed.

As a result of these loose alliances, it's best not to present "bloc" results. Readers are intelligent enough to make any calculations themselves if they wish to see how the different options of government add up. Impru20^talk 12:25, 24 June 2022 (UTC)

I think it would be very helpful to show which bloc is leading in the polls in a column. First and foremost. There are no “loose connections” between these parties. S, V and MP call their parties and C the “red-greens” while M, SD, KD and L call their side “de blågula ~ the blue-yellow”. In the last debate The leaders of all four parties to the right said that they were a bloc.

I don’t see how something that is defining for the election in September is “interesting cherrypicking”.

The Wikipedia side for the Danish election polls has blocs and they have more or less the same connections between parties as we have in Sweden. ShutTheFurgessonUp (talk) 16:12, 24 June 2022 (UTC)

And does C regard itself "Red-Green" too, despite their vocal rejection to align itself with V if not strictly needed? Or should we depict blocs as depicted by S, V and MP? Would C reject allying itself with M, L and KD should the four of them reach a majority without SD? Sources do not represent blocs the way you do, it's you the one unilaterally cherry-picking which parties correspond to each bloc and making an presumption on how they will work in the future for "helpful purposes". Blocs made sense when they had signed agreements in the past (again, Alliance (Sweden) and Red-Greens (Sweden)), not now that alliances are more open due to C's ad hoc alliance with S and M+L+KD having broken their previous rejection to ally themselves with SD. To complicate things further, all parties in each bloc were more-or-less assured in the past to reach the 4% threshold; now, we have MP and L persistently hovering below that line. Adding up these votes that wouldn't actually count for majorities would be more misleading than helpful. Finally, what other countries do is their own stuff; politics for each country work differently, and that's perfectly true between Sweden and Denmark (where sources do specifically depict bloc results in addition to party results, so that's not SYNTH in that case: sources do actually give those numbers on their own!). Let us better not engage in original research here nor pretend than readers are not smart enough to make their own calculations with the provided numbers if they wish so that they need to be led some specific way. Impru20^talk 17:42, 24 June 2022 (UTC)

So your problem is the wording? The fact I wrote “blocs”? There are two “bases” or “sides” if you will in Swedish politics today. One consisting of S-C-V-MP and one consisting of M-SD-KD-L.

Would it be appropriate if the titles were changed to “Gov + support” vs “Opposition”? ShutTheFurgessonUp (talk) 18:17, 24 June 2022 (UTC)

No, the problem is the WP:SYNTH nature of your edits. With each new comment you only keep confirming this is your own doing, rather than the sources's. Impru20^talk 21:33, 24 June 2022 (UTC)

Would it be appropriate to add “Gov + support” vs “Opposition”?

It’s insincere to accuse me of doing something I’m not. If you view Swedish state media as a reliable source then you’d think there were two blocs in Swedish politics. ShutTheFurgessonUp (talk) 06:34, 25 June 2022 (UTC)

Sources do not repict "Government" vs. "Opposition" (also, you seem intent to depict within the "Government" bloc parties that are not inside the government), so again no, it's not appropiate. When speaking of sources I'm referrting to opinion poll sources. Danish opinion polls do show aggregated results by bloc. Swedish opinion polls used to do it in the past when blocs were official, they don't do it now. Impru20^talk 08:24, 25 June 2022 (UTC)

Demoskop, Sentio, Infostat and Ipsos depict gov vs opposition. I explicitly wrote “gov + support”. You seem to have missed that part, among other things. If those were reasons for it not being appropriate and you got both wrong then that means it would be appropriate to add “gov + support”, “opposition” and ”lead” columns..? ShutTheFurgessonUp (talk) 15:24, 25 June 2022 (UTC)

Demoskop shows opposition vs. S, then V+MP+C separately. Infostat requires suscription to see its content. Likewise for the latest Ipsos source, but the one for the previous Ipsos poll does not show any bloc. Same for Sentio.

"Gov + support" means "government + whoever supports the government at any time", which is something that may vary. You say "I" have missed something you say, but then you have missed the issue of the blocs you argue not being established in stone (I repeat: what if M+L+KD+C get a majority of votes without SD? What for parties not crossing the 4% threshold?). I mean, you keep insisting on your own personal view of things here, only furthening the fact that this is a mere SYNTH issue. You want to force readers into your own view of bloc politics, when alliances as of currently are not as closed as they were before. Let readers get their own conclusions rather than force them into the conclusions you want to present them. Cheers. Impru20^talk 16:51, 25 June 2022 (UTC)

It’s not my opinions that dictate who sits in government and who sits in the opposition. What I propose is adding columns that show the support of the current government + their allies aswell as a column for the current opposition parties. Why would that not be appropriate? If it changes before the election then the percentages just have to be redone. ShutTheFurgessonUp (talk) 08:48, 26 June 2022 (UTC)

It’s not my opinions that dictate who sits in government and who sits in the opposition. That's not the point of debate in this discussion, but on whether to show the aggregated sums of parties in a way that you are specifically proposing. I have told you why that is not appropiate (sources not backing such aggregations).

If it changes before the election then the percentages just have to be redone. I think this pretty much says everyhing. We should not be aggregating party results based on a personal perception, then keep changing it whether we see fit once the election comes closer. Again: readers are smart enough to make their own calculations with the numbers as they wish. Do not led them into a specific way as you see fit, nor present them with a conclusion not expressly stated by sources. Impru20^talk 08:54, 26 June 2022 (UTC)

All these sources from some of Swedens largest news sites name the gov + support side and the opposition. You’ve come with many wild claims, but this one is just false. Every reason you’ve given for it not being appropriate to add columns so far has been wrong and misleading. Swedish news talk about politics like there are two blocs. It is a fact that we have a government and an opposition. So I ask once again. Because so far it’s seems like it’s you’re personal opinion that dictates whether this should be done or not. Why would it be inappropriate to add proposed columns? ShutTheFurgessonUp (talk) 14:43, 26 June 2022 (UTC)

You are reasserting the WP:SYNTH concerns existing on this issue. It's not that you solve the issue with wild sources you cherry-pick around the web, put them together and say "hey, sources do say that". It's opinion poll sources the ones that must give the aggregated results by blocs in order for you to have a chance to justify depicting the blocs in an article on opinion polls. You cannot just come to present opinion polls in a way that sources do not present them by throwing random sources from news opinions and editorials speaking of an issue that is not related to opinion polling numbers. That's precisely what SYNTH is about. We cannot reach conclusions not stated in the sources, and your manual aggregations of parties' polling data the way you see fit falls under that category.

Do you want to explain that the general opinion in Swedish media agrees on the formation of two loose blocs ahead of the next election? Fine, describe and source it in prose at 2022 Swedish general election. This article is about opinion poll numbers; opinion poll sources do not present the numbers in the way you are attempting to present them. So please, stop wasting everyone else's time with this since the issue has been clear from the beginning, and get on with it. Thank you. Impru20^talk 15:09, 26 June 2022 (UTC)

References

Opinion poll chart

Unfortunately, the chart is becoming harder to read due to the increasing frequency of opinion poll dots and trend lines using the same colour. How can we make the chart a bit more readable? 92.236.0.141 (talk) 19:21, 2 September 2022 (UTC)

I also think it would be helpful to divide it into the two electoral groups: the Social Democrats and their allies vs. the Moderates and theirs. TFD (talk) 16:31, 9 September 2022 (UTC)

For the 2026 version of this page, I suggest two block sum columns in the table, and a separate plot showing the balance between the two blocks - the opposition and the government with support parties. @Avopeas: Would such a plot be technically possible? See the Swedish wp version of this article, showing that it was possible to handle that one party changed block in the summer of 2021.Tomastvivlaren (talk) 15:04, 12 September 2022 (UTC)