Talk:Opinion polling for the next United Kingdom general election

Polls

@CipherRephic and Chessrat: In the deletion discussion you both stated there is now an opinion poll for the next United Kingdom general election. Can you please provide a link. Thanks --John B123 (talk) 18:48, 15 July 2024 (UTC)

Please ignore the above, its now in the article.--John B123 (talk) 18:53, 15 July 2024 (UTC)

There is a report produced by MoreInCommon and the UCL Policy Lab titled "Change Pending The Path to the 2024 General Election and Beyond" https://www.moreincommon.org.uk/media/e3in12zd/change-pending.pdf Can someone with more experience of these topics decide whether this would be more appropriate here or in the 2024 election topic.LarryJayCee (talk) 12:22, 16 July 2024 (UTC)

Opinion poll flaws should be made transparent

The voting in the 2024 election, as noted in the article, bore little relationship to the polling to the day before. That polling grossly exaggerated the Labour vote in the actual election. It is also the case that the latest poll after the election almost exactly mirrors the incorrect polling before the election. The indication in this article that Labour has improved since the election is utterly wrong: it the same as before the election.

I would suggest that a line giving polling average immediately prior to the election, as well as the result, would increase the understanding of readers to the strange anomalies which we see. RERTwiki (talk) 16:44, 16 July 2024 (UTC)

When the turnout on Polling day is the lowest since universal suffrage was introduced, which was not predicted, you cannot expect the pollsters who were assuming a rather higher turnout to get it right. YouGov's final MRP was pretty good, underestimating the Tories by 19 seats and overestimating Labour (not counting the Speaker) by the same number. It was evident to those who followed the polling closely that Labour's lead was falling rapidly in the last week before Polling day, but to see it you have to strip out the pollsters who had not been polling regularly before the election was called (and YouGov, who changed their methodology in the middle of the campaign). LarryJayCee (talk) 18:14, 16 July 2024 (UTC)

Out of all the numerous pollsters, a few were bound to get close just by chance. By just keeping an eye on what was going on, I got closer to the result than most of them. If you throw enough darts at a board, some will hit the bullseye. Btljs (talk) 21:32, 1 August 2024 (UTC)

Both RERTwiki and Btljs have good points. The polling for the last election really was hit and miss. Mainly miss. Having said that, the general trends were about right, although some would argue that that is stretching the meaning of the word "general". I’ve added a 'See also' wikilink to the last article for easy reference (which for some reason was overlooked) so that people can easily refer to those polls. Oh, and LarryJayCee, predictions of turnout are part of the opinion poll predictions, aren’t they. Boscaswell talk 04:57, 11 August 2024 (UTC)

It is very interesting how Mr. Starmer has rapidly gone from a popular PM to an intensely hated PM, I wonder if this is because of polling flaws or some other reason.173.79.50.25 (talk) 10:49, 14 August 2024 (UTC)

Second place

@FriendlyDataNerdV2 I notice you manually reverted my addition of shading for second place almost immediately after I added it. It hasn't been used in previous articles as largely speaking the second-place spot hasn't been in contention, but I think such information is useful to highlight because 1) the Leader of the Opposition spot is very important in British politics so the question of which party is likely to achieve that is certainly relevant, 2) in the runup to the 2024 general election there was a lot of media coverage when Reform came second in polls, so the question of who is in second place is clearly notable, and 3) the "Lead" column only displays a number and not which party the lead is over, but the question of who the lead is over is probably at least as important as the raw number.

There are alternative layout options (e.g. here's one) but I think the topic of making the second-placed party clear should be considered in general. Chessrat ^{(talk, contributions)} 11:41, 15 August 2024 (UTC)

Dates conducted	Pollster	Client	Area	Sample size	Lab	Con	Reform	Lib Dems	Green	SNP	Others	Lead
7–8 Aug	We Think		GB	1,278	33%	20%	21%	11%	8%	3%	4%	Lab +12	over Rfm
5–7 Aug	BMG Research		GB	1,523	33%	24%	18%	12%	8%	2%	2%	Lab +9	over Con
11–12 Jul	We Think		GB	2,005	39%	20%	16%	11%	9%	2%	TBA	Lab +19	over Con
4 Jul 2024	2024 general election	–	UK	–	33.7%	23.7%	14.3%	12.2%	6.8%	2.5%	6.8%	Lab +10.0	over Con
			GB		34.7%	24.4%	14.7%	12.5%	6.9%	2.5%	4.3%	Lab +10.3	over Con

Chessrat ^{(talk, contributions)} 11:41, 15 August 2024 (UTC)

I get your thinking on this but in my opinion it's unnecessary clutter. The lead is what's important and one can just look at the table to see the in-depth results. 82.14.16.140 (talk) 21:01, 15 August 2024 (UTC)

Agree: pointless addition that would only add unnecessary clutter and take up a lot of wiki markup space (which is unconvenient as we know how large these opinion polling articles can get). Plus, this is not a two-round presidential election where second place actually determines the two top candidates. Impru20^talk 08:06, 16 August 2024 (UTC)

That's true, but it's not a one-round presidential election either, and the lead in national vote and the lead in national seats can and have gone to different parties. So would it make sense to go the other way and remove the Lead column entirely, since it doesn't determine who wins and also duplicates other information already in the table? 2A02:C7C:DAE1:FC00:8904:8F1C:1DAC:A4BC (talk) 08:34, 17 August 2024 (UTC)

First party-lead is widely reported by reliable sources when publishing/commeting polls, so no, it shouldn't be removed. Impru20^talk 10:57, 17 August 2024 (UTC)

That's also true, but *also* widely reported by RS when publishing/commenting polls is the change in each value since the previous poll by the same polling company, and no-one is suggesting cluttering up the table by including all of that; as with the first party lead it can be calculated from other parts of the table if someone wants to know. I don't object to the lead being there but if the aim is to keep table size down it could go. 2A02:C7C:DAE1:FC00:80ED:9872:6070:74E4 (talk) 18:49, 18 August 2024 (UTC)

No one is suggesting that because you can easily compare the change in values between polls by each pollster with the table format. The lead column is not the cause of the big size, but adding another one "with text" is definitely not going to help either. Impru20^talk 08:05, 19 August 2024 (UTC)

I'm open to the idea of this suggestion, iff its relevance is born out by more data. So for now, I don't think it adds any value. However, if we see second place flitting between 2 or even 3 parties over a number of weeks and months, then yes this would relevant. It would show the graphical point(s) of inflection within the data tables. This is something that has been done in other non-UK election articles. For now, let's not make the edit but let's keep an eye on the data and be open to revisiting, perhaps this idea in the New Year. WestminsterWhistleblower (talk) 16:54, 23 August 2024 (UTC)

I agree with other editors above that this is unnecessary clutter. No other Wikipedia polling article does this. Bondegezou (talk) 20:32, 7 September 2024 (UTC)

Merge SNP column into 'Others'

Following the outcome of the 2024 election, we should assess whether we are giving the correct level of weight to political parties. It would appear that this article now gives the SNP undue prominance. Let's recap:

1. The SNP were removed from "Others" and given their own column in the UK wide polling tables following their highly significant general election victory in 2015. Prior to that they were always included within "Others".

2. While the SNP have retained 9 MPs, they are no longer a UK-wide nationally significant party, whether that be in terms of seats, vote share or their status in Parliament. Not only did they They have lost their higher share of PMQ's and representation on select committees etc. The SNP did not even manage to get 750,000, UKIP and the Greens had to surpass 1 million votes each before they were no longer deemed "Others"

3. The SNP only contest the 57 Scottish seats and so we should assess how they compare to other seperatist parties. If you compare the populations of Scotland and Wales, the SNP have a per capita number of MPs similar to Plaid Cymru (who are included in 'others'). Indeed, the SNP only have 2 more seats than Sinn Fein and a similar share of the vote in their constituent countries (again, Sinn Fein are included within "Others").

It is clear that there are 5 nationwide GB parties: Labour, Conservative, Lib Dem, Reform and Green. Everyone else is "others". SNP remain nationally significant in Scotland (no doubt about that) but not in the UK as a whole. We need to change the table to reflect this and not give the SNP undue prominance. WestminsterWhistleblower (talk) 17:23, 23 August 2024 (UTC)

I have made this edit following Wikipedia:Be bold  Done WestminsterWhistleblower (talk) 17:36, 23 August 2024 (UTC)

We can’t have been going by number of MPs as Reform had none and Green had one before the 2024 GE. I do think for parties that only stand in an individual nation, it is meaningless putting their share of GB or UK vote as a separate column. So I agree, just pedantically not for the reasons above. Btljs (talk) 06:42, 24 August 2024 (UTC)

Agree. Makes sense. Gotta drawer a line somewhere and do it consistently. 2A01:4B00:88F4:CE00:30E1:6FE:69B2:76F4 (talk) 07:06, 24 August 2024 (UTC)

If the SNP are permanently removed from the GB/UK table (that is, removed until things change at a future General Election), I suggest they should also be removed from the GB/UK graph. --Wavehunter (talk) 09:46, 24 August 2024 (UTC)

Agreed 2A01:4B00:88F4:CE00:21D3:EB04:2A46:F003 (talk) 19:42, 25 August 2024 (UTC)

Agree, I guess should the SNP have a resurgence they could always be added back, but for now they definitely don't meet the criteria. Icantthinkofausernames (talk) 12:59, 24 August 2024 (UTC)

I think the SNP should be included in its own column. The party has more MPs than Reform and the Greens who are both included. It just stands out as an outlier being the only party with more MPs than both these parties to not be included. I don't think the point stands in regards to its polling level, given its was included in its own column for the opinion polling for the last UK general election when it was typically polling similar to its 2024 result and what its polling at now at between 2 to 4%. Helper201 (talk) 06:18, 7 September 2024 (UTC)

Number of seats is not the only grounds of inclusion. They are back to their pre-2015 status in terms of support in GEs - we have to be consistent. They no longer meet the clear criteria that secured them their own column. 2A01:4B00:88F4:CE00:DD69:8ED7:AF23:7214 (talk) 09:13, 7 September 2024 (UTC)

I don't have strong views on this, but I think there is space in the table for an SNP column and RS mostly report the SNP in their topline results. We should follow RS. Bondegezou (talk) 20:40, 7 September 2024 (UTC)

If we included SNP for this Parliament we would have to include Plaid and also make that retrospective. SNP have clearly gone back to their pre2015 levels of support/representation. I suggest we stick with the clear perspective established above. 2A01:4B00:88F4:CE00:D8E:8338:3688:77EA (talk) 21:23, 7 September 2024 (UTC)

If we included SNP, we would not have to include Plaid. We've long had polling articles with the SNP included without Plaid. The SNP are currently faring poorly in voting and polls, but we don't know whether that's temporary, whether they'll improve again, or whether they'll fall further. We can't just draw those sorts of conclusions: it's WP:OR and WP:CRYSTALBALL! Their national polling is often reported by RS. Bondegezou (talk) 11:05, 8 September 2024 (UTC)

That's just not true. It is only since 2015 (when the SNP had quite an astounding result) that Wiki has included a column for the SNP in General Election polling - to suggest otherwise is just incorrect. For 3 Parliaments (2015-2024) they were the 3rd largest party (more than 4 times more seats than the Lib Dems) in the House of Commons; unquestionably this made them a UK wide significant party, despite them only standing in 57 of the 650 seats.

The reality is that since the 2024 election, the SNP have been reduced back to similar levels of support (votes and seats per capita) as Plaid - the difference is very marginal. And there is no argument that says that election result renders the SNP a nationally significant party for the purposes of UK polling.

I agree that we really need to avoid violating WP:CRYSTALBALL and we therefore need to use the most recent election result as the marker of where things are (as Wiki always has done) e.g. it is on the basis of % of votes received in the previous election result that we order the parties in the table. I completely acknowlage the possibility that the SNP may have a resurgence before the next election but to simply leave their column in gives them undue prominance on the basis of something that has not yet happened - so the Crystallball argument works more against your point than for it.

You also state that "Their national polling is often reported by RS" - only 3 pollsters (BMG, We Think and Stonehaven) have published polls since the election. If you follow the links to the articles that show the headline voting intetions e.g. The i, there is no mention of the SNP. Similarly, if you go into the tables, you will see that Plaid and the SNP are included on the same basis of each other (as they are prompted for in Wales and Scotland respectively). I accept that we are working with very little data here but the data we do have does more to disprove your point than prove it. WestminsterWhistleblower (talk) 11:36, 8 September 2024 (UTC)

I said We've long had polling articles with the SNP included without Plaid. You said That's just not true. It is only since 2015. Forgive me, but since 2015 is 9 years and I think 9 years does count as us long having something! Bondegezou (talk) 13:39, 8 September 2024 (UTC)

Do we not have stand alone opinion polling articles for UK GEs that go as far back as 1974? As in over 50 years? Not sure I agree with your definition of “long time” in this instance. Especially when you consider that 2 of the 3 parliaments covered are among the shortest on record! 2A01:4B00:88F4:CE00:AC8C:A89D:CDA8:1DC9 (talk) 19:00, 8 September 2024 (UTC)

Well, I do not want to spend too long on this minor point, but I was talking about real time, not what history we cover. That Opinion polling for the 1974 United Kingdom general elections article has only existed for 3 years! Bondegezou (talk) 13:14, 9 September 2024 (UTC)

Semi-protected edit request on 8 October 2024

Could someone please add this new poll by More In Common to the article? https://www.politico.eu/article/uk-labour-keir-starmer-lead-one-point-conservatives-new-poll-more-in-common/

It doesn't seem to have been added yet. JPowellOBrien (talk) 10:11, 8 October 2024 (UTC)

 Done by User:FriendlyDataNerdV2 in this edit. Liu1126 (talk) 11:45, 8 October 2024 (UTC)

Hello it's the person who keeps adding polls while logged out

There's a couple new leadership approval polls but I can't add them obviously:

https://www.opinium.com/wp-content/uploads/2024/10/VI-2024-10-02-Observer-Tables-V3.xlsx Summary V006

https://www.moreincommon.org.uk/media/5o5l5kfr/october-approval-tracker.xlsx goodjobbadjob

https://www.bmgresearch.co.uk/wp-content/uploads/2024/10/October-2024-omni-tables-for-the-i.xlsx tables 31-34

Thanks :) 129.234.0.181 (talk) 19:48, 8 October 2024 (UTC)

Semi-protected edit request on 9 October 2024

Leadership approval polling:
https://www.opinium.com/wp-content/uploads/2024/10/VI-2024-10-02-Observer-Tables-V3.xlsx Summary V006
https://www.moreincommon.org.uk/media/5o5l5kfr/october-approval-tracker.xlsx goodjobbadjob
https://www.bmgresearch.co.uk/wp-content/uploads/2024/10/October-2024-omni-tables-for-the-i.xlsx tables 31-34
General election vote polling is on the third link under table 3 too :)

I sent this yesterday but I think that other thing is closed/answered, so I'm reposting here. I'm the person who keeps editing while logged out. 129.234.0.181 (talk) 13:07, 9 October 2024 (UTC)

Edit to JL Partners poll

Reform UK vote share is 19% not 18%. Please can this be amended.

https://x.com/JLPartnersPolls/status/1846627943323521261 2A00:23C5:709A:8C01:4184:8FD0:3B7:B8B1 (talk) 20:12, 17 October 2024 (UTC)

 Done Icantthinkofausernames (talk) 12:02, 18 October 2024 (UTC)

Robert Jenrick Polling

The page includes the Jenrick commissioned polling with the hypotheticals of him/badenoch as leader. Whilst it should definitely be on the page I don't think it should be included as regular polling since it is a hypothetical poll. It should instead be moved to an 'other polling' section of this page. Benocalla2 (talk) 11:30, 18 October 2024 (UTC)

Disagree, I think this can be handled with a footnote indicating the disclaimer, as has been previously done in similar instances such as a constituency poll of Clacton back in January which asked voters how they would vote if Nigel Farage was Reform candidate vs. if it was someone else.

If it was asking people how they would vote under some completely different potential leader I would be more in favour of placing it in an 'other polling' section, but this poll in particular is asking how people would vote under the two final candidates in the Con leadership contest rather than some wild outlandish hypothesis. Even if it's agreed to place these polls somewhere else, I strongly oppose putting them under a whole new heading as polls like these are rare and there will likely be very few conducted between now and the next GE.

Another option I wouldn't be vehemently opposed to is placing the poll based on the eventual winning candidate in the regular section, while putting the one based on the runner-up somewhere else. Icantthinkofausernames (talk) 12:33, 18 October 2024 (UTC)

Agree I don't think polls for scenarios that are prompted should be added as they are conducted differently to the other polls, so having them in a hypothetical section is best. CoaxAndBotany (talk) 00:53, 19 October 2024 (UTC)