Talk:Opinion polling for the 2023 New Zealand general election
![]() | This article is rated List-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects: | ||||||||||||||||||||
|
Maori Party abbreviation
editShouldn't the table abbreviate The Māori Party/Te Pāti Māori as TPM instead of MRI? 2404:4402:3A05:9000:EC7C:473B:5F5A:8114 (talk) 03:47, 30 January 2023 (UTC)
Main blocs chart
editWhat’s the reason for the new “main blocs” chart? The chart doesn’t depict who is likely to form a government as the blocs cannot include parties that could go either way (eg. TPM, NZF, TOP). In the past, even the Greens have presented (at least publicly) that they could form a coalition with National, so the “LAB+GRN” bloc is not definitive. These blocs are one possible interpretation of polling, rather than a factual depiction of results. I appreciate the effort that’s put into maintaining the party vote chart, but I suggest removing the new “main blocs” chart as an MMP election is not a two horse race. 116.90.136.55 (talk) 23:07, 31 January 2023 (UTC)
- There was discussion about this at User talk:Nixinova#NZ Coalition Graph.-gadfium 23:35, 31 January 2023 (UTC)
- The discussion on that user page arrived at the view that the blocs should be LAB+GRN+TPM and NAT+ACT. There has been no reason expressed for including Te Pati Maori with non-parliamentary parties. Unless someone has a good reason otherwise, could this graph be redrawn to include TPM with LAB+GRN? 116.90.136.55 (talk) 03:56, 23 February 2023 (UTC)
Tauranga Polling
editIs it relevant to have a single poll from a single electorate that was taken during the 2022 Tauranga by-election on this page. The demographic differences between the Tauranga Electorate and the whole of NZ mean the vote share is wildly different to the national polling. As it is it just seems to cause confusion. I took action and moved it to the by election page. This move was undone. I think it is better suited there or maybe just removed completely as it seems superfluous and just creates confusion. ShakyIsles (talk) 04:30, 1 February 2023 (UTC)
- Of course the demographics are different to the whole of New Zealand, that's why it's separate. It will eventually be joined by more electorate polling and look less out of place, but it makes more sense here than the by-election article since there is no party vote in a by-election. --Pokelova (talk) 04:36, 1 February 2023 (UTC)
"Smoothing is set to span 65%" for the Party Vote graph
editI understand the principles behind smoothing. I have read the linked Wikipedia page on this topic.
In that Wikipedia page the term "span" appears twice, defining "the span of values of the y-axis".
In that Wikipedia page I could not see the expressions "%" or "percent".
Question: to what does "span 65%" refer in the Party Vote graph?
Discussion: I understand the Party Vote graph is aimed at showing past polling outcomes and attempting to suggest recent trends. In my view the apparent wide gap does not reflect the apparent closeness and possible volatility of reported polling outcomes. (See "Forecast" section)
In my experience, things measured some time in the past have little effect on things measured recently but a trend line can help show the direction of travel. This direction of travel is not apparent to me in the Party Vote graph
In my experience the span is usually best set at: either a certain number of values; or the values that occurred in a given time period.
While more consideration in selecting a span is necessary, my starting point for considering this series/graph would be EITHER 10 values OR all values in the last 120 days (approximately four months) AlwynWellington (talk) 18:37, 1 February 2023 (UTC)
Why is it necessary to use the word "impacted" when "affected" is probably the better word to use?
editSuggestion - Posie Parker
editRecommend adding the Posie Parker incident as a notable political event. Many pundits are predicting this will have a seismic impact on voting intention with old allegiances broken and new ones formed, "culture war" being more important than "class war" in the eyes of some. — Preceding unsigned comment added by 122.61.62.115 (talk) 08:03, 31 March 2023 (UTC)
- I don't think we should add events to the timeline based on what people think they might do to the polls. For that reason I am also opposed to the recent addition of the Nash controversy by @Lord A.Nelson: but would like to seek other opinions before removing that myself. It probably has a stronger case for staying than the terf rally. --Pokelova (talk) 08:22, 31 March 2023 (UTC)
- I think the sacking of a Cabinet Minister is significant enough, but not Posey Parker.NealeWellington (talk) 09:32, 31 March 2023 (UTC)
- Agree with NealeWellington that the dismissal of a cabinet minister is noteworthy. Can't see why Parker will have any effect but could be swayed by reliable sources that said so. Schwede66 18:31, 31 March 2023 (UTC)
Democracy NZ
editIs it time to add Democracy NZ to the polling data? They have been polling over 1% consistently the last few polls and are polling the same or better than the New Conservatives 203.211.104.200 (talk) 01:03, 8 May 2023 (UTC)
- I agree. They have also polled higher than the TOP party in some polls. In my opinion, all registered parties should be included to prevent any bias... 118.93.1.42 (talk) 08:54, 11 May 2023 (UTC)
- Wikipedia is not obligated to include every piece of information that exists. We are allowed to have reasonable limits. Most polls won't even bother mentioning minor parties that aren't polling at or near 1%, so including all registered parties would ultimately just be a whole bunch of blank space. The best you could hope for is an 'Other' column. So far DemocracyNZ has passed 1% a total of two times. I would not call that consistent. If they can keep it up, I'm sure there will be no objections to including them. Have some patience. There will be many polls to come before election day. --Pokelova (talk) 09:17, 11 May 2023 (UTC)
- Agree with Pokelova. Happy to have a discussion about where the threshold should be, but it will need to be above meeting 1% in a poll twice over three years. 2404:4402:3A0F:100:59B:21FB:FAAE:1FC8 (talk) 06:10, 11 July 2023 (UTC)
- Wikipedia is not obligated to include every piece of information that exists. We are allowed to have reasonable limits. Most polls won't even bother mentioning minor parties that aren't polling at or near 1%, so including all registered parties would ultimately just be a whole bunch of blank space. The best you could hope for is an 'Other' column. So far DemocracyNZ has passed 1% a total of two times. I would not call that consistent. If they can keep it up, I'm sure there will be no objections to including them. Have some patience. There will be many polls to come before election day. --Pokelova (talk) 09:17, 11 May 2023 (UTC)
Horizon results
editThe way Horizon reports its results is not consistent with the way other polling firms do. The headline figures in the cited sources that have been included in the table: a) include undecided voters, which lowers the total for all parties, and b)use a "100% likely to vote" filter on the main results they report.
Has any thought been given to reporting the raw figures the sources include instead? Rustie5555 (talk) 06:51, 22 May 2023 (UTC)
- I agree with you. The Horizon results appear statistically out of sync with the other polls. NealeWellington (talk) 09:18, 22 May 2023 (UTC)
- Agreed. I've updated the seat allocation and government formation based on the updated figures since it has NZ First over 5% and getting seats as a result. JJimbo2222 (talk) 04:18, 23 May 2023 (UTC)
Meng Foon resignation
editShould Meng Foons resignation be listed in the nationwide polling just like Meka Whaitiri resignation is? Kiwiz1338 (talk) 06:58, 16 June 2023 (UTC)
- No, it’s irrelevant. Foon is not an MP. Schwede66 17:39, 16 June 2023 (UTC)
- Alright Kiwiz1338 (talk) 23:07, 16 June 2023 (UTC)
Tracker poll
editThe new ‘tracker’ poll does not look reputable to me. Website has no details about methodology, sample size, etc. Site has spelling and grammar mistakes too. Recommend removing all references to it. 210.246.33.157 (talk) 08:30, 2 July 2023 (UTC)
- It Actually Looks like it was the Closet Pollster to the Actual Election Result
- Their Pre-Election Poll Looked Like This:
- Tracker Poll - Election Night Result
- National: 37.9 - 39.0
- Labour: 24.8 - 26.8
- Green: 11.4 - 10.8
- ACT: 10.1 - 9.0
- NZ1: 6.45 - 6.46
- TPM: 2.5 - 2.6
- Personally, I find this as compelling evidence to include this Poll for the Next General Election's Opinion Polling AKiwiAtHeart (talk) 08:00, 15 October 2023 (UTC)
Main block chart - question
editHi I love your work and I find this chart helpful in principle. However, I'm concerned about the accuracy. Adding up the results for the last 7 polls on your list I get the average for the National-Act block to be 46.8 and the average for Labour-Green-Maori to be 45.6 (since the beginning of June). I assume the graph lines are calculated by your computer. Given that the right-wing block is on average a percentage point above the left, I don't see how the graph can be showing that the left block greater than the right block over the last couple of months?
Thanks. FromJohnsonville (talk) 01:11, 23 July 2023 (UTC)
- The chart was last updated for polls ending June 2023. If you ignore the July polls then it should match. Nixinova T C 06:54, 25 July 2023 (UTC)
Graph Createion Dates
editI think the two graphs in the Party Vote section should more prominently display the date they were created. For example, right now they were last created on 12 July and therefore don't include data from the last three polls. 119.148.98.205 (talk) 14:33, 25 July 2023 (UTC)
Suggestion - Preffered PM Polling
editI have a chart made from Google Sheets of the PM Polling, if any mod or whatever whould like to use it till the actuall one starts working then I'd be more then happy to share it. AKiwiAtHeart (talk) 06:56, 30 July 2023 (UTC)
Graphs issues.
editThe party vote graphs are showing a June poll as the latest poll and it seems to be dramatically skewing the graph. Other graphs on the page aren't working at all and haven't been for months and months. 103.14.71.113 (talk) 23:28, 1 August 2023 (UTC)
- I agree, really needs sorting out AKiwiAtHeart (talk) 06:18, 2 August 2023 (UTC)
- the preferred Prime Minister graph and the country direction graph are still not working! Why? 103.242.24.107 (talk) 01:41, 11 August 2023 (UTC)
- the preferred pm and direction graphs still "temporarily" unavailable. When does temporary become permanent. This has been going on for months. 103.242.24.107 (talk) 23:58, 2 September 2023 (UTC)
- This is still not fixed! 210.54.242.219 (talk) 00:26, 18 September 2023 (UTC)
Where are the "don't knows" in all this?
editFrom the current article text: "Refusals are generally excluded from the party vote percentages, while question wording and the treatment of "don't know" responses and those not intending to vote may vary between survey organisations."
As RNZ MediaWatch are constantly pointing out, representing political polls without including the people who answered "don't know" seriously skews the results: https://www.rnz.co.nz/national/programmes/mediawatch/audio/2018725722/undecided-and-uninterested-undermine-poll-conclusions
If, say, 20% of the electorate hasn't decided who they're voting for, presenting a poll as implying that National will get 33.5% of the MPs, and Labour 26% of the MPs is highly misleading. Because even if everyone who gave an answer votes the same way on Polling Day, if even half of the "don't knows" decide to vote for someone other than those two, their percentage of seats would be much lower than the representation of the poll is implying. Or if the majority of them vote for National or Labour, that could totally change the result suggested by this representation of the poll results.
Note that it would be unwise to dismiss the "don't knows" as non-voters (although they are block). The "swing voters" - who are widely considered to decide elections - may not decide who to vote for until very soon before the election, or even on the day. Danylstrype (talk) 07:02, 3 August 2023 (UTC)
- One reason why don't know isn't included is because the propotion who give that option tends to map with the number who don't vote. 49.224.215.119 (talk) 06:29, 17 August 2023 (UTC)
"Lead" column highly misleading in MMP elections
editFrom the current article text:
"The 'party lead' column shows the percentage-point difference between the two parties with the highest figures."
The lead of any one party over another has no bearing on anything in an MMP election. What matters is relative support of any combination of parties that might be capable of forming a government. The "Lead" column in the "Nationwide polling" table is meaningless and misleading, and ought to be removed. Danylstrype (talk) 07:16, 3 August 2023 (UTC)
- Agree, it would be much more informative to show the lead between the main blocs. Imnofox (talk) 00:28, 5 October 2023 (UTC)
Guardian poll
editgiven the fact that the Guardian includes undecided in their pv numbers, unlike everyone else, wouldn't it be better to use the adjusted numbers (ie % of decided voters) to keep it consistent? 47.72.182.90 (talk) 03:57, 10 August 2023 (UTC)
Talbot Mills Poll
editWhy are we including the Talbot Mills poll? These are private polls which are occasionally selectively leaked publicly. This creates a publication bias. 147.161.216.255 (talk) 12:54, 21 August 2023 (UTC)
- They have 2 types of polls - their ones that they do for Labour, which are only occasionally leaked, and then their commercial ones, that they release regardless of the outcome. 47.72.182.90 (talk) 10:56, 2 September 2023 (UTC)
CFM Coromandel Polls
editIve seen the local radio station in the Coromandel have recently been issuing monthly polling out for the election, I was wondering if the Electorate Polling was reserved for all Electorates or tightly contested ones's since incumbent Scott Simpson has a astounding lead but there are a few supprises coming from the Party Vote with ACT having a 6% lead. Here is a refrence : https://cfm.co.nz/news/scott-simpson-looking-safe-act-continues-lead-in-coromandel-party-vote-in-august-cfm-poll/ AKiwiAtHeart (talk) 06:44, 27 August 2023 (UTC)
- It seems that the poll respondents are self-selected through this station's Facebook page, which is obviously not good practice. It's evident enough in their results, no reasonable person believes ACT is winning the party vote in that electorate. 203.211.79.215 (talk) 13:26, 27 August 2023 (UTC)
- I wouldnt put it past the people of the Coromandel. Many people that I have spoken to have said that they would vote for ACT in the Party Vote and National in the Electorate. Also im curious where you found the info for how people are self selected from the CFM polls. AKiwiAtHeart (talk) 05:14, 28 August 2023 (UTC)
- By looking at their Facebook page where they invite their followers to participate. Even if it wasn't so apparent as that, the fact that the link you provided makes no mention of methodology would be suspect enough. 203.211.79.215 (talk) 05:28, 28 August 2023 (UTC)
- I wouldnt put it past the people of the Coromandel. Many people that I have spoken to have said that they would vote for ACT in the Party Vote and National in the Electorate. Also im curious where you found the info for how people are self selected from the CFM polls. AKiwiAtHeart (talk) 05:14, 28 August 2023 (UTC)
Internal polls
editIs there a policy for including internal poll results in this article? I see Curia's recent poll for Ilam has been added, but I'm wondering if TOP's internal Ilam poll that was leaked should also be displayed alongside it. It's not clear which polling company was used (if any company), but the methodology seems to be described at the end of this Herald article. Specifically it says:
A leaked survey from the Manji campaign showed a very different result.
Campbell was still ahead on 27.2 per cent, followed more closely by Manji on 23.4 per cent.
Pallett was a distant third on 17.7 per cent.
That survey picked random addresses in the electorate from the 2020 electoral roll, and interviewed 2591 voters (1339 male and 1252 female) between August 9-18.
It's concerning that it's very different from the Curia poll, but if the sample size and methodology is accurate and representative then (to me anyway) it seems at least as likely to be reliable. Izogi (talk) 21:41, 30 August 2023 (UTC)
- It was conducted by party volunteers. That's deeply problematic in terms of getting a reliable result. People tend to say what they know that someone wants to hear. 47.72.182.90 (talk) 11:00, 2 September 2023 (UTC)
NZ Loyal
editShould perhaps replace New Conservatives, as New Cons are polling at 0.1% and NZ Loyal at 1.2%. New Cons are dying as a movement whereas NZ Loyal is taking off. Anna Nilsen (talk) 09:31, 22 September 2023 (UTC)
- I concur. NZ Loyal is virtually as popular as New Conservative and The Opportunities Party. If winning seats in parliament isn't a baseline for earning a place in the polling table then NZ Loyal has surely earned it by regularly polling above 1%. This is just my two cents though. Clesam11 (talk) 05:37, 13 October 2023 (UTC)
Swarbrick in Preferred PM section
editShould Chlöe Swarbrick be unhidden from the preferred PM table? At some point the criteria was set such that a person "must reach at least 3 percent in three separate polls" to be included, and Swarbrick has already exceeded this. So I assume it follows that either Swarbrick's numbers should be uncommented, or the criteria statement should be discussed/changed? 125.239.192.39 (talk) 12:43, 6 October 2023 (UTC)
- It doesn't mean consecutive polls right? If not then I think we can uncomment her. Kiwiz1338 (talk) 06:42, 11 October 2023 (UTC)