Jump to content

Talk:Opinion polling for the 2018 Pakistani general election

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


Dispute

[edit]

Clearly we are in dispute about what to do for this page. A few points:

1) Should the facebook poll be included or not? 2) Should events be allowed to show whether these events influence the voting intentions of parties, such as the disqualification of NS? Masterpha (talk) 17:06, 1 March 2018 (UTC)[reply]

Facebook is not a reliable source. Also, we should talk about Nawaz's disqualification.--Panam2014 (talk) 20:25, 1 March 2018 (UTC)[reply]
Ok, but why was the GSP (Global Strategic Partners) poll removed? It was cited by many independent news agencies: here, here and here. I only used the google drive version of the poll because it contained all of the information for voting intention at both the provincial and national level. Masterpha (talk) 20:57, 1 March 2018 (UTC)[reply]
We should have reliable source. --Panam2014 (talk) 23:07, 1 March 2018 (UTC)[reply]
Ok, but what do you deem to be reliable? Surely the GSP poll is reliable as it has been cited by many news agencies? Masterpha (talk) 07:29, 2 March 2018 (UTC)[reply]
Yes, all sources but no FB. --Panam2014 (talk) 11:37, 2 March 2018 (UTC)[reply]

Facebook poll

[edit]

Okay. So I understand that many of you might be hesitant to use the facebook poll. But may I ask you, why is this so? I will revert the poll that has been added by an anonymous user but can anyone explain to me what is so wrong with a poll which has a clear methodology explained and is by all means reliable, expect for the fact it is facebook? Regards, Masterpha (talk) 06:42, 24 March 2018 (UTC)[reply]

People are adamant on these polls being included, so unless there are any objections I will let them on the page Masterpha (talk) 12:48, 24 March 2018 (UTC)[reply]

To comment on the reliability of the poll, if the sample size is under 1,000 nationwide it will not be deemed reliable, or if the results are too extreme in relation to other polls, like the PakPolling one Masterpha (talk) 20:19, 26 March 2018 (UTC)[reply]

Facebook news sources are unreliable, and this pollster in particular is a primary source. We should be particularly wary of this source which only exists on Facebook (can you find coverage of this operation's polling results in any other publication?), conducts its polls through Facebook and Survey Monkey, and does not publish its methodology or sampling error or any of the other critical information that professional poll operators publish along with their results to inform their accuracy. A reliable polling source looks like this: pages and pages of statistics, methodology, and discussion, not just a pretty pie chart with some stock art, and most importantly for our purposes it has secondary source coverage. The OOP polls are not WP:RS and should not be included. Ivanvector (Talk/Edits) 20:23, 28 March 2018 (UTC)[reply]

OOP just released a new poll, and I contacted them for information and had a chat with them. First of all, their polling is NOT done online, they have teams on the ground surveying people in person. Second of all, I asked their methodology and they sent it to me and even added it to their latest post, they have a margin of error of 3-4% at a 95% confidence interval with a sample size of 3355 respondants, their previous polls all had sample sizes above 1000. Thirdly, they are even providing polls to one of the three major political parties, who are using the polling data for their own decision making. In my opinion there is enough evidence to suggest that their polling is reliable enough to be included on Wikipedia. Have a look at their latest poll and feel free to contact them too https://www.facebook.com/PakistanOnlineOpinionPolling/posts/1968865646517233 --Hussain.r97 (talk) 12:48, 20 June 2018 (UTC)[reply]

@Hussain.r97:, Roshan Pakistan opinion polling should explain their polling methodology rather than just say ‘it is good’. Also, if they can, they should create a website where this into can be stored as it does not fulfill WP:RS if it is in Wiki Masterpha (talk) 16:55, 20 June 2018 (UTC)[reply]
It is also pertinent to mention that creating websites is cheap. Wordpress.com allows you to create a great website for cheap, and if they do provide surveys to political parties I would assume that they have sufficient money to create a custom domain. Masterpha (talk) 20:55, 20 June 2018 (UTC)[reply]
@Masterpha: I totally understand your concern on the fact that they don't have a website. However, when it comes to methodology, they have explained it: METHODOLOGY: Our coverage was national for the first time (all four provinces including FATA), and rural/urban, based on 3355. 56% of our respondents were male, the balance women. 33% were young (under 30), 49% middle aged (30-50), and 19% old (50+). 65% of our respondents were in rural areas and 35% of the respondents were in urban areas. Our mode of surveying was face to face, and the survey was conducted from June 8-13. You can expect a margin of error of 3-4% at a 95% confidence interval. I had a chat to them and they are actually relatively new, but they seem quite genuine. --Hussain.r97 (talk) 21:53, 20 June 2018 (UTC)[reply]
@Hussain.r97: the main point is that we will not use a facebook source. If they wish to create a website and put their polling results and methodology on this then we might allow this. Masterpha (talk) 06:35, 21 June 2018 (UTC)[reply]
@Masterpha: Roshan Pakistan Opinion Polling have created a website now, I think we can definitely add them to the page now? https://roshanpakistanopinionpolling.wordpress.com/ --Hussain.r97 (talk) 10:20, 25 June 2018 (UTC)[reply]
We are already discussing this, on a seperate discussion on the talk page. Masterpha (talk) 11:04, 25 June 2018 (UTC)[reply]
[edit]

@Ivanvector: why are you removing external links? It is perfectly acceptable in opinion poll pages for there to be external links to the polls rather than references. I don’t even need to give you examples, they are in plain sight. I am returning the external links. Masterpha (talk) 07:53, 30 March 2018 (UTC)[reply]

@Ivanvector:Sorry, that might have seemed agressive. There might be a good reason to use references instead of external links, so I just want to understand why before I revert. Regards, Masterpha (talk) 10:17, 30 March 2018 (UTC)[reply]

(edit conflict) I didn't remove the links, I just converted to inline references, then linked the polling firms' names to our articles on those firms where I could find one. This preserves the citation in the reference, allows it to be reused throughout the article where it's used in different polling tables, and gives the reader a link to follow to additional content on the polling group itself, which is the usual expectation of bluelinks in a Wikipedia article. I haven't actually removed any links, they're just elsewhere on the page. There are not supposed to be external links in the body of an article at all, according to our guideline. If you can find many examples of external links in polling articles, you have found many errors that should be corrected. Ivanvector (Talk/Edits) 10:33, 30 March 2018 (UTC)[reply]
[edit]

Masterpha, the source seems to be a WSJ article that is locked for non-subscribers. Is there a link for it that is on the domain of Gallup Pakistan, the pollster that conducted it? If so, please change it to that so the information remains verifiable for everyone.

Thanks, Wiki.0hlic (talk) 12:49, 29 April 2018 (UTC)[reply]

Wiki.0hlic, thank you for reaching out to me in this regard. Whilst I have not found a non-wsj source for the poll, I have emailed one of the directors at gallup about this poll, for which I am still awaiting a response. Regards, Masterpha (talk) 19:04, 29 April 2018 (UTC)[reply]
Masterpha, thank you for your response. I will watch this space for any updates on the issue. That said, if Gallup Pakistan doesn't respond to you on the matter within a reasonable timeframe, we would have to rollback on this edit until a proper reference source is available. Best, Wiki.0hlic (talk) 11:56, 30 April 2018 (UTC)[reply]
Wiki.0hlic, I have just received an email from Gallup, saying that the poll was authentic and conducted in March 2018. They did not, however, elaborate further as to when in March this poll was conducted or what the minor parties position was in the poll. Best, Masterpha (talk) 15:30, 30 April 2018 (UTC)[reply]
Thank you for the update, Masterpha. Weird that WSJ has access to this poll before it is officially released by Gallup Pakistan. There is no mention of it on their website or any Pakistani news channel/website. Regardless, will watch for any updates on Gallup Pakistan's Twitter handle. Best, Wiki.0hlic (talk) 11:47, 1 May 2018 (UTC)[reply]

Semi-Protection Request

[edit]

Due to increasing number of edits without proper and disputed sources and expecting an increasing in such cases as the elections near, I am proposing that the main page be semi-protected Jibran1998 (talk) 20:04, 20 June 2018 (UTC)[reply]

support There are far too many unregistered users insistent on adding the facebook poll without discussing it on the talk page. I hope this move will encourage them to put their ideas on the talk page Masterpha (talk) 20:22, 20 June 2018 (UTC)[reply]
support There seems to be over-reliance on weak Facebook data as source for some poll results. The source in question has no publications to their name. Their only claim to adding data on Wikipedia is a Facebook page, something which is inherently against the guidelines of Wikipedia's reference source policies.Wiki.0hlic (talk) 21:19, 20 June 2018 (UTC)[reply]
This needs to be implemented fast. We are still getting vandals. Masterpha (talk) 16:31, 21 June 2018 (UTC)[reply]

Roshan Pakistan Opinion Polling

[edit]

Here they have created their own website. I realise this may be classed as a ‘self published source’, but I believe it meets the criteria for being allowed. For example, it is neither making exceptional claims, nor is there any reasonable doubt to its authenticity. Masterpha (talk) 06:05, 25 June 2018 (UTC)[reply]

@Wiki.0hlic:, @Jibran1998: Masterpha (talk) 06:06, 25 June 2018 (UTC)[reply]

@Masterpha:, yes the website does not make exceptional claims and they even claim to have a "polling firm" in Islamabad, so support it to be posted- Jibran1998 (talk) 10:24, 25 June 2018 (UTC)[reply]
@Masterpha: I disagree. Take a look at Wikipedia's policy on this. I am not for the inclusion of their data unless they get some sort of credence - like a reliable source (newspaper, TV program) citing them. As of the present rationale, I can create a blogspot page and put up my own 'research data'. Will it be fit for citation on Wikipedia?
As for this: "Our polling firm, located in I-8 Markaz, Islamabad, provides quality opinion polling data to two major political parties in Pakistan at the moment, amongst other organizations." You can't get any more ambiguous than that. No address, no contact info, just ridiculous claims. Did these 'major' parties hire them off a Facebook page while passing on reputable local pollsters like Gallup and IPOR?
That said, I am not accusing these Roshan Pakistan pollsters of skimping on their research but I have some serious doubts regarding their veracity. - Wiki.0hlic (talk) 11:08, 25 June 2018 (UTC)[reply]
@Wiki.0hlic:, apologies, only read this after I completed the edit. I suppose they have given their address seeing as they have told us I-8 Islamabad? Masterpha (talk) 11:23, 25 June 2018 (UTC)[reply]
@Masterpha: That's a whole sector of Islamabad, not a pinpoint address. Sort of saying like having an office in DHA Lahore or Clifton.Wiki.0hlic (talk) 14:17, 25 June 2018 (UTC)[reply]
@Wiki.0hlic: I agree with you because of poor sourcing. At the same time Duniya / ARY news surveys are also not reliable sources. I have serious reservations on including them. For example a TV anchor visiting two or three locations in a Urban centre like Multan and drawing inferences about voting pattern of whole region South Punjab comprising of 46 constituencies. We must stick to professional survey organizations most preferably Gall up Pakistan. Such organization select strata based samples and use scientifically and statistically proven technics. If we want to keep this page as per quality standards then we have to be very particular about professionalism. Jawadmdr (talk) 14:32, 25 June 2018 (UTC)[reply]
@Jawadmdr: Dunya News is a reputable media house in the country and their surveys are being conducted throughout Pakistan. Based on your point of only Multan being surveyed out of South Punjab - the surveyor, Habib Akram, has visited several S. Punjab constituencies including the ones in Multan, D. G. Khan, Rajanpur, R. Y. Khan, Vehari, and Khanewal. Complete list is here. Furthermore, these surveys are backed by video evidence. I will take the Dunya News surveys over anything conducted by Roshan Pakistan or Poll Maker. @Masterpha:, please re-add the Dunya News surveys. No consensus was reached on their removal and yet @Jawadmdr: removed them based on his single message on the Talk page. Wiki.0hlic (talk) 15:36, 25 June 2018 (UTC)[reply]

Agreed. So we will not have roshan. But. @Jawadmdr: what are you doing adding online polls and removing dunya polls? If anything dunya is much more reliable than any ‘election watch maker’ Masterpha (talk) 15:05, 25 June 2018 (UTC)[reply]

@Masterpha: I appreciate your good work on this page. Poll maker is backed by a election watch. You can visit election watch and see professional credentials reflected in Swing seat analysis. Incumbency analysis, Electable Vs Party analysis, Seat Hattrick analysis and so on. Thanks Jawadmdr (talk) 15:15, 25 June 2018 (UTC)[reply]

@Masterpha:, @Jawadmdr: I believe the Dunya News polls should be added to the article, on what basis can we assume that the poll might be biased? When it is being streamed on National Tv viewed by millions more than here so makes no difference if we add it here Jibran1998 (talk) 15:34, 25 June 2018 (UTC)[reply]

@Wiki.0hlic: Your answer is actually making my point more valid. You said Dunya News is a reputable media house. Yes it is but Is it reputable survey organization ? @Jibran1998: Who said they are biased ? They must not be But are they professional surveyors ? May be I have to involve a ADMIN for POV push by @Masterpha: by repeatedly adding Duniya news survey and removing ARY news survey, May be double standards ???? Jawadmdr (talk) 16:06, 25 June 2018 (UTC)[reply]
@Jawadmdr:, news organizations conduct public opinion surveys throughout the world. Here are some survey pages from USA, India, and United Kingdom. You can clearly see who is going for a PoV push.Wiki.0hlic (talk) 16:56, 25 June 2018 (UTC)[reply]

@Jawadmdr: There is consensus between 3 editors here, re edit the ARY News polls with proper reference. You are being warned for edit warring Jibran1998 (talk) 16:12, 25 June 2018 (UTC)[reply]

@Jibran1998: consensus is not majority. It is Consensus. Read the policy If you add any thing its burdened on you to get consensus otherwise we have to move to arbitration. Jawadmdr (talk) 16:16, 25 June 2018 (UTC)[reply]
@Jawadmdr: you are free to move to arbitration, but till then the edits stay with 3 editors majority Jibran1998 (talk) 16:21, 25 June 2018 (UTC)[reply]

@Saqib:, @SheriffIsInTown:, please input your opinion on this Dunya News source Jibran1998 (talk) 16:25, 25 June 2018 (UTC)[reply]

Even 100 users can not put anything against the WP policy of reliable sourcing. Jawadmdr (talk) 16:28, 25 June 2018 (UTC)[reply]
@Jawadmdr: your interpretation of WP policy is against the majority of users. Masterpha (talk) 16:49, 25 June 2018 (UTC)[reply]
  • A big no for Roshan Pakistan poll, Wordpress is a blog and not a reliable source, anyone can put anything on WordPress. ARY is known to be biased in favor of PTI and against PML (N). I am ok with Dunya poll, they are more neutral of the two although Pakistani media organizations are prone to making mistakes. If we can avoid then avoid them otherwise it does not hurt to include Dunya poll. Rely more on Gallup, PILDAT etc. Sheriff | ☎ 911 | 16:52, 25 June 2018 (UTC)[reply]

Comment There has been severe edit warring at this article, to the point where several of you were in danger of getting blocked for violating WP:3RR. I have fully protected the article for two days, so that you can continue to discuss this here on the talk page and hopefully reach some kind of consensus. --MelanieN (talk) 16:59, 25 June 2018 (UTC)[reply]

@MelanieN: most users are in consensus. It is just that one user in particular wants to implement his point of view regardless of what the talk page consensus is. None of this would have happened had it not been for him. Masterpha (talk) 17:02, 25 June 2018 (UTC)[reply]
@Jibran1998: I second what Sheriff said above. --Saqib (talk) 17:16, 25 June 2018 (UTC)[reply]
@Saqib:, @SheriffIsInTown: THe Roshan Pakistan thing was over, we were talking about the Dunya News Poll which was being constantly removed by one user Jibran1998 (talk) 17:21, 25 June 2018 (UTC)[reply]

::I am fine with inclusion of Dunya News poll! Sheriff | ☎ 911 | 17:53, 25 June 2018 (UTC)[reply]

Dear ADMIN kindly reverse Duniya news link. In Pakistan Media houses take sides of political parties therefore their surveys are biased. Please see other Pakistani user is also confirming this fact [1] and [2] . Even the person who is adding Duniya link has confessed the fact in his previous edit summary [3]. I will further add, No editor has any objection on reputable survey organizations like Gallup Pakistan, Pulse consultants and GSP but Media houses such as Duniya and ARY are known for taking sides in favour of political parties. So during Talk page discussion period please remove controversial edits [4] . Jawadmdr (talk) 18:06, 25 June 2018 (UTC)[reply]

Example of a reliable survey on Latest Pakistan election One example of an independent survey by SDPI, secondary sourced by Herald and thirdly sourced by DAWN [5]. This survey, financially and technically was supported by the Sustainable Development Policy Institute, an Islamabad-based independent think tank, was carried out in the first two weeks of April 2018 in 136 districts and regions across Pakistan. Chosen in accordance with the 2017 census data, the total number of respondents who participated in the survey was 1,497; 849 males and 648 females. They were further divided by locality: 652 respondents were from urban areas and 845 from rural areas. Respondents were also split on the basis of provinces and regions: 794 were from Punjab, 348 from Sindh, 219 from Khyber Pakhtunkhwa, 88 from Balochistan, 34 from the Federally Administered Tribal Areas (Fata) and 14 from Islamabad. The last parameter for division was mother language: 584 were Punjabi speakers, 266 were Pashto speakers, 231 were Urdu speakers, 216 were Sindhi speakers, 87 were Seraiki speakers, 61 were Balochi speakers, 20 were Hindko speakers, 8 were Gujarati speakers and 24 respondents spoke other languages.

This is how a professional survey organization works not like a media anchor with mike in hand. Jawadmdr (talk) 18:24, 25 June 2018 (UTC)[reply]

@Jawadmdr:You have to realise that the majority of respondents in Habib Akram surveys are not done in camera. Only significant responses are to show how people think. Also, he very clearly explains his methodology in tv shows. You also bring up that i reverted an ARY poll. This is because ARY has a pro-PTI tendency. I realise that my summary was incorrect, and I apologise if you think I am being hypocritical. But for now, consensus has been reached that they will be added, and you should ideally concede this. Masterpha (talk) 20:27, 25 June 2018 (UTC)[reply]

Dunya News

[edit]

Please use this section for the discourse on the issue of adding\removing Dunya News Polls. Add either support or oppose before continuing with the rationale so @MelanieN: can tally the final count. @Masterpha:, @Jibran1998:, @Saqib:, @SheriffIsInTown:, @Jawadmdr:. Thanks.

support Dunya News is a reputable media house in the country and their constituency-centric surveys are being conducted throughout Pakistan. Complete list by constituency is here. Furthermore, these surveys are backed by video evidence. Regarding @Jawadmdr:'s point of news organization not conducting surveys (or being biased), here are some survey pages from USA, India, and United Kingdom. All of these use TV channels/media houses as sources for certain surveys. Finally, a news organization can have inclinations towards a certain brand of politics. For example, it is conservatism for Fox News and liberalism for CNN and MSNBC but that doesn't merit the exclusion of their data in the aforementioned USA page. And Dunya News haven't had any history of journalistic favoritism unlike ARY News (pro PTI) and Geo News (pro PMLN). So, unless, evidence is brought to suggest that Dunya News has doctored the survey results or has an established history of overwhelming favoritism in their news reports, there is no cause for removing their survey data - which, I believe is some of the most comprehensive being conducted by any news organization in Pakistan.Wiki.0hlic (talk) 18:33, 25 June 2018 (UTC)[reply]

  • Oppose/Don't include Duniya well, CNN collaborated with ORC international- a specialist, NBC with Survey monkey, Reuters with Ipsos. Others have a history. Duniya News neither have history of O poll nor collaboration with researchist. EXCLUSION will be good & stop inclusion of any future dummy surveys by Media houses. Using Media houses as 2ndry source will be fair not as primary source. StLouis2 (talk) 07:16, 26 June 2018 (UTC)[reply]
  • Oppose per StLouis2. --Saqib (talk) 07:25, 26 June 2018 (UTC)[reply]
@Saqib: first you support, now you oppose? Masterpha (talk) 08:20, 26 June 2018 (UTC)[reply]
What? Where did I supported? I said I second what Sheriff said by which I meant that I would not cite Roshan Pakistan poll due to concerns Sherrif pointed and that we should avoid polls by news media organisations and rely on pollsters such as Gallup, PILDAT. --Saqib (talk) 09:43, 26 June 2018 (UTC)[reply]
Apologies, misunderstood what you said. Sheriff said that he was ok with dunya polls because dunya is neutral so by seconding that i thought you were endorsing dunya polls. Masterpha (talk) 10:20, 26 June 2018 (UTC)[reply]
  • Comment: I was thinking about opposing since I read the last comment by Jawadmdr, I think they are right in their assessment of what the right way of polling is. You cannot do opinion polling the way Habib Akram did although I do not question his or Dunya's neutrality but why insist on including it when certain requirements of proper opinion polling are not met. I will also like to suggest to the folks who wants to include it to not make this as a matter of ego and let it go. The example of these channels reliability is that Geo News was showing this same Wikipedia page (while trying their best to not show the statistics from Dunya) to discuss election trend in one of their shows (Aapas Ki Baat) yesterday and this debate and edit warring might have started from there as someone working on that project might have wanted to object Dunya's inclusion (infighting of channels coming to Wikipedia). It looks like while we are relying on them, they are relying on us. On the other side of the debate, the question is whether we should succumb to infighting and jealousy of these channels towards each other and not include something on Wikipedia just because they don't like each other. Sheriff | ☎ 911 | 11:25, 26 June 2018 (UTC)[reply]
  • Oppose: Addendum to my above comment, not enough details in source regarding the methodology to warrant an inclusion! Sheriff | ☎ 911 | 16:19, 26 June 2018 (UTC)[reply]
  • Oppose per my above comments. I also swear to God that I have no link to Geo, Duniya, ARY or any media channel. Neither I am member of any political party. In fact I have never voted in my life due to logistics issues. I simply want Fair elections and Fair Wikipedia. I also request that no survey must be updated before discussion on talk page. This will help us evaluate and make a fair call on inclusion. I have seen pathetic Duniya link and Habib akram episodes. EXAMPLE: Shangla which is PMLN's Ameer Maqam winning constituency, when someone said I will vote PTI he smiled and when someone answered PMLN he started arguing Nawaz is corrupt, Ameer maqam is failure. I mean how pathetic was this interview cum so called Opinion polling. I was laughing on his biasness and non professional approach. He was influencing and humiliating a person who was answering PMLN as his party. All these programmes are in Urdu language. May be Foreign WP users wont understand and judge this biased poll. Jawadmdr (talk) 18:00, 26 June 2018 (UTC)[reply]

Final Tally at the end of the protection date is 3 supporters and 4 dissenters. Hence, as most users do not support the motion, the dunya polls shall NOT be allowed on the page Masterpha (talk) 21:01, 27 June 2018 (UTC)[reply]

@Masterpha: Great, nice resolution to the issue. I think this is a good method for the addition/removal of potential contentious surveys before things get ugly with edit warring.Wiki.0hlic (talk) 16:46, 28 June 2018 (UTC)[reply]

GSP/IPOR poll

[edit]

@Jibran1998: @Wiki.0hlic: Ok, so I realise that we were calling the pollster from the survey dated 24 Oct 2017 as ‘GSP’. Upon further inspection of the poll, it is actually collected by IPOR consulting. Here on the last page it says that the survey was collected by IPOR. Should we change the pollster name from GSP to IPOR? Masterpha (talk) 21:02, 2 July 2018 (UTC)[reply]

@Masterpha:, please follow the format here. Where there is more than one party involved in the publication/data collection of a survey, the format is 'Publisher'/'Data Collector'. For example, CNN/ORC, NBC/Survey Monkey. So, in this scenario we should signify the pollsters as 'GSP/IPOR'. - Wiki.0hlic (talk) 22:03, 2 July 2018 (UTC)[reply]

Aggregate polls

[edit]

Just out of interest, what is your source for the aggregate polls? I mean, i see no sourcing for this. If you just calculated these percentages by yourself that is not good enough Aggregators are fine, but the aggregation here is self calculated. In us elections, like your example, aggregations are not calculated by an individual. Additionally, the aggregations are by pollsters not by all polls. NEXT TIME, actually respond to me on the TALK PAGE rather than proceeding to revert my edit. I had already engaged in a discussion with you. Follow WP. Additionally, do not engage in edit wars. Talk about controversial changes like this before proceeding with them. You also claim that aggregators are used in all pages. You gave one example. My examples of pages which do not use aggregate: United Kingdom, France, Turkey, Denmark, Belgium, Germany, Italy, Ireland, Spain, I can keep listing. Now tell me a page other than US and one which uses self calculated aggregate polls. My final point is that the US page is of a presidential election while this page is of parliamentary elections. Masterpha (talk) 18:38, 4 July 2018 (UTC)[reply]

Masterpha in statistics aggregation is used to beat 2 sampling errors, One Random error, Two System error (Biasness), On wikipedia we have a summary on every article in the lead that sums up all sourced data. I am doing the same. So stop edit war. Dont try to own this page.StLouis2 (talk) 18:42, 4 July 2018 (UTC)[reply]
@StLouis2: your points are entirely incorrect. We do not have such a consensus on all opinion poll pages. I am not trying to own the page at all, just pointing out that your points are invalid on many accounts. Firstly, even the aggregations on the US page are 1. Not self calculated, calculated by seperate pollsters and 2. Are not based on such large time periods as 365 days and 5 years. In no opinion polling page will you find self calculated aggregations with no reliable secondary source. Finally, barely any polls for PARLIAMENTARY ELECTIONS (which are the kind of elections in pakistan) hold aggregations. You said in an edit summary that all opinion poll pages have aggregations, but only those with presidential elections do, and those are never self calculated. Masterpha (talk) 18:57, 4 July 2018 (UTC)[reply]

@Saqib:, @SheriffIsInTown:,@Wiki.0hlic:, @Jawadmdr: @Jibran1998: please input whether self calculated aggregate polls without proper sourcing are allowed on the page.Masterpha (talk) 19:01, 4 July 2018 (UTC)[reply]


  • Oppose deletion by Proposer: and move them to Lead of the article per this style guide [6] It says The lead, or introduction, is the most important text in any article. Summarize the most important points of the article.Use discretion when putting citations in the lead means citation not mandatory but discretionary. Don't include information that is not covered later on. Here individual surveys are covered later on.Jawadmdr (talk) 18:27, 5 July 2018 (UTC)[reply]
@Jawadmdr: so you are ok with self calculated percentages with poor sourcing being included in the article? Masterpha (talk) 05:13, 6 July 2018 (UTC)[reply]
Sourcing is not required as I hv shared Wikipedia policy on Calulation (below) StLouis2 (talk) 09:55, 6 July 2018 (UTC)[reply]
  • Support deletion by Proposer inclusion of self aggregated polls. Base them on reliable sources. I might have been okay if it was a shorter time period but not for such long period as StLouis2 is proposing. Sheriff | ☎ 911 | 18:41, 5 July 2018 (UTC)[reply]
So you are OK with aggregate average but you have issue with period. Sourcing is not required as I hv already shared Wikipedia policy on Calulation & Jawadmdr has on Wikipedia policy on summary. Hope You will suggest a time scale of moving average.StLouis2 (talk) 09:53, 6 July 2018 (UTC)[reply]
@StLouis2: provided thay they are meaningful and agreed to by users. These aggregates are not meaningful as there is a large time period on them. Also I would advise you to stop editing what people write on the talk page and removing entire statements because of your likes and dislikes. Masterpha (talk) 06:59, 6 July 2018 (UTC)[reply]
@Masterpha: don't confuse the reader by deleting proper vote tagging. Who support/oppose the proposer's view should be clear. You had already given your consent for adding aggregates. Now you want to delete or change period then propose and get concensus. StLouis2 (talk) 07:55, 6 July 2018 (UTC)[reply]
@StLouis2: don't edit other users' comments. Period. Ivanvector (Talk/Edits) 19:02, 6 July 2018 (UTC)[reply]
  • Support deletion by Proposer No need for Aggregate number yet, if everyone agrees then it can be added a few days before elections. New Verdict against a major political leader is about to be announced, it may affect polls very significantly and many pollsters will definitely conduct polls for the month of july before election day Jibran1998 (talk) 09:09, 6 July 2018 (UTC)[reply]
Jibran1998 , In statistics there is concept of moving averages. Election are only 20 days away so it is best time to include averages.StLouis2 (talk) 10:07, 6 July 2018 (UTC)[reply]
  • @StLouis2: I propose that a separate section be made for the Aggregate then, the article is about opinion polls so no harm in dedicating a section to Aggregate number
  • Support deletion by Proposer - No need for an unsourced Aggregate Poll on a parliamentary election page. Particularly when there is no precedent for it on Wikipedia - at least none that I have found despite looking extensively. Furthermore, the time periods you have used are counterintuitive to the basic process of poll aggregation (where only polls from separate pollsters who have conducted them within a few days of one another are aggregated). Your calculation, on the other hand, aggregates polls across 5 years and thus doesn't factor in the significant voter movement trends which are induced by major political events. And Pakistani political landscape has been in turmoil since the Panama Leaks. How does an aggregation of 1800+ days given a close estimate of how people will vote on July 25th? Further still, given the political volatility over the past year, the 365-day poll aggregate is also effectively useless in predicting a general trend. It basically uses poll data that is months apart and has had significant voter movement from poll to poll. - Wiki.0hlic (talk) 09:24, 7 July 2018 (UTC)[reply]


  • Comment StLouis2 Please do not engage in disruptive editing by altering others comments. You are not understanding the consensus policy. The onus to get consensus is not on Masterpha for deletion but on you for inclusion. Sheriff | ☎ 911 | 09:53, 6 July 2018 (UTC)[reply]
  • Comment regarding period used SheriffIsInTown I think you have not gone through Article history. EVENT 1, I included aggregates. EVENT 2, Masterpha removed with edit summary that add them in separate section.(so he gave consent). EVENT 3, I added in new section. EVENT 4, Masterpha backed out now and proposed deletion on Talk page. Hope fully it will be clear to you. Most importantly All of us has said that adding aggregates is ok if period used is reduced Well we have to look the context. It is election 2018 happening after 2013 so X2013 fulfils the context. One year is reflecting short term. This is also part of ELECTION 2018 article under section Polls conducted one year before electionStLouis2 (talk) 10:02, 6 July 2018 (UTC)[reply]
@StLouis2: it’s called changing your mind. I agreed to aggregates at first, but then I took a deeper look and decided they were not fit for the article. Additionally, you cannot delete my statement because you think I have already expressed my views. The emboldened support or oppose helps one to count up the people who dissent and those who support a change. By removing this you are, whether inadvertently or not, tampering with the functions of the talk page. Also, may I remind you that YOU have in fact put an ‘oppose deletion by prosper’ for yourself’, despite making your views clear Masterpha (talk) 18:33, 6 July 2018 (UTC)[reply]

I have summed up this disarrayed conversation. Jawadmdr (talk) 12:44, 7 July 2018 (UTC)[reply]

Objection By Answer Answered by Recommendation
Self calculated OR Masterpha, Sheriff Routine calculations are not original research] StLouis2 who don’t agree must read policy
Aggregates unsourced Masterpha, Wiki.0hlic, Sheriff citations not necessary , JawadMdr who don’t agree must read policy
Add few days b4 elections Jibran1998 18 days 2 election StLouis2 Agreed by Jibran1998
No precedent Wiki.0hlic USA aggregates JawadMdr who don’t agree must visit page
Situation is volitile Wiki.0hlic WP has nothing to do with this JawadMdr We ll keep updating
Agree but Period is long Masterpha, Wiki.0hlic, Sheriff Context 2013 election, 1 year already used StLouis2 New section added 4 views on period
Edit war Masterpha On dis page U r warring most JawadMdr Stop everyone
Changing comments StLouis2 Masterpha Stop it Ivanvector StLouis2 must stop
I Agreed 1st , Mind changed Masterpha Keep it up JawadMdr Change it Again (LOL)

@Jawadmdr: - stop shaping the discussion. I didn't say Agree but Period is long. I am not supporting the inclusion of the aggregate polls. I have disagreed on the basis of unsourced content (the WP you have linked to is for the lead section, not the body - there is difference between the two), no precedent (parliamentary election pages, not a presidential one as you have linked above) and a flawed approach to aggregation (poll aggregators aggregate polls from different pollsters that are within a few days from one another to give a general trendline - they don't aggregate across months and years as is the case here).Wiki.0hlic (talk) 15:17, 7 July 2018 (UTC)[reply]

Oh my God, what a mess. Wiki.0hlic I think you have not visited this article but you are participating on Talk discussion. Please first Re visit article. See its Title is ‘’Opinion polling for the Pakistani general election,2018’’ . Then see its content includes all polls including 2013 election numbers. NO ONE EVER objected to the content of article for months, it implies all had consensus to all polls and 2013 numbers. So if a user summarises them by a simple calculation of average which can be checked by anyone. Then what is this noise all about. We should include aggregates in the lead. It is compliance of Wikipedia policy that lead should have summary and it doesn’t need any citation. Another policy that simple calculations involving addition , division (Both used in average) is not original research. Please stop wasting others time for ego issues. 2 policies are very clear. StLouis2 calculation was double checked by me.Jawadmdr (talk) 17:41, 7 July 2018 (UTC)[reply]
Try to stay calm and refute the facts I have put up. Content cannot be added to Wikipedia on one's whim and wish. Also, I'd like to know how you are tracking my Wikipedia activity? How do you know that I have never visited this article and yet am participating in the discussion? It is clear as to who has an ego issue here.
  1. Aggregations are sourced everywhere they are used on Wikipedia. And they are not included in the lead section. If you don't know, the lead section is the one before the Contents table, i.e., the introduction.
  2. There is no precedent for poll aggregations on a parliamentary elections page on Wiki.
  3. A flawed approach to aggregation is being used here. To reiterate, poll aggregators aggregate polls from different pollsters that are within a few days from one another to give a general trendline - they don't aggregate across months and years as is the case here.
- Wiki.0hlic (talk) 03:47, 8 July 2018 (UTC)[reply]
Oppose deletion by Proposer: and move them to Lead of the article per this style guide , (reproduced My edit 5 July 2018), I think even you have not read this talk page so I reproduce my comment. By the way all Your questions are already replied. So read back this time please and then these key points
  1. This is a separate article, Its a Talk page to discuss this article. Anything that is according to policy can be added here irrespective what other articles do.
  2. Period discussion is in below section. If you have objection to 2013,2014,2015,2016,2017 polls then you add a new disscussion and get concensus for deleting them from this article on the grounds that those surveys are not relevant to election 2018.
  3. Un less you get all old survey deleted. Aggregate will contain all surveys displayed on article.
  4. Nothing personal, Best of luck Jawadmdr (talk) 08:57, 8 July 2018 (UTC)[reply]
What a mess!, Seeing the table, I never mentioned anywhere that I oppose the deletion, aggregate polls should not be included in the main Section, because they are not sourced. Second mentioning that other countries election opinion article included aggregate, I saw none in Turkish opinion article. Finally if there was precision in the number, there "might" have been some precedent to include. Please get this discussion over with now Jibran1998 (talk) 10:34, 11 July 2018 (UTC)[reply]
Jibran1998 refer above all your questions are answered above. Also reproducing what you said earlier No need for Aggregate number yet, if everyone agrees then it can be added a few days before elections. New Verdict against a major political leader is about to be announced, it may affect polls very significantly and many pollsters will definitely conduct polls for the month of july before election day Jibran1998 (talk) 09:09, 6 July 2018 (UTC)[reply]
Jibran1998 , In statistics there is concept of moving averages. Election are only 20 days away so it is best time to include averages.StLouis2 (talk) 10:07, 6 July 2018 (UTC)[reply]
  • @StLouis2: I propose that a separate section be made for the Aggregate then, the article is about opinion polls so no harm in dedicating a section to Aggregate number
Jawadmdr (talk) 16:11, 11 July 2018 (UTC)[reply]
  • Final vote count at the end of the discussion is 4 supporters and 2 dissenters of deletion by prosper. Hence, the majority opinion is that these aggregate polls should NOT be on the page. The decision will be implemented as such. Masterpha (talk) 11:19, 11 July 2018 (UTC)[reply]
Mr. Masterpha you are proposer you can not close discussion. All queries were answered as per policy. You failed to get consensus. Jawadmdr (talk) 16:14, 11 July 2018 (UTC)[reply]

@Jawadmdr: Questions were answered but answers were rejected. As far as me, Jibran1998, Wiki.0hlic and SheriffIsInTown are concerned, there is no consensus to keep the polls. Only two editors, you and StLouis2, are in favour of keeping them. Don’t act like an expert on WP. Simple fact is that you are not willing to let go on your argument. @Ivanvector: @MelanieN: please close this discussion seeing as there is no consensus for these ‘aggregates’ to be added. Or, at least in the views of StLouis2 and Jawadmdr, for them to be kept. Masterpha (talk) 17:37, 11 July 2018 (UTC)[reply]

@Jawadmdr: - act professionally. Look at the consensus on the Dunya Polls voting and how it resulted in their amicable removal from the page. This time the consensus is against the Aggregate Polls so I second @Masterpha:'s motion of closing the voting with their removal from the Wiki article. Wiki.0hlic (talk) 18:13, 11 July 2018 (UTC)[reply]

@Ivanvector: @MelanieN: - please take note of @Jawadmdr:'s disruptive editing on the article. No one has changed their vote on this discussion with 4 clear supporters of the deletion and 2 dissenters, yet they still maintain that no consensus is reached and even consensus cannot override the policy. - Wiki.0hlic (talk) 18:36, 11 July 2018 (UTC)[reply]
@Masterpha: shame on your attitude. Policy is clear. If you lack argument. Wait for administrator decision. I will accept admin decision. Show some grace against your fellow country man. Show big heart to remove cheap tactics like Sock puppet investigations.Jawadmdr (talk) 18:34, 11 July 2018 (UTC)[reply]
@Jawadmdr: our argument is clear. You have not made any sufficient counter argument to change our mind. It is not a cheap tactic for sockpuppet investigations: it is a genuine way to check whether a user is abusing multiple accounts. If you are not sockpuppeteering, great. But carrying out this investigation is my right. A cheaper tactic is playing the patriotism card on wiki: I love my country and I don’t have to prove that by dropping investigations on people just because ‘they are countrymen’. Masterpha (talk) 18:43, 11 July 2018 (UTC)[reply]
@Masterpha, Jawadmdr, and Wiki.0hlic: Guys you all three have been doing good work lately. I should advice you to don't get stressed out and stay civil. --Saqib (talk) 17:44, 13 July 2018 (UTC)[reply]
@Saqib, Masterpha, and Wiki.0hlic: Ok inshallah I will, despite my country man using failed SPIs against me. I also acknowledge good work by all of you as well.@SheriffIsInTown, StLouis2, and Jibran1998: Jawadmdr (talk) 08:45, 14 July 2018 (UTC)[reply]

@Jawadmdr: Jawad bhai the SPI was merely to check whether you were abusing multiple accounts, please do not take it personally. I just saw something fishy with the way the two accounts were being handled. I am more than happy if you are not sockpuppeteering Masterpha (talk) 11:02, 14 July 2018 (UTC)[reply]

  • Request for closure of discussion. By proposers logic all summary map / aggregators are unsourced. While I support Both on grounds of WP Lead and WP CALC @Ivanvector: @MelanieN: please make a final decision based on above discussion. Thanks all and apologies for hearting egos. Please see double standards of keeping graphical summary but not aggregates. Here is that [8] Jawadmdr (talk) 16:36, 12 July 2018 (UTC)[reply]
@Jawadmdr: You are not listening. There is no precendent for such self calculated aggregates over such a long period of time. However, there is precedent for graphical summaries. [[9]], [[10]], [[11]], [here], [[12]], [here] and [here]. None of these pages use aggregates, yet all of them use graphics. There is no double standard in this. There are also many more pages like this. What you are doing is clear revenge vandalism and disruptive editing.Masterpha (talk) 16:56, 12 July 2018 (UTC)[reply]
  • I'm only going to offer my opinion here: the aggregates should not be included in the article, and no tables or graphical elements should be included in the lede at all. My problem with the aggregates is they are an unfair reflection of voting intent. Opinion polls are a snapshot of voter intent at a specific moment in time, you can't really aggregate instantaneous metrics over a long period of time in the way that is being proposed here without stripping the data of meaning - although the data contributing to the individual polls that are being aggregated may be reliably sourced, it's synthesis to combine those sources (through aggregation) to imply a conclusion that none of the individual polls intended. The situation could be different if a polling organization released aggregate polling data and we were sourcing that directly, but as I read it that is not the case here.
As for the graphical representation, I think it's fine and should be included in a separate section above the nationwide voting intention, as it had been for some time prior to this discussion. This is common practice in our opinion polling articles, we even have tools to generate them (see links below the graphical summary at Opinion polling in the 43rd Canadian federal election for example). As long as you explain what data is being graphically represented, the data itself is referenced (e.g. in the table below), and you don't use the graphical representation in a way that implies an unreferenced conclusion, then all your bases are covered.
-- Ivanvector (Talk/Edits) 17:26, 12 July 2018 (UTC)[reply]
@Jawadmdr: Double Standards? Since when are we taking sides on Wikipedia unless you are the one who is taking sides and proving again and again with your edits and remarks Jibran1998 (talk) 17:17, 12 July 2018 (UTC)[reply]
@Jibran1998: okay, that's enough of that. Discuss content, not contributors. Ivanvector (Talk/Edits) 18:48, 12 July 2018 (UTC)[reply]
Actually, current lead is perfect! Sheriff | ☎ 911 | 17:57, 12 July 2018 (UTC)[reply]
I've tweaked the lede just now to better reflect the dynamic situation in the actual campaigning period. I'm not familiar with Pakistani election process and it appears that the assembly dissolved right at the start of Ramadan and so not much campaigning happened for the first month, but the polling results during the campaign period seem to me to be more relevant to the outcome of the election than stagnant inter-election polls. I'm open to other opinions on this, of course. Ivanvector (Talk/Edits) 18:58, 12 July 2018 (UTC)[reply]
@Ivanvector: I think it would be fair to mention PML (N)'s slight lead over PTI. It might be a close race but no poll conducted after 31 May showed PTI leading! Sheriff | ☎ 911 | 19:12, 12 July 2018 (UTC)[reply]
The most recent (in the table) Gallup and Pulse polls have margins of error of 1.6% and "2-3%" respectively, according to the sources. With those error margins the results are tied within the margin of error and we can't say that either party is actually leading. The July IPOR poll has a margin of error of 2%, resulting in a PML(N) lead of less than two confidence intervals, which is very slim indeed. I'm more comfortable describing it as a close race and leaving it to readers to interpret the results, but that could change if more polls are released before the election. Ivanvector (Talk/Edits) 20:27, 12 July 2018 (UTC)[reply]

Page full-protected, 4 July 2018

[edit]

I have upgraded the page protection here to full protection due to the ongoing edit war. @Masterpha and StLouis2: discuss your issues with the proposed content here, and if you cannot come to a resolution, please try dispute resolution, request a third opinion, or start a request for comments to seek input from uninvolved editors. If I have to come back here because of an edit war again I will start blocking editors instead. Ivanvector (Talk/Edits) 18:58, 4 July 2018 (UTC)[reply]

@Ivanvector: I request that once the full protection has been lifted for semi protection to be reinstated. We have discussed this on a previous talk page section and every time the page’s full protection expires it seems that all of it’s protection expires. Masterpha (talk) 05:23, 5 July 2018 (UTC)[reply]

That's correct, full protection replaces semi-protection, and when a protection level expires it resets to no protection. I'll come back to check on this in a few days (or ping me) and will reset MelanieN's semi-protection when the dispute resolves, but I expect to see some discussion of the issues or I may just decide to protect the page for a longer period of time. Ivanvector (Talk/Edits) 12:29, 5 July 2018 (UTC)[reply]
@Ivanvector: Thank you for handling this. I'm sorry to see that the edit warring resumed, and I agree that some blocks may be necessary if it happens again. I will be away from the computer for the next week or so, so please continue to monitor this article and act accordingly. --MelanieN (talk) 13:17, 5 July 2018 (UTC)[reply]

@MelanieN: @Ivanvector: the protection has expired. Masterpha (talk) 06:56, 8 July 2018 (UTC)[reply]

@MelanieN: @Ivanvector: please implement this, vandalism is already happening Masterpha (talk) 17:58, 8 July 2018 (UTC)[reply]

 Done. I was waiting to see if vandalism would recur, but it did, so I've restored MelanieN's protection level and duration. We were both away from the project for the weekend, but if this had been urgent you could have made a request at WP:RFPP or left a message at WP:AN referring to my note above. Ivanvector (Talk/Edits) 14:37, 9 July 2018 (UTC)[reply]

Leading paragraph added and to stop edit war I sourced it

[edit]

I added this

General elections are scheduled to be held in Pakistan on 25 July 2018. Since the previous general election, several pollsters have conducted opinion polls at both the national and provincial levels. Pakistan Muslim League (PML-N) leading most of polls nation wide and strong 15-21% lead in key province Punjab (54% of all Pakistan constituencies) in all region Central, North, West and South Punjab. Pakistan Tehreek-e-Insaf (PTI) is second most popular party in most of nationwide surveys and is having comfortable lead in KPK province, while Pakistan Peoples Party (PPP) in third place with significant lead in Sindh.
and sourced it [13] Jawadmdr (talk) 19:04, 15 July 2018 (UTC)[reply]


@Jawadmdr: - its you who are edit warring. 3 separate editors have reverted your edits to the intro in the past day. What you have written is not a lead, this is basically the whole page. Do you even know how to write a lead? Take a look at other opinion polling pages to see how it is written: 1, 2, 3, 4, 5, 6, and 7.

@Ivanvector: - please take note of this user's perpetual disruptive behavior on this page. - Wiki.0hlic (talk) 19:39, 15 July 2018 (UTC)[reply]

Lead section discussion - July 17

[edit]

@StLouis2, Wiki.0hlic, SheriffIsInTown, Masterpha, and Jawadmdr: I accept the criticism that the lead I wrote falls into original research territory: it is my own summary of the recent polls. It shouldn't be there unless there's a source to support that opinion, notwithstanding the brief discussion above about confidence intervals. Adding what I wrote without a source is no different than Jawadmdr adding their own apparently pro-PML(N) lead with inadequate sourcing. We should find a reliable source or else not add anything at all.

Times Now News published an article summarizing recent poll results just yesterday. Is this publisher reliable for this purpose? It doesn't set off the typical red flags for me but I'm not really familiar with Pakistani news sources. It does contain a subheading "Overall mood of the voter: PML-N, PTI are neck and neck", which seems like it would support what I had written. What do you think? Ivanvector (Talk/Edits) 13:59, 17 July 2018 (UTC)[reply]

I am not sure about the credibility of that source as I have never heard of that specific publisher before. That being said my understanding was that lead does not need a source if it correctly summarizes the article and what you wrote was correct that polls released after 31 May put the race between these parties very close. Although, I was more in favor of summarizing the whole article since this article is not about post 31 May polls, it contains pre 31 May polls as well so we should be right to write that PML (N) has an overall lead in almost every poll released so far! Sheriff | ☎ 911 | 14:36, 17 July 2018 (UTC)[reply]
@Ivanvector: honestly, we don’t really need a summary. There are so little polls conducted that a summary is basically useless at the point. Masterpha (talk) 15:26, 17 July 2018 (UTC)[reply]
Your suggestion does not comply with WP:LEAD and manual of style. Lead must summarize the article content. I suggest to add something like Most of the opinion polls suggest an overall Pakistan Muslim League (N) (PML (N)) lead with Pakistan Tehreek-e-Insaf (PTI) being the close second. Short and sweet and summarizes all the polls, not just the ones conducted after 31 May! Sheriff | ☎ 911 | 19:26, 17 July 2018 (UTC)[reply]
I think the best way is mention that PML(N) has a small lead Nationwide with a good lead in Punjab, PTI lead in KPK etc.... Jibran1998 (talk) 19:35, 17 July 2018 (UTC)[reply]
@Ivanvector: the biggest opinion poll was conducted by IPOR interviewing approx 215,000 people in key province Punjab (54% of Pakistani constituencies) that is not showing any close contest. Other opinion polls also maintain more then 10% lead of PMLN. Above 10% can not be called close by any means. Remaining small provinces have separate patches of different small regional, religious Parties such as MQM, PSP, MMA, ANP QWP, PKMAP, NP, BNP, BNPA and GDA as well as PPP, PTI and PMLN.
@Jawadmdr: biggest does not mean most factual. The data was collected from a period of 2 months, meaning that significant voter movement could also have had occured during this. Methodology is what matters in opinion polls. Masterpha (talk) 00:05, 18 July 2018 (UTC)[reply]

What I am saying is that there are not a lot of polls conductsd in pakistan but the most recent ones have shown a close competition. Hence, why do polls from before mean anything now? Sure, pmln may have had a large lead THEN, but now they has clearly been a swing against PMLN/for PTI detected in the polls. Saying that pmln is ahead and pti is behind is lying. So in this instance I support @SheriffIsInTown: with his summary about pti being in close second... Masterpha (talk) 00:13, 18 July 2018 (UTC)[reply]

Additionally, provincial polls vary by pollster. The most recent ones have shown varying results, even in punjab: the geo/jang polls show 9 to 14 point leads, while the IPOR polls show 19-21 point leads. Sindh is quite complicated too: we know that ppp is ahead but who is second? KPK shows PTI ahead but by how large of a margin? Polling in Balochistan is just a mess, full stop. Are we meant to summarise all of this in the lead? If we wish to summarise, we should do them in the respective sub sections of the article. Masterpha (talk) 00:16, 18 July 2018 (UTC)[reply]

Comment - Opinion polling pages across Wikipedia have a concise lead that don't even summarize results: 1, 2, 3, 4, 5, 6, and 7. I don't think there is need to summarize content which isn't hidden in a wall of text but is easily visible in graphs and tables. StLouis2/Jawadmdr have been writing the same lead over and over again focusing on the Punjab province and using a single source regional breakdown (IPOR) - no other surveyor has given a regional breakdown of Punjab province. It is clear from their edits that they want to paint the lead as biased towards a particular party. - Wiki.0hlic (talk) 05:35, 18 July 2018 (UTC)[reply]

New survey

[edit]

@Ivanvector: @Enigmaman: @MelanieN: there is a new survey. However, seeing as the page is full protected once again (despite there no longer being any content dispute) I cannot add it. I request that either semi-protection be reinstated in the place of full protection, or one of you administrators manually add the poll into the tables. This is the poll. маsтегрнатаLк 10:49, 21 July 2018 (UTC)[reply]

@Oshwah: as per above Jibran1998 (talk) 11:50, 21 July 2018 (UTC)[reply]
@Ivanvector: @Enigmaman: @MelanieN:, please stick with the page protection but add the data of the new survey to the article. - Wiki.0hlic (talk) 19:35, 21 July 2018 (UTC)[reply]

Oppose I oppose removing the full protection due to un stoppable edit wars . Jawadmdr (talk) 19:01, 21 July 2018 (UTC)[reply]

Masterphal, I have no idea why you keep saying there's no content dispute when there very clearly is. I fully protected in the hopes that it would be worked out. Enigmamsg 02:23, 22 July 2018 (UTC)[reply]
@Enigmaman: before this there was no longer any content dispute on the issue of the lead. There is now dispute about a survey. Apologies. маsтегрнатаLк 02:34, 22 July 2018 (UTC)[reply]
A road trip survey in which random people from only 56 districts were questioned was added by Masterpha. Even a school going children will not believe that PMLN PTI PPP three are almost equal in 20s as per this survey. Read the online resources and news papers. Contest is between PMLN and PTI. PPP is even weakest in its strong hood Interior Sindh but this joke survey shows PPP at 20% . Salutes masterpha sir for wasting our time. Jawadmdr (talk) 09:04, 22 July 2018 (UTC)[reply]
  • I also oppose removing protection for what it's worth. I requested at RFPP for it to be upped to full protection until after the election, due to ongoing content disputes over various polls being included or not, and several editors revert warring in place of discussing. So discuss. Ivanvector (Talk/Edits) 12:57, 23 July 2018 (UTC)[reply]

Protected edit request on 21 July 2018

[edit]

Please add the data of the the following survey to the article. Wiki.0hlic (talk) 19:37, 21 July 2018 (UTC)[reply]

I second this. At first I was asking for semi protection to be reinstated but now with jawadmdr continuing his disruptive editing, I appeal to the administrators to add this new survey into the article, keeping full protection. маsтегрнатаLк 01:25, 22 July 2018 (UTC) @Enigmaman: I am no longer asking for unprotection, just for the new poll to be added in. маsтегрнатаLк 02:36, 22 July 2018 (UTC)[reply]

I asked Ivan to do it. I'm not comfortable doing it unless you can lay out exactly what you want added and where in the article to put it. Enigmamsg 02:55, 22 July 2018 (UTC)[reply]
I'm also not going to do anything through protection, per WP:INVOLVED, for what it's worth. Ivanvector (Talk/Edits) 12:59, 23 July 2018 (UTC)[reply]

@Enigmaman: this is a controversial survey. Similarly these twin wiki users have added one blog and another Non neutral media based survey. Now this survey which is already controversial nationwide. We all previously concluded that we will not add any controversial survey and first discuss on Talk page. Please see new Talk page section above. Jawadmdr (talk) 08:44, 22 July 2018 (UTC)[reply]

Serious calculation errors in this survey. It says PMLN has 25% support. But if we multiply all 4 provinces survey %age with each province number of seats then we end up 26.3 %. I am adding a table to show this blunder in the reported nationwide numbers.

@Jawadmdr: where are your numbers coming from? 141, 61, 51, 16? маsтегрнатаLк 10:47, 22 July 2018 (UTC)[reply]
Brother see provine wise summary of this survey . It says PMLN popularity @ 40% in Punjab. 4% in Sindh. 10% in KPK 4% in Baluchistan. We all know Pakistan has 272 constituencies and how they are province wise. so when I multiply province percentages with 272 province wise breakdown. I ended 26.3% as shown above. Serious calculation errors in this survey. It says PMLN has 25% support. Jawadmdr (talk) 10:56, 22 July 2018 (UTC)[reply]

@Ivanvector: @Enigmaman: @MelanieN: please delete this new controvertial survey which has self contradictory errors and also controversial on Pakistani media and social media. Jawadmdr (talk) 10:59, 22 July 2018 (UTC)[reply]

@Jawadmdr: evidence for being controversial? Social media is not acceptable: pti supporters will have rejected earlier polls putting pmln ahead as will pmln have rejected those putting pti ahead. His point is entirely invalid.маsтегрнатаLк 11:09, 22 July 2018 (UTC)[reply]
@Jawadmdr: pakistani NA seats are not representative of the population. For example, FATA is overrepresented and constituencies are much smaller in size. Additionally the question being asked province wide may be different: voting intention for provincial assemblies tend to differ from the national assembly маsтегрнатаLк 11:03, 22 July 2018 (UTC)[reply]
No there is no such difference in all other surveys.
Second no survey shows PPP at 20% all keep at 13-15 and last survey which was just conducted at same time also keep it at 13%. So there sample looks like Sindh centric and non representative. Jawadmdr (talk) 12:03, 22 July 2018 (UTC)[reply]
@Jawadmdr:Look at my answer on the general election talk page. NA seats are not totally representative of the population. There are quite representative, but out of 272 they cannot be exactly representative of 210,000,000 people. And due to rounding even 25.49 would be considered as 25%, which is only a 0.751% difference, one which is not entirely implausible. Also just because a previous poll has not mentioned it does not mean that it is necessarily incorrect. (Would like to point out that you have broken the 3RR. Disappointed that you did not fully read the definition of edit warring like I suggested)маsтегрнатаLк 12:11, 22 July 2018 (UTC)[reply]

My argument is about election 2018 which is going to be conducted on election constituencies 2018 based on 2017 population census. If I accept your logic of 6 extra FATA seats more than population then that will further increase PMLN representation to 27%.Jawadmdr (talk) 13:27, 22 July 2018 (UTC)[reply]

  • Oppose SDPI all surveys show unreliastic numbers for PTI and PPP see 2014 and 2015 surveys. Everyone is surprised in pak media and joking this faulty survey. God knows they can show 200% for PTI on order of those who are openly riging 2018 election. Those who are not aware of pre poll rigging in Pakistan election 2018 can read article on Pakistani election 2018 and reffered sources there. StLouis2 (talk) 18:58, 22 July 2018 (UTC)[reply]
@StLouis2: who is everyone in Pak Media, care to share links of their views on this survey? - Jibran1998 (talk) 19:03, 22 July 2018 (UTC)[reply]
@StLouis2: Your opinion makes sense as I see three out of four polls which show PTI in lead or conducted by SDPI. Why others never find PTI in lead? I am wondering if there are any reliable sources which show that SDPI have a history of conducting faulty surveys under the influence of some specific political party or Pakistani establishment (i.e. military and intelligence agencies)! Sheriff | ☎ 911 | 21:24, 22 July 2018 (UTC)[reply]
@SheriffIsInTown and StLouis2: Support please do not try to make this another ground for calling out a ‘conspiracy’ against pmln. Those polls were conducted in 2014-15 where no other polls were conducted hence we can’t compare them to other polls. The recent one is fairly in line with the other polls. And even if there is pre poll rigging against pmln, then these results would prove that, no? That the military’s influence has worked and people are now shifting to pti? Pre poll rigging is not the same as normal rigging: it is designed to change the opinions of people.маsтегрнатаLк 02:56, 23 July 2018 (UTC)[reply]
Additionally, there is not enough ground for the survey’s to be labelled as unreliable. Everyone in pak media? I have genuinely never heard any respectable media organisation say anything against this poll. If you are talking about twitter and facebook, then go ahead. Polls showing one party ahead will always be controversial to the party that is behind. For example, pti supporters call surveys showing pmln ahead ‘paid surveys’. Are we meant to remove them for being controversial? маsтегрнатаLк 03:08, 23 July 2018 (UTC)[reply]
And as for the issue of ‘faulty surveys’: we will see which surveys are faulty in 3 days time :) маsтегрнатаLк 03:11, 23 July 2018 (UTC)[reply]
@Masterpha: I never expressed my personal opinion, I never said that there is a conspiracy theory against PML (N), there are countless sources proving that. People from inside judiciary (such as Justice Shaukat Aziz Siddiqui) are speaking out regarding this. As for the survey is concerned, if it is faulty and influenced by military and intelligence agencies then it should not be allowed regardless of what the result is in three days time. They can use whatever tool they have available in their own country to manipulate the elections, Wikipedia will not become their tool in this game of thrones. This is definitely quite fishy that there are three surveys conducted by SDPI at different times during last five years and all are lead by PTI, all other surveys conducted by others before, after and in between SDPI surveys have PTI trailing! Sheriff | ☎ 911 | 04:11, 23 July 2018 (UTC)[reply]
@SheriffIsInTown: SDPI were the first ones to conduct voting intention surveys: no surveys of the sorr were conducted by any other polling organistaions until mid-2017. What I would like to say is that the two surveys conducted in 2014 an 2015 may well have been accurate: the first showed a large pti lead, the second one showed a smaller one, hence showing that pti’s popularity was decreasing. In 2016 there were no surveys, hence we can say that significant voter movement could have had occured in favour of pmln in this year. In 2017 the first poll showed pmln with a 16 point lead. This double digit lead was echoed throughout the next year. However voter movement started going in the favour of pti in late may and early june: a pulse consultant poll which showed pti with a 13 point lead in november 2017 now shows a 3 point pti lead in may 2018. Similar voter movement is evident in IPOR and Gallup surveys: a 8 point pmln lead from gallup was reduced to a one point pmln lead in may/june, and a 11 point pmln lead from IPOR was reduced to a 3 point pmln lead in June/July. The poll showing a 4 point pti lead is reasonable: it is not showing a massive 6-14 point point lead like it did previously. Now, seeing as SDPI did not conduct any polls between 2016-2017, we cannot say that they would not have shown a pmln lead there. In fact I would bet that if an SDPI poll was conducted in 2017, it would have had shown a pmln lead. The general message, which is being echoed by all pollsters, is that polls are showing that the race is too close to call and that there is no clear leader. The point about methodology is that while provinces may not have been given absolute equal weightage, they have been given generally good representation: 25-26 is not a huge error! Also, what evidence do we have that SDPI is influenced by the military apart from it is showing PTI leads? In fact the last poll that SDPI conducted in 2013 showed a pmln lead I believe.маsтегрнатаLк 04:59, 23 July 2018 (UTC)[reply]

Also justice siddiqui’s remarks should not be blindly believed. I would like to reiterate that there is no absolute evidence that there is military and judicial interference in favour of pmln and against pti: something which was rightly called out by BBC hardtalk’s stephen sackur against hameed haroon in this interview. We can’t prove these allegations just because political commentators are saying that they are true. маsтегрнатаLк 05:04, 23 July 2018 (UTC)[reply]

I would also like to point out that PMLN, PTI, JI and PMLQ are all in fact clients of IPOR. PPP is not. Should we remove their survey’s because of this, due to a possible ‘anti-PPP’ bias? No. of course not. SDPI and IPOR both gain nothing by publishing fake surveys. маsтегрнатаLк 05:12, 23 July 2018 (UTC)[reply]

  • Support inclusion of the new survey. I see nothing to indicate that the poll is part of some vast government conspiracy to be righted, and Jawadmdr's mathematical analysis of the poll's flaws is itself deeply flawed, and shows a fundamental lack of understanding of statistical analysis. As I observed before and the polls support, PML-N and PTI have been in a statistical dead heat since the start of the election; the survey itself calls the results "too close to call". Ivanvector (Talk/Edits) 13:22, 23 July 2018 (UTC)[reply]
Seconded. - Wiki.0hlic (talk) 16:50, 23 July 2018 (UTC)[reply]

@Enigmaman: @MelanieN: Three users are against inclusion of survey. Please immediately delete the survey until consensus develops Already slow decision making is making our discussions un necessary dragging with foolish arguments. Jawadmdr (talk) 17:18, 23 July 2018 (UTC)[reply]

I'm washing my hands of this. You can make an editprotected request, but there's disagreement on so many things here. Enigmamsg 17:26, 23 July 2018 (UTC)[reply]
  • Support Nothing in the survey indicates a conspiracy of establishments, just because a survey continuous lead of PTI doesn't means it is bias, there is a close race between PML-N and PTI, and if considering a margin error as indicated in all survey, nothing wrong in posting this one aswell. - Jibran1998 (talk) 18:21, 23 July 2018 (UTC)[reply]

I hate to be dragged in here where I know nothing about the subject, but I will try to evaluate the discussion here since somebody has to. From this discussion I conclude there is consensus to include the SDPI/Herald poll. (I am puzzled by Jawadmdr’s request to delete it; it isn’t actually in the article, is it?) I will add it, if someone will lay out for me, on this talk page, exactly what should be added, where, and in what format. In the meantime the full protection is clearly still needed. --MelanieN (talk) 18:27, 23 July 2018 (UTC)[reply]

@MelanieN:, the survey has already been added, under "Nationwide polling intention", the 12 Jul 2018 SDPI poll - Jibran1998 (talk) 18:32, 23 July 2018 (UTC)[reply]
(edit conflict) I believe it's already been done - see Masterpha's series of edits yesterday from about 4:25 - 5:15 UTC, while Enigmaman had unprotected the page for that purpose. Jawadmdr is maintaining opposition to its inclusion but (involved opinion) there is consensus to include it, or at least no consensus that it should be removed. Ivanvector (Talk/Edits) 18:34, 23 July 2018 (UTC)[reply]
OK, thanks. That's good. That wasn't clear, and people were still begging for action here. I concur that there is consensus to have this in the article. And that full protection is still needed. --MelanieN (talk) 18:41, 23 July 2018 (UTC)[reply]
I voted in this thread assuming that the survey in question might be removed if although I already knew it was included Jibran1998 (talk) 18:45, 23 July 2018 (UTC)[reply]
  • Clarification I did not oppose the inclusion, my latest comments were just intended to give opposing view a consideration as there is nothing very straightforward in Pakistan as some folks would like every one to believe. There is heavy involvement by elements of military and intelligence agencies in every step during the election as being reported by countless international media reports so if someone says that they might have skewed the survey results, we ought to pay attention but I did not change my position as I did not find proof of involvement in this specific survey so I am still in favor of inclusion. Sheriff | ☎ 911 | 18:50, 23 July 2018 (UTC)[reply]
Yes, I understood both of you to be in favor of inclusion. --MelanieN (talk) 19:01, 23 July 2018 (UTC)[reply]
Herald Magazine is the publisher of this survey and it belongs to the Dawn Media Group, whose CEO appeared on HARDtalk a few days ago with a clear agenda to expose the army's meddling in political affairs. He couldn't support his allegations with facts apart from an insinuation that army must be involved because "ISPR has trolls on social media". - Wiki.0hlic (talk) 12:20, 24 July 2018 (UTC)[reply]

Polls of Polls

[edit]
  • Support A aggregate of all polls conducted have been computed by Gallup Pakistan [14], I vote for its inclusion in the article - Jibran1998 (talk) 19:58, 24 July 2018 (UTC)[reply]
  • Comment - I'm inclined to oppose an aggregate poll's inclusion for the same reasons I opposed when this came up last week. The source provided for this new aggregate is pretty wishy-washy on whether conclusions can be drawn from it at all, let alone what those conclusions might be. It's just not really giving us any new information. Ivanvector (Talk/Edits) 21:31, 24 July 2018 (UTC)[reply]
  • Oppose: It includes Roshan Pakistan’s polls, which we havent given the go ahead to yet. Will only support if Roshan Pakistan is allowed. маsтегрнатаLк 02:56, 25 July 2018 (UTC)[reply]
Retrieved from "https://en.wikipedia.org/w/index.php?title=Talk:Opinion_polling_for_the_2018_Pakistani_general_election&oldid=1219535821"