Talk:Opinion polling for the 2022 Brazilian presidential election

Lula release ordered today

With the release of Lula (Inacio da Silva) authorised today , he can be restored to the potential Workers' Party contenders. Culloty82 (talk) 20:54, 8 November 2019 (UTC)

References

Danger of deletion

Based on the recent precedent at Wikipedia:Articles for deletion/2022 in the Philippines, and that apparently consensus has in fact changed, this article could be deleted. I can't begin to state how categorically I disagree with this change in consensus, but it is what it is. Bearian (talk) 01:22, 7 February 2020 (UTC)

The chart data is out of sync

across series and dates (one can see that by pasting it into a spreadsheet, and expanding the csv's into columns). Not sure how best to fix it, though, other than remaking it from scratch. If it comes to that, then perhaps it'd be a good idea to include the data table, collpased, next to the chart, so that both editors and readers can verify the chart is plotted correctly. — Guarapiranga ☎ 23:59, 29 June 2022 (UTC)

I'm trying to fix the chart, Panam2014 (re: Special:Diff/1095730851). — Guarapiranga ☎ 01:27, 30 June 2022 (UTC)

Ah, now I see the problem: I broke the tables below. — Guarapiranga ☎ 01:45, 30 June 2022 (UTC)

 Fixed — Guarapiranga ☎ 03:58, 30 June 2022 (UTC)

Article title

There is currently no polling data in the article for the legislative election. Unless data for that is added, the title of this article should be changed to 'Opinion polling for the 2022 Brazilian presidential election' to ensure readers are not misled. N Panama 84534 06:37, 21 July 2022 (UTC)

True (not to mention the state government elections!). The sister page at ptwiki also has the qualifier: pt:Pesquisas eleitorais para a eleição presidencial de 2022 no Brasil. — Guarapiranga ☎ 08:35, 21 July 2022 (UTC)

 Done — Guarapiranga ☎ 08:38, 21 July 2022 (UTC)

I have moved the article to a title that actually reflects current (and well-established!) practice in the English wikipedia. A move to a format different than that should require a more extensive discussion than what just transpired here. Cheers! Impru20^talk 18:04, 21 July 2022 (UTC)

In general, do not perform a unilateral moves on any article where the move might be controversial, as was done here yesterday. An uncontroversial move, would be changing Brasil to Brazil in the title, because the former is a misspelling in English Wikipedia; no reasonable editor could disagree with this. In order to do a controversial move—which is anything more serious than a typo—please go through a move request first; see WP:RM#CM. Impru20's return of the article to the original title of "Opinion polling for the 2022 Brazilian presidential election" was the correct move; any change to this must go through a move request.

As a secondary point: whatever the article is called in pt-wiki is completely irrelevant here. The only thing that counts, is what reliable sources in English call it, along with a dollop of consistency, which Impru20 alluded to above. Thanks, Mathglot (talk) 08:45, 22 July 2022 (UTC)

a controversial move [...] is anything more serious than a typo
That is in stark contrast to the policy you just quoted, Mathglot:

A move is potentially controversial if either of the following applies:
 • there has been any past debate about the best title for the page;
 • someone could reasonably disagree with the move.

I moved it to Electoral polls for the 2022 presidential election in Brazil, and Impru20 moved it to Opinion polling for the 2022 Brazilian presidential election, which is not the original title. All good. WP:BRD, peeps; wash-rinse-repeat. — Guarapiranga ☎ 06:35, 28 July 2022 (UTC)

@Guarapiranga: The bit about 'anything more serious than a typo' is an oversimplification, perhaps, but not really in contrast to the quotation. To me, it is essentially equivalent to "someone could reasonably disagree with the move". Some editors are unable to see anything that other editors do that differs from their own view as "reasonable disagreement", so the conservative thing to do, is to interpret is as 'anything more serious than a typo'. "After all, how could anybody possibly disagree with *my* obvious choice of move to another name? Clearly I'm a reasonable person, or I wouldn't have done it, if it weren't reasonable; everyone else will necessarily think the same thing." I'm sure you see what I mean. Anyway, I'm fine with your move, and I hope you understand about the 'typo' business; it's an oversimplification that works for editors not as familiar with moves perhaps, and doesn't necessarily apply to you, so, yes, all good. Cheers, Mathglot (talk) 06:53, 28 July 2022 (UTC)

Graph

@Teratornis06: There is already a graph that is generated with a code in the article, and it gets frequently updated by many users. It's redundant to show two graphs with essentially the same information, and only you have the code to generate your graph so it hasn't been updated so frequently. In addition, your graph is incomplete because it doesn't show a line for the sum of other candidates, like the green line in the other graph. If you insist on keeping your graph in the article, I suggest adding a line for the sum of other candidates and updating your graph every few days, so in this case we can remove the other graph and keep only yours. In addition, when you update your graph, it's better to upload it as a new version of the same file instead of making a new file name. If you don't want to update your graph so frequently, I suggest removing it for now, and adding it only after the election to replace the other graph. Thank you. Heitordp (talk) 16:07, 2 September 2022 (UTC)

A Commons file used on this page or its Wikidata item has been nominated for deletion

The following Wikimedia Commons file used on this page or its Wikidata item has been nominated for deletion:

• Lula campanha 2022.jpg

Participate in the deletion discussion at the nomination page. —Community Tech bot (talk) 17:07, 19 September 2022 (UTC)

The Economist "poll aggregator" showing 50/50

So, The Economist poll aggregator has switched to show a 50/50 scenario. This wouldn't be issue, except that no poll shows a Tie or a Bolsonaro lead. At this point, no poll shows a Lula lead lower than 3 and the average margin of all five polls published since the 1st round is 5.4, so it's impossible for any poll aggregator to show such a perfect Tie except for it being an error of some kind. I hid the aggregator until The Economist fixed it, but I was reverted on grounds that '"This is probably an error of them" is not an argument' and that 'the fact is that they have the aggregation tied, until they change it (assuming it's an error) this should remain and it's not misleading'. Yes, it is an error and it is misleading since it does not reflect any of the published polls. If a decision to show it anyway is given, then we should present it as something else (a projection, an estimation or whatever), but as of now it is for sure than that cannot be the result of any poll "aggregation". That's simple maths. Impru20^talk 16:48, 10 October 2022 (UTC)

This is entirely speculative on my end, but it looks like they are extrapolating results to extend the trend line from the latest poll to the current date. If this is the case, it will probably be adjusted once the next poll is released. At any rate I agree that this isn't an "aggregation" in the traditional sense. Jacoby531 (talk) 17:34, 10 October 2022 (UTC)

Well, if they are making an extrapolation, then that is not a poll aggregator. I see that Gregorysher36 refers to The Economist numbers as a "poll" in their latest edit: if it's a poll, then its place is the main table, not together with the other poll aggregators. That's the point of the whole affair. We are misleading readers when putting this as a poll aggregator.

I don't mind adding it as a separate projection, but we should not be misleading people into something that isn't. No other polls show such a tie, so this cannot be an average of polls or anything related to other published polls. Also pinging the other involved users to the discussion since this seems to be degenerating into an edit war @91.127.69.193 @SantanaZ Impru20^talk 18:07, 10 October 2022 (UTC)

Agreed with all of your points above, folks. There simply is no second round poll that I know of which shows such a close tie between Lula and Bolsonaro, not even after the first round was actually tabulated. It is certainly a closer election than initially expected, but an aggregate supposedly showing that Lula and Bolsonaro are evenly split even though there is no such average to see from the polls we have thus far is a mathematical miscalculation to its extreme. If Gregory insists on keeping this particular poll around, then it should be moved to the main area with all the other polls, and not misleadingly marked as a calculated average, since it quite simply is not. 91.127.69.193 (talk) 20:52, 10 October 2022 (UTC)

The guy running the average is “exploring fixes” to it, so he’s clearly not very confident in the results right now. Seems best to wait until there’s an adjustment (or they fully stick behind the result it’s spitting out) https://twitter.com/gelliottmorris/status/1579474512198144000?s=46&t=PW8IDK7LxNK6NzZvo8dsjg Dingers5Days (talk) 23:18, 10 October 2022 (UTC)

They have now updated the model and it now shows a 54/46 situation, which excluding undecided is fairly similar to the current polling situation. It seems that it was an error indeed, after all (no special knowledge was needed to ascertain it). I have re-added the aggregation to the table. Impru20^talk 09:21, 11 October 2022 (UTC)

Polling aggregators?

The reference for the PollingData poll aggregator takes one to another website (veja), which has an aggregator that shows a different estimate than the one given in the main body of the article. Furthermore, the aggregator referenced in the article (https://www.pollingdata.com.br) also gives a slightly different estimate than the one listed in the article. What's up with that? 2A02:A319:8040:3000:C936:6C3E:9553:61F9 (talk) 20:22, 16 October 2022 (UTC)

For clarity's sake (because I realize that this was not very clear): I mean the part of the article that lists polling aggregators for the run-off. In particular, the third of them (PollingData) links to a different aggregator (veja). 2A02:A319:8040:3000:C936:6C3E:9553:61F9 (talk) 23:33, 16 October 2022 (UTC)

Valid poll being removed for no reason

A user keeps deleting a correctly verified poll because according to him it is not a serious pollster when it is registered in the TSE. He also threatens not to let me edit this page anymore when I have been posting polls every day. A source to the poll: https://oantagonista.uol.com.br/brasil/instituto-verita-divulga-pesquisa-sobre-2o-turno-da-eleicao-presidencial/ 2800:A4:33EE:B200:2CF1:F544:2271:EEDC (talk) 00:05, 17 October 2022 (UTC)

Second this. Verita is an established poll, and some overly-aggressive editors are threatening anyone who adds this poll onto the page. Clearly biased WinniethePooh136 (talk) 02:22, 17 October 2022 (UTC)

Also second this. Both reasons stated for its removal - disagreement with other pollsters or Verita not having previously published a poll are not standards kept on any other wikipedia election page. Wikipedia should not editorialize which polls are displayed. Otto_von_Saxony (talk) 02:48, 17 October 2022 (UTC)

Verita was not included originally, so it looks weird for that to be the first poll included. But it should’ve been included originally anyway, so yes, throw it in. However, all the previous Verita polls should be inputted as well. Dingers5Days (talk) 16:16, 17 October 2022 (UTC)

Verita must stay off, until there is genuine compelling evidence that it is a legitimate poll. We aren't going to take the word of people uttering threats and yelling at admins that it is accurate, and Wikipedia has a very strict standard on false information, a standard that your poll quite likely violates. The fact that you are using sockpuppetry to advance your claims is not helping. (2607:FEA8:7227:B323:1420:D06B:A472:255D (talk) 16:31, 17 October 2022 (UTC))

Is there a more reliable source than the TSE (Superior Electoral Court of Brazil)? The poll is deleted by users who believe it is false when it is clearly registered with an official body and therefore meets all the requirements.

Even one of the arguments used is that "since it is the only pollster with Bolsonaro as the winner, it must be false" which makes no sense. 2800:A4:33E5:8300:5D17:72A8:EE5A:B1F9 (talk) 19:41, 17 October 2022 (UTC)

You have the burden of proof backwards. Those wanting to remove a poll have to show its not legitimate. Since its registered by the TSE it should implicitly be accepted as legitimate until proven otherwise. Any other standard is ridiculous.

Also why are you going on about sockpuppeting? Otto_von_Saxony (talk) 20:08, 17 October 2022 (UTC)

Your claims are inaccurate. When it comes to polling, especially a new poll that no one has ever heard of or seen, it is on you to prove it is accurate, especially for a poll that has a slew of red flags. The TSE is not the standard we use here, it is Wikipedia's standards against false information, which are very strict. You have not provided a single shred of genuine evidence that it is a real poll, such as other surveys from this alleged poll before or during this election in the past this election, or a history, nothing that proves it is genuine. That is why it will not be accepted and screaming about bias is not going to help. (2607:FEA8:7227:B323:1420:D06B:A472:255D (talk) 20:27, 17 October 2022 (UTC))

New polling firms emerge all the the time, and this is not a standard used anywhere else on wikipedia. F.e. Quebec 2022 election Segma's poll was included even though Segma has never polled before and only released one poll the entire election. You have also not listed a single red flag, only merely asserting there are red flags which is not evidence.

Also This isnt against Wikipedia's standards against false information, because this a real poll published by a real news agency from a real market research firm. It may or may not be accurate, but the quality of the poll has nothing to do with it being "false information" or else basically this entire page should be deleted because the round 1 miss was so bad as to be laughable. Otto von Saxony (talk) 21:31, 17 October 2022 (UTC)

Also, Instituto Verita did in fact poll before, in 2018 (and its on Wikipedia no less!) and was perfectly adequate. Otto von Saxony (talk) 21:40, 17 October 2022 (UTC)

Why should I believe that the poll is legitimate? Because you say so? Listen, the poll is junk. And I can’t trust the “TSE” when I have never heard of them before. I need you to understand that this poll has not been verified by a trustworthy source. The TSA and NOT trustworthy, and could be swayed by any number of factors. I will now close by listing the red flags you seek:

1. Verita have seemingly shown up from nowhere, and have failed to release any polls before or since, except for one occasion in 2018, previously mentioned.
2. The poll claims the incumbent president to have won, even though EVERY other poll over the last few months has proclaimed otherwise. (In fact, I can make the poll make sense again if I swap the two candidates’ names, which suggests that the agency mixed up the names)

I hope these are of some use. I will also note that the round 1 “miss” was caused by third party voters abruptly switching allegiance. This CANNOT happen in the 2nd round, as it only has two candidates. Therefore, I must conclude that this anomaly is much less likely to occur in round 2. Pablothepenguin (talk) 21:52, 17 October 2022 (UTC)

1. I literally just disproved this point. Wikipedia is fine with one off polls from pollsters, and Instituto Verita has in fact polled before.

2. Polls can disagree with each other, every other contentious race has some polls showing different sides winning. Look at any competitive US election as an example.

3. Number of candidates has no impact on polling error. Polls can miss by miles with 2 candidates, or with 10. Otto von Saxony (talk) 21:58, 17 October 2022 (UTC)

1. Wikipedia needs verification, and that is more difficult when the pollster shows up from nowhere. They cannot be trusted due to lack of independent verification.

2. No poll is as off-kilter as the Verita one. Disagreeing with EVERY other poll is not acceptable, especially when all of the other round 2 polls have all borne resemblance to one another.

3. I wasn’t denying the possibility of error, I was merely proving beyond reasonable doubt that the round 1 errors CANNOT reoccur in the second round. Pablothepenguin (talk) 22:04, 17 October 2022 (UTC)

1. That is just flatly not true. You keep repeating yourself, they have polled before, and nowhere else on wikipedia has the standard of infrequent polling barred a poll from being published. They were fine in 2018 so they are fine now.

2. What are you talking about? Verita is 3 points above the Futura thats within margin of error. Additionally there are polls in the US election wikipedias like Arizona Senate Kurt Jetta showing D+17 and OnMessage Inc showing D+3 which is accepted just fine while clearly not being within MoE or any other level of tolerance.

3. You didnt prove anything. You just asserted the error was polls overestimated third party voters because third party voters changed votes last minute which is completely unsupported. Ipec missed Bolsonaro by 11 PP, thats methodological error not "last minute vote changes". 13 million people dont change their mind in a day. By your standard we should toss Ipec because they clearly have bad methodology. Otto von Saxony (talk) 22:11, 17 October 2022 (UTC)

You still have not given a scrap of evidence to prove it is real. You are in no position to state we are wrong when we are providing evidence that justifies deleting it and not using it. If you really do have proof, then present it and let us judge if that is enough to justify using that poll. 2607:FEA8:7227:B323:1420:D06B:A472:255D (talk) 22:18, 17 October 2022 (UTC)

Define real? Its objectively a poll published by a market research firm that is on the approved list by the Brazilian TSE, carried by a real polling aggregator.

Besides the point the burden of proof is on people calling the poll fake. Not on people saying the poll should be published. You have no given a single piece of evidence that the poll is legitimately fake, only that you dont like its result. Otto von Saxony (talk) 22:21, 17 October 2022 (UTC)

Stop making up rules, it makes you look like a retard. I suggest that that is not an ideal look for you. Please give it a rest, we are tired of your bullshit.

I have given you plenty of evidence, but you just won’t believe it, because it doesn’t suit your narrative. The Brazilian TSE are about as trustworthy as my dog when it comes to Wikipedia verification. Also, YouGov and Ipsos MORI are real pollsters, not some backwater Brazilian weirdo organisation that I have never seen in my life. Pablothepenguin (talk) 22:28, 17 October 2022 (UTC)

You blatantly ignore every single time I contest your "evidence". You completely refuse to engage in good faith, and completely refuse to acknowledge the double standard this would create with multiple other wikipedia pages which were completely fine to have infrequent polls.

Why do you expect to have a perfect knowledge of every single pollster ever? Also pollsters you know of is not a standard.

Literally all of your evidence is it disagrees with other polls and Verita polls infrequently. That is an insane standard we have to start deleting hundreds of polls of wikipedia by this standard. Otto von Saxony (talk) 22:37, 17 October 2022 (UTC)

I am not ignoring you, you’re just stupid. Now you’re just insulting me and you are crossing the line. It doesn’t just “disagree”, it is violently and misleadingly wrong. It is borderline propaganda and it’s veracity cannot be verified. Pablothepenguin (talk) 22:44, 17 October 2022 (UTC)

Lmao stop projecting its a bad look. I never once insulted you.

I have cited several actual examples on wikipedia where your standard isnt accepted and you refuse to accept that. Otto von Saxony (talk) 22:46, 17 October 2022 (UTC)

You did insult me, and your arguments are not going to me made any more valid by repeating them 100s of times. Pablothepenguin (talk) 22:50, 17 October 2022 (UTC)

When did I insult you pics or it did not happen. You have called me a "retard", "thick", etc. so so far this is one way street buddy.

You've never responded to my arguments once. You just ignore them. Otto von Saxony (talk) 22:54, 17 October 2022 (UTC)

I did respond to them. Repeatedly. Also, you insult me by claiming that I ignore you and accusing me of “projecting`’. Pablothepenguin (talk) 22:55, 17 October 2022 (UTC)

I don't really care about this dispute, but I have to say, calling someone a "retard" is definitely not acceptable behavior. Jacoby531 (talk) 04:18, 18 October 2022 (UTC)

1. Wikipedia needs Verification. How thick must you be to not understand that!? That was the point that I was making, that you were blatantly and conveniently ignoring.

2. Disagreeing with EVERY other poll is still wrong. Also, making up your own “margin of error” won’t save you, the poll is still off by too much.

3. Funny how the round 2 polls are much closer to the round 1 result. It’s almost as if I was right about the 3rd party candidates and their voters. Pablothepenguin (talk) 22:25, 17 October 2022 (UTC)

1. What is your standard of verification? And why is it different than 2018? Instituto Verita clearly exists, clearly does market research, and the TSE has the company registered.

2. Disagreeing with every other poll literally doesnt matter and is not a standard anywhere else of wikipedia you refuse to acknowledge any other real example of other wiki pages that were fine with polls more divergent than this.

3. Round 2 polls have no impact on round 1 results. Feel free to think you are right, but all you have is a theory with no data supporting it. Otto von Saxony (talk) 22:33, 17 October 2022 (UTC)

We have given reasons and our evidence. We have clearly stated that we need actual verification, because in Brazil, a poll could be made up by anyone, and thus just relying on a Brazilian source does not work under Wikipedia's standards. You can provide real evidence that meets the standards, or you can drop the whole argument.  (2607:FEA8:7227:B323:1420:D06B:A472:255D (talk) 22:37, 17 October 2022 (UTC))

Why dont you engage with my arguments instead of just asserting the samething 20 times in a row? If you refuse to actually engage I propose this page should just be perma locked/deleted as it will serve no purpose with people deleting polls arbitrarily. Otto von Saxony (talk) 22:40, 17 October 2022 (UTC)

We can't engage with someone who refuses to give actual evidence, otherwise they are just peddling in falsehoods and is being actively hostile, which is unacceptable. (2607:FEA8:7227:B323:1420:D06B:A472:255D (talk) 22:45, 17 October 2022 (UTC))

What are you talking about, myself and others have given plenty of evidence that Instituto Verita is a real company, and multiple other wikipedia pages allow polls just like Instituto Verita just fine! I dont know what evidence you want and you refuse to accept anything I say. Otto von Saxony (talk) 22:48, 17 October 2022 (UTC)

You havent given us any evidence that is real and geunine according to Wikipedia's standards. Now you are flat out lying to us. 2607:FEA8:7227:B323:1420:D06B:A472:255D (talk) 22:50, 17 October 2022 (UTC)

What are you talking about? What is my lie. Name it. Otto von Saxony (talk) 23:04, 17 October 2022 (UTC)

Claiming that we haven’t given you evidence is a lie. Pablothepenguin (talk) 23:05, 17 October 2022 (UTC)

Its not evidence though. You dont like the headline number, and You dont trust Brazilian sources. That isnt evidence Verita is any different than any other Brazilian poll. And once again Wikipedia was fine with them in 2018. Otto von Saxony (talk) 23:08, 17 October 2022 (UTC)

It is evidence Pablothepenguin (talk) 23:10, 17 October 2022 (UTC)

We would like independent (Preferably Non-Brazilian) sources. Verita being a real company does not exclude the possibility that it is corrupt and wrong. Pablothepenguin (talk) 22:51, 17 October 2022 (UTC)

1. Verification refers to acquiring independent sources to confirm the poll. TSE is NOT an independent source and therefore, it and anyone copying it cannot be accepted as valid verification here.

2. It does matter, as it a rare occurrence indeed.

3. Are you That thick that you don’t realise that I said that round 2 polls seem to agree with the round 1 result? You have got this the wrong way around, the round 1 result had an effect on round 2 polls, not the other way round. Pablothepenguin (talk) 22:47, 17 October 2022 (UTC)

1. So you refuse to accept the government of Brazil as an independent source? You refuse to accept Poder360 as an independent source? On what grounds are those two not legitimate?

2. Almost every single election has polling divergences of more than 3PP, Futura -> Verita is not notable. In fact its incredibly unnotable. Otto von Saxony (talk) 22:51, 17 October 2022 (UTC)

Of course, we won't accept the government of Brazil as independent. From what I understand, pretty much anyone can submit a poll, and there are no rigors checks on poll validity there like we have here. That is why we cannot use it.  (2607:FEA8:7227:B323:1420:D06B:A472:255D (talk) 22:53, 17 October 2022 (UTC))

Ok if we are ignore the Government of Brazil then cite me for every single poll on this page an independent verification or they should be deleted too. Otto von Saxony (talk) 22:55, 17 October 2022 (UTC)

Real pollsters such as YouGov and Ipsos MORI can be trusted due to years of experience and countless 1,000s of polls. That is what matters. Pablothepenguin (talk) 22:58, 17 October 2022 (UTC)

What is the verification for Datafolha, Futura, Ipespe, Atlas, Pesquisas, Gerp, PoderData, IPEC, Genial/Quaest? Otto von Saxony (talk) 22:59, 17 October 2022 (UTC)

None of them are YouGov or Ipsos MORI. The fact that these people cannot be replicated in Brazil is concerning. In US or UK elections, the polls are run by pollsters with decades of experience and are capable of being wrong, but what matters here is that they do polls weekly or even more, which allows for better comparisons. A pollster that shows up from nowhere can’t be used like that. Pablothepenguin (talk) 23:03, 17 October 2022 (UTC)

So you agree, since you cannot give me independent (non-Brazilian) verification of these polls they should all be deleted yes? Otto von Saxony (talk) 23:04, 17 October 2022 (UTC)

That is a tough one. A lot of them could be deleted, but it is complicated to work out which are good enough to be kept. Therefore, this process must be done very carefully. Pablothepenguin (talk) 23:07, 17 October 2022 (UTC)

If every poll is suspect is there a point to this entire wikipedia page? Either we accept all Brazilian polls are sus and publish them, or we dont publish any. I dont really see how we can just pick and choose here. Otto von Saxony (talk) 23:12, 17 October 2022 (UTC)

Don’t Non-Brazilian agencies ever cover this kind of stuff? Pablothepenguin (talk) 23:13, 17 October 2022 (UTC)

I've been following Brazilian politics for a few years now and I havent seen any firms from Canada/US/Europe publish polls for Brazil, at least, with any meaningful frequency. Otto von Saxony (talk) 23:15, 17 October 2022 (UTC)

1. By that logic, I should accept the Government of Russia as valid, and they’re absolutely bonkers right now. Same for North Korea, Iran, Belarus, and others.

2. This divergence is ALOT more than 3%, a figure you likely made up on the spot. The real figure is more like 8-9% Pablothepenguin (talk) 22:54, 17 October 2022 (UTC)

1. Read above

2. Futura had Lula +0.4, Verita has Bolsonaro+2.3 thats 3%. Overall average its Lula+3 to Bolsonaro+2 so 5PP. Thats not a big discrepancy Canada, Quebec, Alberta, Every single US Senate Race, etc. have similar or larger gaps. Otto von Saxony (talk) 22:58, 17 October 2022 (UTC)

1. So you’d accept Russia Propaganda then? World governments can sometimes be a little bit crazy, it is a fact of life.

2. Lula had bigger leads than that which have been conveniently ignored here. He usually leads by around 4-8%. Pablothepenguin (talk) 23:00, 17 October 2022 (UTC)

1. No? I didnt want to ask for what is independent verification in two places.

2. Even at biggest Lula+8 to Bolsonaro+2 is par for the course. We see swings that big all the time. Like I said check, Pennsylvania, Arizona, Georgia, New Hampshire senate races, US general ballot, Quebec 2022, etc. All has swings of similar magnitude from pollster to pollster. Otto von Saxony (talk) 23:03, 17 October 2022 (UTC)

A lot of those American pollsters are known to be biased. That is why they sometimes are marked with (D) or (R). This indicates possible bias. Pablothepenguin (talk) 23:04, 17 October 2022 (UTC)

Why is their bias allowed on Wikipedia just fine, but Verita, which you are accusing of being biased, has to be removed? Otto von Saxony (talk) 23:06, 17 October 2022 (UTC)

Because they are open about hteir bias and we can adjust that quickly, and the bias is only partial. In a place like Brazil, this could easily mean fabrication, aka false information. That is why we cannot use it. This petulant trolling from you has gone on far long enough. 2607:FEA8:7227:B323:1420:D06B:A472:255D (talk) 23:08, 17 October 2022 (UTC)

Because the American pollsters have been there for fears, and their biased is clearly marked, thus preventing anyone from taking them at face value. Pablothepenguin (talk) 23:08, 17 October 2022 (UTC)

Is a better resolution then not just marking Verita as (B) - deleting it while accepting other Brazilian polls seems suspect. Otto von Saxony (talk) 23:10, 17 October 2022 (UTC)

But no-one knows what (B) means, because you just made it up. And, most importantly, that (B) will not make any sense in a few decades time when completely different candidates are up for election. Pablothepenguin (talk) 23:12, 17 October 2022 (UTC)

Right I did just make it up to try to come to a compromise here. We can footnote the (B) as being potentially biased towards Bolsonaro. Otto von Saxony (talk) 23:13, 17 October 2022 (UTC)

It would still confuse people. It is not immediately obvious that it is a reference to the incumbent president. Pablothepenguin (talk) 23:15, 17 October 2022 (UTC)

If you dont like (B) we could go with (C) with a footnote marking Verita as being potentially biased towards Conservatives. With the C representing Conservative. Brazils parties change too frequently for bias to a party to be useful like in the US. Otto von Saxony (talk) 23:18, 17 October 2022 (UTC)

Utterly unacceptalbe. Until we knew for certain, we would be asking for all th polls to be hit with bias warnings. That is why we cant use it, it risks opening up a whole can of worms. 2607:FEA8:7227:B323:1420:D06B:A472:255D (talk) 23:19, 17 October 2022 (UTC)

People won’t know what (C) means either. Pablothepenguin (talk) 23:20, 17 October 2022 (UTC)

That wont work, because then it opens a whole can of worms. That is why we cannot use it. The potential of it being fake or corrupt means that we cannot use it and that has been our point all this time. Just drop it. It won't be used because it is against Wikipedia standards, and I dont care if I have to see this page protected intil the election ends. 2607:FEA8:7227:B323:1420:D06B:A472:255D (talk) 23:16, 17 October 2022 (UTC)

Im not going to drop arbitrary deletion of polls. What has come out here is the anti-Verita camp doesnt trust Brazilian verification (fair enough), but that casts doubt on every single poll on this page. We can either come to some compromise here or accept this page should be permalocked. Otto von Saxony (talk) 23:19, 17 October 2022 (UTC)

Its a large enough discrpeancy that we cannot use it because it violates Wikipedia's rules. This poll cannot go on there, and you acting hostile to everyone is not going to help. 2607:FEA8:7227:B323:1420:D06B:A472:255D (talk) 23:00, 17 October 2022 (UTC)

Ok, then toss me a compromise then. I'm not going to just accept the deletion of a poll when by the same standard Im not sure how we can trust any Brazilian poll. Otto von Saxony (talk) 23:22, 17 October 2022 (UTC)

That poll is worse, other polls may remain until a solution is found. Pablothepenguin (talk) 23:24, 17 October 2022 (UTC)

If all the polls are in doubt, then the page should remain locked until we come to a conclusion. Wikipedia cannot allow untrustworth polls and none can be verified by your standards, which if we are going to use we apply them everywhere. Either we only allow polls that come from non-Brazilian sources or we come up with another system to mark the less trustworthy Brazilian polls. Otto von Saxony (talk) 23:26, 17 October 2022 (UTC)

Obviously a new system must be invented. The page is currently locked, and I would advocate keeping this until after the election is over. Pablothepenguin (talk) 23:29, 17 October 2022 (UTC)

We dont need a new system, and we cannot make one. That is up to Wikipedia, not us, so any chance of changing it is impossible. 2607:FEA8:7227:B323:1420:D06B:A472:255D (talk) 23:59, 17 October 2022 (UTC)

I withdraw my previous remark. No system for marking polls will be anything but confusing. I still think the page should be protected, though. Pablothepenguin (talk) 00:02, 18 October 2022 (UTC)

Well until Wikipedia comes out with a system here I concur the page should be protected. I dont see a good way to resolve this dispute otherwise. Otto von Saxony (talk) 00:45, 18 October 2022 (UTC)

It would be nice to have polling info for the second round. --38.106.246.201 (talk) 21:38, 17 October 2022 (UTC)

It literally takes you a minute to google the pollster and see that they have published some polls on the election

Their official page: https://www.institutoverita.com.br/

A poll they have published before this one:

https://portaldeprefeitura.com.br/2022/09/24/bolsonaro-e-lula-estao-empatados-tecnicamente-na-terceira-rodada-de-pesquisa-para-presidente-da-republica-do-instituto-verita-veja-os-dados/

In addition, Poder360 has the Verita poll in its polling aggregator: https://www.poder360.com.br/agregador-de-pesquisas/ 2800:A4:33E5:8300:B0AB:2FC0:C58:E4B7 (talk) 21:48, 17 October 2022 (UTC)

Just because they have a page does not mean it is a valid poll. Wikipedia has a very strict procedure on this, and just because it is a poll deemed valid by Brazil's standards does not mean it is valid by Wikipedia's standards. Furthermore, Wikipedia can refuse to use polls if they are deemed with too many red flags and that is what happened here. Too many red flags to be used. Please drop it. (2607:FEA8:7227:B323:1420:D06B:A472:255D (talk) 21:51, 17 October 2022 (UTC)).

You havent listed a single red flag. Otto von Saxony (talk) 22:01, 17 October 2022 (UTC)

Actually, Pablothepenguin listed at least two red flags that are grounds for refusal, you have not given a single scrap of hard evidence that it is genuine or accurate, both of which are required to prove that it does not qualify as false information, instead you just rant about bias. We have given proof it is false and cannot be trusted, you have to give proof it is real. (2607:FEA8:7227:B323:1420:D06B:A472:255D (talk) 22:03, 17 October 2022 (UTC))

We cannot rely on Wikipedia to create a new system, especially when the current one works so well. And besides, the protection will expire in two days. By then, more accurate and valid polls will likely be out. (2607:FEA8:7227:B323:1420:D06B:A472:255D (talk) 00:54, 18 October 2022 (UTC))

Semi-Protect

Administrators, could you please semi-protect this article? I hope this request reaches you in good faith. Thank you. Pablothepenguin (talk) 00:30, 17 October 2022 (UTC)

I second that opinion. These two users, https://en.wikipedia.org/wiki/User:Otto_von_Saxony and https://en.wikipedia.org/wiki/Special:Contributions/2800:A4:33EE:B200:2CF1:F544:2271:EEDC are blatantly ignoring the red flags about this poll, and there have been no other polls released from this alleged pollster. Even worse, the two users are resorting to threats and damaging this site's emphasis on article information being accurate, which this poll is demonstrating to be heavily inaccurate, and Otto Saxony has a history of initiating edit wars on political topics as well such as the Nevada election. I ask that they both be banned, and Otto be banned permanently. (2607:FEA8:7227:B323:7581:D27E:EC2F:CC6C (talk) 00:58, 17 October 2022 (UTC)).

Furthermore, Otto has created a sock account. https://en.wikipedia.org/wiki/Special:Contributions/WinniethePooh136.

This user is using the exact same arguments about the poll to justify it, without acknowledging the slew of red flags regarding whether or not it might be a fake. I ask that both accounts be banned immediately. (2607:FEA8:7227:B323:7581:D27E:EC2F:CC6C (talk) 02:14, 17 October 2022 (UTC)).

Actually could we have this page fully protected then? IanDBeacon (talk) 02:34, 17 October 2022 (UTC)

I already requested this at RfP. Augusthorsesdroppings10 (talk) 02:44, 17 October 2022 (UTC)

Why are there no current polls shown in the article?

The latest one has been conducted already a week ago and there have to be some more recent one`s. 93.206.53.221 (talk) 20:57, 20 October 2022 (UTC)

The reason there are no current polls was that the page was protected after a user attempted to insert a poll that violated Wikipedia's standards, and for several days, no one but administrators have been allowed to edit. You are free to add the polls that have since come out, barring the Verita one, as Wikipedia has deemed that one unacceptable. (2607:FEA8:7227:B323:846:A53C:298D:FA08 (talk) 23:12, 20 October 2022 (UTC))

Banning the Verita poll makes the editors look biased because it was the first one showing Bolsonaro leading in the run-off. Deleting that one poll in particiular confirms the already existing notion that wikipedia has a left-wing bias on political issues. 93.206.55.108 (talk) 21:10, 21 October 2022 (UTC)

Wikipedia is not censored, and "Wikipedia" as an entity does not "deem" sources "unacceptable". There was no consensus that that poll violated Wikipedia's standards, as you can read above. The reason the article was locked was a particularly frothy edit war that developed over the inclusion of the poll. Blippy1998 (talk) 23:20, 21 October 2022 (UTC)

The portuguese version of the same article on wikipedia shows the latest polls. By hiding them on the english speaking site you are basically siding with Lula, the socialist and lefty in this election because the latest numbers are showing a trend favoring Bolsonaro which the left does not like. By hiding these latest numbers you are siding with the left and this should never occur on a objectivity driven encyclopedia. 80.131.59.31 (talk) 18:35, 23 October 2022 (UTC)

Anexo:Sondeos de opinión para las elecciones generales de Brasil de 2022 - Wikipedia, la enciclopedia libre

"libre" means free. Apparantly there is no longer freedom to write and communicate openly on all areas of wikipedia. 80.131.59.31 (talk) 18:38, 23 October 2022 (UTC)

Of course there is; add any poll that's missing if you want to and can find a reliable source for it. As for me personally, I have never added or removed polls because I don't know where to find them. I'm just telling you what happened earlier and trying to clear up any idea that there is some sort of moratorium on adding polls. There is not. Blippy1998 (talk) 00:29, 24 October 2022 (UTC)

This is now becoming absurd. Several polls show a close race. There is nothing in Wikipedia which says "I don't like their findings" and "I've never heard of them" are valid reasons for rejecting the inclusion of a poll. It is clear that Verita correctly picked up on a trend of the Lula lead narrowing or disappearing. Calling something a "red flag" doesn't make it so. This page should be brought into line with the page on Portuguese Wikipedia, with the same polls included. 82.5.236.90 (talk) 14:54, 25 October 2022 (UTC)

You have to add polls that are actually valid. According to Wikipedia, you cant just add polls without proving they are genuine, and polls can be removed if people have reason to believe they are false, and the site wont allow polls that go so far against the mean as Verita did. The rules are the rules, and you cannot just blatantly break them. (2607:FEA8:7227:B323:6D27:C2FC:538C:9B3B (talk) 16:41, 25 October 2022 (UTC))

There is no Wikipedia policy that says outlier polls are not allowed. Blippy1998 (talk) 20:18, 25 October 2022 (UTC)

It's not even particularly an outlier poll! If you take the 125 2nd round polls done since the start of February, and rank them based on how far they are from the average of the 10 polls before them, the Verita poll in question is not in the top 10 outliers. It is 16th.

In November 2021 there is a poll with a 7% Lula lead and a poll with a 30% Lula lead. One of those must surely go further against the mean than the Verita poll? Yet there they are.

In February 2022 a poll with a 24% Lula lead and a poll with an 8% Lula lead.

In May 2022 a poll with a 25% Lula lead and a poll with an 8% Lula lead. The poll with the 25% lead is actually the biggest "outlier poll" in the series, as the average at the time was around 14%.

So polls in Brazil are generally not very good and can span about a 16% range, though the average poll is about 3% from the rolling average. More than a quarter of polls are 5% or more away.

A poll with a 2.3% Bolsanaro lead is therefore wholly consistent with a position where other polls have a 13% Lula lead and the truth is Lula being about 5% ahead.

Does anyone have a link to any of these "According to Wikipedia" rules, or any actual reason to believe the Verita poll is "false"? The sample size is strangely large and invites questions, but are there any RS questioning the poll, as against several reporting it and the decision of Brazilian Wikipedia editors to include it? 82.5.236.90 (talk) 20:33, 25 October 2022 (UTC)

See WP:Verifiability and WP:Reliable sources for the policy and guidelines, respectively, on inclusion of sources, but also WP:REDFLAG, which is perhaps what the other editor is citing. Of course, as you said, variance is expected, and shouldn't ordinarily count against a poll under the red-flag standard; see perhaps WP:CALC for a related policy. In any case, even if the red-flag standard would have been met by this poll, the emergence of new polls in the same ballpark would qualify as "exceptional sources" under that standard. See also WP:SOURCESDISAGREE for a possible way to handle the emergence of reliable sources that contest the methods, etc. of any specific poll.

I'm sure one could talk all day about what counts as "fringe" and how reliable a polling firm is and what kind of data we should seek to include in an article like this, but the policies, as well as the practice of translating pages to English, to me seem to strongly favor updating the article to match the Portuguese version, which I will do right now. Since the Verita poll had an edit war over it, I'll add it last so that it can be reverted on its own if I'm wrong about it. Blippy1998 (talk) 21:15, 25 October 2022 (UTC)