Wikipedia talk:WikiProject Elections and Referendums/Archive 29

This is an archive of past discussions on Wikipedia:WikiProject Elections and Referendums. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Archive 25

←

Archive 27

Archive 28

Archive 29

Should there be additional rows about events in polling tables?

Latest comment: 1 month ago60 comments13 people in discussion

Hi all. Some opinion polling articles are like Opinion polling for the next Italian general election or Opinion polling for the 2024 French legislative election: you have a table where each row is a different poll. And that's it.

However, some opinion polling articles have tables where there are additional rows to note other events. With UK articles, we have a standing consensus that this should be limited to other elections (e.g., local elections, by-elections), changes in a party leader, and when an election is called. You can see this at Opinion polling for the 2017 United Kingdom general election. Other countries' articles seem to have moved to this sort of approach, e.g. Opinion polling for the next Japanese general election.

We then see what one might call mission creep! What gets a row sometimes then expands further. Opinion polling for the 2017 United Kingdom general election also includes two notable terrorist attacks during the election campaign as rows. Opinion polling for the next Israeli legislative election has all sorts of events noted. The matter has come up recently at Talk:Opinion_polling_for_the_next_United_Kingdom_general_election#Add_creation_of_Your_Party_to_the_main_table because a new party is being created in the UK, so should we note that?

I think it would be good to get a broader consensus from the WikiProject on this. What do you all think?

Personally, I favour no or few event rows. I think adding these rows is a form of editorialising, because you are saying to the reader that this is an important event that affects polling, and that's WP:OR. Most of these events don't actually affect polling and I think any big events can be described in prose text accompanying a table (with citations), rather than breaking up the table. These articles are meant to be tables of polling, not timelines of events. If people want a timeline, they can create a timeline article or graphic. But that's just me! Other views welcome. Bondegezou (talk) 10:35, 28 July 2025 (UTC)

No. Entirely agree with Bondegezou. My view on this is the same as always has been, as this is a recurrent debate in opinion polling articles, such as here and here (also a practical example here and the real example on which it was based). First and foremost, opinion polling tables are not meant as listings or timelines of events, nor as a way to project a narrative. So: why would we be adding them to the tables? Most of the time, the answer will be "because X or Y event has an impact and/or is relevant/important to opinion polling". Which is a problem, because this will mean that adding events into opinion polling tables means making an editorial decision that such events have an impact and/or are relevant to opinion polling (and implies so to any casual reader), despite this being, most of the time, a very difficult (if not outrightly impossible) situation to prove. Polling trends tend to be influenced by many events, sometimes by a chain of them, and many times their impact cannot be ascertained until weeks or even months later (at which point, subsequent events will have taken place that may have made the initial changes moot). In order to add an event, we would have to solve several issues, among which:

What does "event" mean: it's an executive order? An election? A leadership change? A whole campaign? A single protest? A chain of protests? A war? A pandemic? An announcement? A declaration? An interview? A speech? If it's a speech, it's a particular word or sentence? It has to be an event from the polled country or an event happening in another country? What if there are several inter-connected events? Etc.
What should we consider as a "impactful or meaningful" event?: A party gaining (or losing) 1 point would count as a relevant enougn impact to justify having an event for it? It has to be 10 points? 5? 20? No impact at all, just a random event that is considered as important (and if so, how is such importance determined)?
If we base the addition of events to tables on the measure of their "impact", for how much time does such impact have to last in order for it to be considered as "significant" for inclusion?: Most opinion polling trends tend to not last forever. They usually last some weeks, then subside. And for trends lasting longer, you'd typically have multiple other events that could be contributing to such a trend. A government will typically wear out after many years in power and a chain of unrelated, insignificant (by themselves) events may end up triggering a new polling trend. Also, new trends will make previous ones moot, meaning that the relevance of previous events will be cancelled by future ones.
Once we have somehow agreed on which events to add, how should they be described? What's the particular text to be added?: How an event row is described? What text is to be added? This may seem a nuisance, but practice has shown that edit wars and long discussions may spark just because of the exact wording of an event row in an opinion polling table. Is it really worth the effort?

Most of the time, you would have to end up justifying that an event affects polling with circumstancial evidence, but that's it. And that is wholly within WP:SYNTH (if not outright WP:OR) territory. In order to avoid that, you would have to list every-single-event that "may" "possibly" affect polling, which would mean you would clutter the table with events in order to avoid cherry-picking which events should be in and which should not (thus going against WP:NPOV). This is not feasible.

Some opinion polling articles provide an alternative solution, which is to add a line above the tables linking to articles that actually cover events that took place during the opinion polling period: Opinion polling for the next United Kingdom general election#National poll results links to 2025 in United Kingdom politics and government and 2025 in the United Kingdom; Opinion polling for the 2024 Austrian legislative election#Poll results links to 2019 in Austria, 2020 in Austria, 2021 in Austria, 2022 in Austria, 2023 in Austria, 2024 in Austria; and so on. In my view, choosing this over the "event row approach" would not only solve the aforementioned issue with events, but would also encourage people to edit those other articles (which at times are left underedited), without the need to duplicate information nor creating additional problems to opinion polling articles and leaving to the readers' themselves to reach any conclusions they wish without us pre-establishing what influences polling and what not for them.

All in all: no to any event rows in opinion polling tables, open to alternative proposals that help build Wikipedia stronger. Impru20^talk 13:50, 28 July 2025 (UTC)

I agree with you about political events such as executive orders or but no one (at the very least in the UK polling thread) is arguing for their inclusion and you have already been reassured about this. Whilst debating their inclusion makes this, as you correctly point out, a scale that is unreasonable to do in a completely neutral manner.

However, the true scale of suggestion (and what has been done successfully in the 2019 page, although I do understand you wish that to be removed) is much lower in scope and much more reasonable. We are talking about mentioning changes to the actual political structure of parties such as a change of leadership, creation/deletion of a party, or the merging of two parties.

This is does not make the page a list of political events, it merely explains very direct impacts and a change of Prime Minister (for example) is something you'd find on most graphical summaries of voting intention polling and they manage it perfectly fine.

As for "how do we decide what is too big or too small", we do not. If it is a party that exists within the table, it will be mentioned. If the party does not, it will not be mentioned. I'm sure you remember well, we have already decided how to judge whether a party should be included in the page, this follows the same rules.

I understand completely you wish them not to be included and believing them not to be necessary is a completely valid place to stand but these fears of violating some larger wikipedia guidelines are blown out of proportion in my opinion. Kirky03 (talk) 15:24, 28 July 2025 (UTC)

This topic is about event rows in polling tables in general, not just at the UK opinion polling articles, so that is the approach of my comment (it applies to UK articles as well, but not only). Executive orders are just one out of many examples that some users have cited in the past as "significant enough events that should be added" to these tables. Obviously, the perception of what is signicant is a very subjective one, and we can have as many "significant events" as users, if not more.

The UK 2019 example you cite is not truly "successful": I already pointed out how some of the examples you cited at the UK talk page were done without consensus and without an explanation on how they adhered to the established consensus there, so I would not say that is successful (indeed, I would have removed these right away if the issue was not being currently under discussion).

I think your reply hints exactly at my point: you say that the scope of application is much lower, but it isn't. Throughout the years (and up to the present day), some users elsewhere have used the UK opinion polling articles as a reference on why event rows should be added elsewhere (in some sort of circular argument that, just because something is done in some place, it should be automatically done elsewhere), despite that one being a very specific compromise solution made for a specific set of articles a couple years ago. Time has shown that such solution is not working properly. It does not work even for the specific UK articles themselves, as the compromise solution has frequently been used to attempt to justify including even more events than those agreed for (even to the present day: "if leadership changes are added, then the creation of a party should be added too"). For the specific issue of party creation, it was already mentioned here and here that, if a party is significant enough to get polled, it will be shown in the tables with a column. You do not need an event row to establish a party as significant, it would be a pointless and redundant feature anyway.

All those events (and any that you may think of) ultimately meet the issues and concerns I raised above. You cannot have a truly "neutral" set of events, because the issue is not the event themselves but the narrative (and, ultimately, editorialism) their addition attempts to establish. You want to consult events? Use the "[Year] in [country]" articles which are (much) better suited to list all relevant events in a given country. Impru20^talk 15:52, 28 July 2025 (UTC)

I agree that the [year] in [country] articles are useful, but I disagree on the idea that these articles mean there shouldn't be events of major significance to the election shown in the table. The biggest problem is that articles such as 2025 in the United Kingdom or even 2025 in United Kingdom politics and government are highly verbose and actually include far too much information to effectively replace the listing of only the most important events, as is currently shown in the opinion polling articles. When trying to follow opinion polling in relation to the most significant events like party leadership changes, removing the events from the list would make this very difficult. Those most significant events being included improves readability, and often provides important context to the numbers shown (e.g. with events surrounding Liz Truss in the 2024 UK polling article).

In my opinion the events should be kept, and I maintain that it's reasonable to consider a new party creation (of a party significant enough to have its own column) to be more significant than a leadership election. I think it only makes sense, given the existing "compromise" standard. It's a leadership change (from no-one to someone), and a creation of a new party, combined into a single event. Looking at the 2019 UK polling page, I'm not sure that specifically the closure of candidate nominations should be included, although I'm sympathetic to everything else that is shown there, which currently seems to be leadership changes, party creation, major events like the beginning of the campaign period, and by-elections or similar. I think these four categories at least should remain. Eastwood Park and strabane (talk) 16:46, 28 July 2025 (UTC)

I think I'd also personally add an event to the 2024 UK polling article on 4 January 2021, for the renaming of the Brexit Party to Reform UK. Eastwood Park and strabane (talk) 16:53, 28 July 2025 (UTC)

Events listed at 2025 in the United Kingdom and 2025 in United Kingdom politics and government are no more than 2-3 lines long (and most are 1 line long). I cannot see how this is "highly verbose"; I'd say that, in most cases, this is the required info to properly describe an event. You cannot do that with event rows without distorting the whole table.

"actually include far too much information to effectively replace the listing of only the most important events" Just to remind everyone here: opinion polling tables are not timelines of events. The "[year] in [country]" do not "effectively replace" any listing because opinion polling tables are not meant as listings of events. There is a misconception on what opinion polling tables are meant for, here.

"e.g. with events surrounding Liz Truss in the 2024 UK polling article" It is so? The death and state funeral of Elizabeth II is not mentioned. Neither is the mini-budget, nor the pound sterling fall, nor the IMF criticism of UK fiscal policy, nor the government's cancellation of their plan to abolish the highest income tax band, nor Truss' Conservative Party Conference speech (together with the Greenpeace protests), nor the rail workers' strikes, nor the Bank of England's warning of a material risk to financial stability, nor Kwasi Kwarteng's dismissal, nor Jeremy Hunt's emergency statement, nor the events surrounding the chaotic 19 October 2022 parliamentary votes. Do you think these events are not needed to provide context? Do you think we need to add them? Do you think an opinion polling table is suited to include these and other events "to provide context to readers"? And these are just a handful of examples related to the specific case you brought: I could go on with many other events related to this and other cases.

"When trying to follow opinion polling in relation to the most significant events like party leadership changes, removing the events from the list would make this very difficult." Opinion polling tables are not meant for this. You are acknowledging that the purpose of event rows is to create a narrative that readers should follow, which is exactly what I am criticizing, and is exactly what poses a problem. Who determines the narrative? Who determines what are the "significant events"? Who determines how should readers follow opinion polling? In relation to which events? Even leadership changes may prove insignificant in opinion polling trends, or easily eclipsed by other events.

"I'm not sure that specifically the closure of candidate nominations should be included" This may be one of those events that some user added because they thought that it "added context". Just as you may think that some other events may "provide context". And another one will think that other events do "provide context". See my initial comment as to why this may (and does) pose a problem. "I think I'd also personally add an event to [whatever]" And yes, this just highlights why this is a problem: we cannot keep adding what everyone does personally think that should be added. This is exactly the point I have made in my initial statement. Impru20^talk 17:50, 28 July 2025 (UTC)

I shouldn't have said "verbose", rather I should have said "excessive". Most of the events on the list are trivial, the actually important events for the next election such as leadership changes are lost in the sea of "random councillor resigns seat", or "politician said something". I agree it's difficult to get a consistent standard for events that doesn't get blown out of proportion, but I think we've got a fairly good balance as is - the number of events is miles away from being a problem right now. I think the Israeli article is an example of too many events clogging up the actual content. Eastwood Park and strabane (talk) 21:45, 28 July 2025 (UTC)

You don't even have to go to the Israeli example. Take Opinion polling for the next Australian federal election, which allegedly only shows the "minimalist approach" of events (leadership changes, party creation). It depicts five events in a row (some of them cancel themselves) without any intermediate poll. What impact do they have? How are they relevant? What context do they provide to opinion polling? Why are those even selected?

And you still miss the key point. You speak as if events should be taken for granted as a feature in opinion polling tables, as if these were designed for that, but they weren't and that isn't their purpose. The correct balance for something that doesn't belong in a place is for it to not be placed there. Impru20^talk 22:16, 28 July 2025 (UTC)

That looks fine to me, those are all major events that occurred right after the previous election, so makes sense they were consecutive. "but they weren't" They're just wikitables, you can design them to include whatever you want. "The correct balance" That's your opinion, there's no objective right or wrong. If you want to get rid of excessive details, then these tables don't need to include whether every poll was online or by telephone or what the same size is. We're already presuming that these are reputable pollsters who strive for a representative sample and moderate margin of error (right?), so these aren't really significant details or informative to most readers like these contextual notes can be. — Reywas92^Talk 23:19, 28 July 2025 (UTC)

The difference is that you cannot fully provide context for an event (or chain of events) with the design of a wikitable for opinion polls. They are not intended or designed for that, and it is not that you are proposing a design revamp either. The SYNTH and NPOV concerns are not even remotely addressed, which I find troubling. Impru20^talk 06:53, 29 July 2025 (UTC)

However, in the event of the creation of a whole new party, whatever they are polling at is by the nature of its existence changing polling. It is also, if I may be blunt, common sense that a change in leadership would.

But there are also more complex changes that may need to be recorded. If two parties merge (or a party splits in two), it may help the reader of the article to be able to see a note telling them that that happened so that they can better interpret the information.

It is a small matter for these articles which ultimately are incredibly negligible but can provide useful bits of information that make the articles better for their users. Arbitrarily using wikipedia guidelines in dubious ways ignores the reason they exist. They exist to make wikipedia better for the readers, not the editors, and obscuring incredibly relevant information does not create a better experience for the user.

Does it really hurt in any way for these not to exist? Kirky03 (talk) 15:30, 28 July 2025 (UTC)

You do not explain why do the opinion polling tables need to explain or record that, when that is not their purpose. This is ultimately the issue that almost no one seeks to address when confronted with it. Also, when you say abitrarily using wikipedia guidelines in dubious ways seems a serious accusation: what are the "dubious ways" here? I could say the same about using opinion polling tables in "dubious ways" and for purposes that are not their own. You yourself acknowledge that adding such event rows seeks to lead readers to specific interpretations, which is exactly what I am countering (and denouncing) here. Impru20^talk 15:53, 28 July 2025 (UTC)

Why do we do anything? Why do we colour code the parties, or put the sample size, or (in the case of UK polling) whether it is UK or GB? Why, why, why? We do them to make a better experience for the user where all the relevant information is easily obtainable. The balance is between ensuring the information is there without clutter.

What is currently in place provides information without adding clutter. There is little reason to remove.

Also, to add further context to my comment on wiki guidelines being used in dubious ways. I was simply trying to say that they are only relevant in a strenuous way and so are at best a thought to be considered rather than a leading argument in of itself. I am not sure it is original research to have a note explaining that a new party was founded on x date and that's why a new column suddenly appears following that date. It is simply clarity. Kirky03 (talk) 12:44, 29 July 2025 (UTC)

Yes It should be limited to items directly relevant to the election such as changes of party leaders, but these events can often provide relevant context of the election timeline like what parties might be included in polling and major swings so they should not be completely excluded. The events in the Israeli election article are mostly quite relevant because they directly explain changes in what might be included in the Government bloc or Opposition bloc or what parties are being polled in the first place; however, the resignations of some members are not very explanatory here and should be removed or limited. Reywas92^Talk 15:46, 28 July 2025 (UTC)

Just as stated above: all of these purposes seem legit, but that is not the opinion polling tables' job to do. If this is relevant, explain it in prose in the relevant article. You cannot explain the full context of an event in a single row in a table, and you can get too many "relevant events" that you may feel that need to be added "to provide context". There are better (and more policy-compliant) ways to do that, but not in a table whose purpose is a different one. Impru20^talk 15:58, 28 July 2025 (UTC)

I disagree. These are unobtrusive and informative, and readers shouldn't have to cross-check the campaign prose with the table to get the context. Per WP:NOTSTATS, context should be provided for statistics, and that's better within the polling page than only on the main article. There can be better limits on what type of events to include as is appropriate for that country's politics, but there should not be a ban. Nor can a sweeping consensus be made here, especially with no other pages notified about this. — Reywas92^Talk 18:20, 28 July 2025 (UTC)

I should remind you that opinion polling articles are content splits from the main election article, due to size. The main election article is still (and should be) the primary topic of information regarding the election and polling trends. If a polling trend is relevant for election context then it should be included into the main article. If you do not wish to include it into the main article, then it is not relevant for the election and should have no place elsewhere either. Further, event rows do not even properly give context or analysis: they are purely decorative, while forcing readers to interpret that such event has an impact in polling without even caring to explain how and why. If you cherry-pick a few of them, even if individually sourced, you can lead readers to your preferred narrative without the need of sourcing that narrative as a whole (that is a blatant WP:SYNTH violation). And yes, Opinion polling for the next Israeli legislative election is a perfect example of what is wrong in including such events (some parts of the tables have more event rows than polls): what context do these provide? What impact do these have in polling? Why is each of these events relevant? Where is even any of it explained? Of the aforementioned concerns I raised above, no one is making the slightest attempt at addressing any of them. Impru20^talk 19:08, 28 July 2025 (UTC)

Oh, and I would agree that a consensus in any way would require a properly formatted RfC (specially considering this affects a wide number of articles), but I think this was aimed at getting some preliminary input, though. Impru20^talk 19:41, 28 July 2025 (UTC)

I agree with your judgement about member resignations. Those are unnecessary amounts of detail. Kirky03 (talk) 12:48, 29 July 2025 (UTC)

Yes, within reason. The UK and Israeli pages brought up here definitely have gone to an extreme; these pages supposed to be for Opinion polling, not an exhaustive list of every event between elections. That's more what in the Next Foo election Background or Timeline sections are for, and in the case of by-elections and floor crossings, even nth Foo Legislature. But some events can be relevant and useful information for readers. In countries where politics are very leader-centric, a party changing leaders is both a significant event in the lead-up to an election, and one that can have an impact on the party's polling. Likewise, it seems reasonable to highlight the formation of new parties, assuming they're ones that get media attention and appear in polling surveys. Anything more than that should be included sparingly, and only events that are truly exceptional: things like war, massive and sustained protests, a coup attempt, etc. that could reasonably be expected to significantly affect public opinion. One editor suggests linking to 2025 in Foo articles offers a compromise, but I don't think this is actually very helpful: if a reader is going through the data here to see how a leadership change affected something, or is curious about a sudden change, they're going to have to open two tabs are cross-reference the pages. That just sounds cumbersome and obnoxious to do. Again, we should absolutely be culling which events we put in here, but I think a blanket ban on all events is throwing the baby out with the bathwater. — Kawnhr (talk) 18:22, 28 July 2025 (UTC)

"could reasonably be expected to significantly affect public opinion" And how do you determine this?

"if a reader is going through the data here to see how a leadership change affected something" If you think a reader needs to see how a leadership change has affected something, and you have sources backing up that claim, then by all means add such bit of info into the context/background section in the main election article. Opinion polling tables are not meant to lecture readers into how X event affected something (and indeed, adding an event row does not even properly address this, since you cannot explain the full context in a single row). Basically every "Yes" !vote is (indeed) falling into the same issues I just denounced in my initial comment. Impru20^talk 19:01, 28 July 2025 (UTC)

Have you read WP:BLUDGEON? I'm just curious. — Kawnhr (talk) 23:15, 28 July 2025 (UTC)

I am merely asking people to be coherent and answer simple questions that were raised in the process and that affect nuclear points of the issue at hand. Worry not, I will assume you just cannot. Impru20^talk 06:49, 29 July 2025 (UTC)

(And yes, I am attempting to back down on this to write less, but it is fairly frustrating how people here seem to be falling under the exact same problems that have plagued this issue for years, that I have raised on the basis of my own experience with this issue and that, despite all of it, it does not even look like a slightest attempt to read what I said is being made). Impru20^talk 07:03, 29 July 2025 (UTC)

I Appreciate you mentioning under which context leadership changes/new parties should be mentioned. It is an important aspect. I would personallynargue this should be every party listed in the columns and no party not listed, that way we avoid any amount of original research that goes beyond that which determines which parties get a column.

I'm not personally affronted by the idea of including major events such as a war however this is what gets a bit subjective and is the issue that other editors have been raising, so I would argue against their inclusion for that reason. Kirky03 (talk) 12:57, 29 July 2025 (UTC)

Yes, within reason - If reliable secondary sources such as national newspapers indicate an event, such as a change of leadership in a political party or a split, had a major effect on polling, then that is sufficient to include that event in the table and help give the reader context to major shifts in polling. CeltBrowne (talk) 00:19, 29 July 2025 (UTC)

Yes Coming from someone who mostly edits Eastern European polling pages, these events are really important to the understanding of any user of the set-up of the table and the polls. For example, in Czechia there has been quite a complex process of coalition forming/joint lists. I understand that this could be explained in the main page (and it still should be), but equally when the information can be right there and provide a short helpful explainer, that is no bad thing. Or, in another example, take Hungary, where parties have begun to announce they are not running in the upcoming election; especially when few pollsters to begin with include the minor parties, it is not obvious to see when they ceased to run and makes it easier for the user to locate it with an event in the table. I do not see the point in removing information like this when it is just helpful to anyone who arrives on the page. Quinby (talk) 07:24, 29 July 2025 (UTC)
Never thought about that, but I think this is useful context! iamthinking2202 (please ping on reply if you would be so kind) 09:16, 29 July 2025 (UTC)

Yes, but only if multiple RS report that an event influenced the polls.. Koopinator (talk) 07:44, 29 July 2025 (UTC)

I don't think that's possible to establish in all cases, and might create some unusual events that by all other metrics shouldn't be there. Quinby (talk) 07:50, 29 July 2025 (UTC)

I considered this standard myself but I think I agree with the response above - It'll likely create a flood of events when opinion polling changes, so I think we should stick to the four that seem to be included already - leadership changes, party creation (controversial - see above), by-elections and similar, and important dates such as the parliament dissolving. I'd add to this list the beginning of a major war, particularly one that involves the territory of the country being polled for as has been done in the Israeli and Ukrainian articles. Eastwood Park and strabane (talk) 09:50, 29 July 2025 (UTC)

I also agree.

It's a good solution but I am concerned it is open to the issues others have been raising when backing a no stance. Even if those issues don't come to pass, we probably need to be cautious about creating a solution everyone can be happy with which unfortunately I don't think this would be. Kirky03 (talk) 13:03, 29 July 2025 (UTC)

Yes event rows within reason. Agree that eg Opinion polling for the next Australian federal election has it excessive amount right now, but even then I think it’s just a matter of being flexible and cutting where needed, rather than outright bans.

I think it is helpful providing context - eg the pre-campaign polls for Opinion polling for the 2025 Canadian federal election, and I think it wouldn’t be good to prevent these markers being added here and there. It also isn’t a major change visually, unlike eg changes between TIE and TILE election infoboxes iamthinking2202 (please ping on reply if you would be so kind) 09:15, 29 July 2025 (UTC)

Yes but only if its affect the polls Braganza (talk) 10:58, 29 July 2025 (UTC)

Yes, I've added my opinion further up, I personally think there are four categories that I would include.

Leadership changes/party changes (e.g. new parties founded or party renames)
Major events effecting the country (like invasions or similar)
By-elections and similar (like local elections)
Election related events like the election being called

I agree that sometimes they go too far, but I actually am not sure if most of the examples mentioned have. The Australian example is unfortunate because there were so few polls in the chaotic post-election period, I'm not sure if anything included here is unjustified though. Maybe Dutton's resignation itself could be merged with Ley's election? Eastwood Park and strabane (talk) 11:14, 29 July 2025 (UTC)

Post-election periods tend to be chaotic (typically, parties will see leadership changes and such most frequently in this phase), whereas very few or no polls at all will be published (because interest in conducting opinion polls temporarily wanes during this time). We cannot say "X example is unfortunate" while still defend the value of adding event rows, because those unfortunate examples are the consequence of enforcing event rows into opinion polling tables (yes, adding event rows has many undesirable and unfortunate consequences, so of course you have to account for them). Also, of these four categories, the first and second ones are wildly problematic ("leadership changes" and "party changes" when it comes to decide which parties are considered as relevant enough to have their leadership changes mentioned, if we account for both resignations and appointments, etc.; "major events affecting the country" is a wildly broad category subject to interpretation and synthesis, probably the most problematic category by far). While this has been discussed ad nauseam throughout Wikipedia, Talk:Opinion polling for the 2022 Australian federal election#RfC on commentary rows together with Talk:Opinion polling for the 2022 Australian federal election#Macron claim that Morrison lied and its relevance to voting intention provide an example of a particular "major event affecting the country" being discussed. Impru20^talk 11:29, 29 July 2025 (UTC)

I would say that determining which parties are relevant is not a problem, it can be narrowed down to the parties that are regularly polled and as such have their own column in the table. If a particular figure forms a new party or becomes its leader, and then the party is polled, that event can be retrospectively added (or often, as there has often been hypothetical polling, added at the time e.g., Omtzigt, Magyar). Quinby (talk) 11:59, 29 July 2025 (UTC)

For party formation/changes: if it is narrowed down to the parties that are regularly polled, one can ask what the purpose of such event row is if you are gonna have a column for that party anyway... wouldn't it be a redundant and unnecessary event row? What info would it provide, exactly? Wouldn't it be already visually evident with the additional column?

For party leadership changes: would regional parties that get polled regularly (but that poll consistently around 1-2% of the vote) be considered as "relevant enough" to get their leadership changes mentioned, even if the leadership change itself is not relevant in reliable sources related to/mentioning the country? Also, leadership changes do not necessarily take place simultaneously: frequently, it is the leader announcing his/her departure, then his/her formal resignation, then the leadership contest, then the formal appointment of his/her successor. If that leader is a prime minister, then the date of his/her appointment as PM may not be the same as their appointment as party leader. This could account for four/five event rows alone if these do not take place on the same day. Should event rows for all of them be included? If yes, why are all of them relevant? If not, which criteria are you going to establish to determine which ones are relevant/have an impact in polling and which ones aren't/haven't?

This while other issues still remain unaddressed. Impru20^talk 12:58, 29 July 2025 (UTC)

I think you are confusing arguments, I don't think 'polling impact' is measurable from events nor should it be used to determine whether something should be added to the page. The events should be for user understanding of the relevant parties shown in the table, and the elections that go on. The only other events I can imagine including are ones that pause the campaign, such as terror incidents as briefly mentioned above. Quinby (talk) 13:07, 29 July 2025 (UTC)

So, who decides "which events should be for user understanding of the relevant parties shown in the table"? Then, if it's only for "understanding of the relevant parties", surely that information can (and should) be in the main article of the election, so I do not understand how its presence in the polling tables is relevant. I see some attempt at squaring the circle here: somehow we should aim both to have some events in the polling tables to "provide context/information", while being careful at it becoming absolutely chaotic, but the question remains at how would you determine which events "provide context" and which ones do not. Particularly, when having event rows will inevitably lead to other users jumping to add event rows of their own. We do not need a discussion to reach that situation as that is the situation at present.

This, this or even this are perfect examples of what should not be done: event rows being added (and, at many times, stacked one on top of the other in a long continuum of events without any opinion poll in-between) that provide little to no useful context to any casual reader, while cluttering the table and making it more complex to read and navigate. Here we have up to ten event rows compared to just one poll. Impru20^talk 14:56, 29 July 2025 (UTC)

I would be perfectly fine agreeing to remove events from all tables that are not the main polling table, that's just repeating content, and leads to some of the problems you outline. Please see below for my proposed categories Quinby (talk) 15:53, 29 July 2025 (UTC)

Well, I am against those categories and I think I have thoroughly explained why in my numerous comments. I will not repeat myself, but those fail to address any of the raised concerns (and indeed, leave way too much room to interpretation, which is precisely the problem that Bondegezou and myself have denounced). Impru20^talk 16:07, 29 July 2025 (UTC)

Yes, to some extent I don't think a hard and fast rule is achievable on this, since different countries inevitably have different event and political situations, which may vary different treatment. I would say that in general the tables are not supposed to contain a list of every event that can possibly affect polling. I would say it should be reserved for events that are so significant that knowing about them is neccesary for understanding subsequent polls. Not all leadership changes neccesarily qualify (particularly for minor parties or immediate post-election changes), but some do, particularly if it's in the ruling party. Joint lists/coalitions/parties being formed or dissolved do qualify in my mind, at least if the parties in question are politically significant. Also for other events, it should be a case-by-case assessment. Personally I would use the format sparingly. Gust Justice (talk) 13:39, 29 July 2025 (UTC)

Yes, within reason – as per the rationale of the users above DimensionalFusion (talk · she/her) 18:26, 29 July 2025 (UTC)

Arbitrary break

There has been a lot of discussion, so thanks to all for their contributions. I'm adding a break to make it easier to continue the discussion. Just as a summary so far (and I hope I've not misrepresented anyone), broadly there are three camps. Myself and Impru20 favour no event rows. The largest group (Kirky03, Eastwood Park and Strabane, Reywas92, Kawnhr, iamthinking2202, Quinby) favour a limited set of event rows, broadly what has been the consensus on UK articles to date (leadership changes, party mergers, etc.), although with maybe some disagreement on the precise details. Finally, CeltBrowne and Koopinator favour event rows when there is WP:RS that it has affected polling; Braganza is possibly also in this group. Bondegezou (talk) 12:23, 29 July 2025 (UTC)

I would describe that as a fair characterisation.
I would propose the following categories building on those from @Eastwood Park and strabane

Party changes: the date a new leader takes over (discussion over whether resignations should also be done, especially as they often happen immediately after elections), the day of a merger/split of a party, or the day a party announces it is not running. Name changes can be covered with a note next to the party's name/logo in the table.
Election-related events: this can either be done (depending on the page/integration) in the table or by separating tables, e.g., having one for the campaign period.
Other elections: by-elections, local elections, supranational elections (though these often are included with their results, e.g., EU). It is worth saying that many by-elections (especially in the non-anglosphere world) are not covered on ENWIKI, so this may be harder and require more country-specific consensuses (also if by-elections are seen as particularly important potentially).
Country changes: changes of government and the leader of the government. This one I am not sure about its full practical implications.
Major events: events like invasions (e.g., Ukraine) should be included. This can be done on an ad-hoc basis, as (one hopes) events like these do not come about regularly. This should also include events like the killing of a candidate/politician (e.g., 2016 UK referendum, 2023 Ecuador presidential), which can pause campaigns.

I know this potentially throws up quite a few questions, but I think considering the plurality support some events, we should agree on the buckets those events should fall under. Quinby (talk) 13:20, 29 July 2025 (UTC)

If we have party leader changes, I think it should just be the day the new leader takes over. We don't need, as some have added in the past, a separate row for when the prior leader resigns. I note Gust Justice's comment above and also feel that changes in the leadership of minor parties seems less relevant, but then it gets complicated deciding which leaders count. Bondegezou (talk) 13:53, 29 July 2025 (UTC)

I think the main instance where leadership changes is kind of irrelevant is where a party has two leaders (e.g. Green Party of England and Wales) and one (not both) of its leaders resign. In those cases it doesn't seem like a pivotal event. Same thing for political parties where the "official" party leader is not really the face of the party. Gust Justice (talk) 14:01, 29 July 2025 (UTC)

If we have party leader changes, I think it should just be the day the new leader takes over. We don't need, as some have added in the past, a separate row for when the prior leader resigns. I think this depends on how quickly the replacement is named. If we're looking at a country like Australia, where a leader stands down and is replaced within a week, you might be right that just a single event (for the new leader) is sufficient. But in Canada, leadership races can take years, during which the party is led by an interim leader… so not mentioning the first leader's departure can be a little misleading, as if they had been the face of the party when they actually hadn't. — Kawnhr (talk) 16:30, 29 July 2025 (UTC)

In prose, you can easily explain this situation in a single, continuous line. In a table, you need to have at least two separate rows at different dates that may or may not be contiguous (and, if they are not, how are you even expected to "provide context" years apart? Are readers supposed to know the separate dates in which each event takes place and make the connection themselves?).

Most, if not all, of the problems discussed up until now come from the very simple fact that such tables are not meant nor designed to explain or give any written context. So yes, it is prone to be misleading either way. Impru20^talk 16:59, 29 July 2025 (UTC)

Sadly, we cannot nor should not be providing the full context behind such things for reasons you've pointed out but that doesn't mean we should not provide what we can. Kirky03 (talk) 17:27, 29 July 2025 (UTC)

I am really not seeing the objection here. How is having two rows years apart (separated by several polls, if not dozens) excessive clutter? How is "January 15, 2017: Smith stands down as leader" and "March 8, 2019: Jones becomes leader" supposed to be impossible for readers to decipher? But I also didn't use the words "provide context", so maybe this comment is in the wrong place. — Kawnhr (talk) 19:02, 29 July 2025 (UTC)

Well, I have not pronounced the work "clutter" in this section at all, and indeed the comment you are replying to relates to an entire different issue, so is it possible that it is your comment, not mine, that is in the wrong place? As for providing context: wasn't that the reasoning behind adding event rows? What is their purpose, then?

Answering Kirky: we can (and should) indeed provide the full context whenever needed. You can do that in the main election article, where it should be, and where you can do it in an easier and more effective way. Obviously it is sad that opinion polling tables cannot fully meet our desires in that regard, but maybe that is because such tables are not meant nor properly designed to provide such context in prose. Impru20^talk 19:48, 29 July 2025 (UTC)

Oh my god, I'm not playing these games with you. Peace. — Kawnhr (talk) 21:36, 29 July 2025 (UTC)

Which "games", if I may ask?

Earlier on, you accused me of bludgeoning without even caring to address any of the points I legitimately raised in the discussion.

You then accused me of placing a comment "in the wrong place" after claiming it stated something it did not. Again, without addressing the (legitimate) points I raised within it.

You now say I am "playing games", again avoiding to actually reply to the comment's content. I dare you instead to play the game of assuming good faith.

You are free to ignore me if you do not wish to address my points, but please stay on topic without making any personal considerations on other editors. Thank you. Impru20^talk 22:30, 29 July 2025 (UTC)

I suppose the general solution would be to consider the moment an interim leader takes over to be a "new leader" however really we'd have to just trust the individual editors. Kirky03 (talk) 17:26, 29 July 2025 (UTC)

As for which leaders count, I don't think this has to be complicated. Just the parties that get polled and are included in the table already. — Kawnhr (talk) 16:31, 29 July 2025 (UTC)

This seems to make the most sense. Kirky03 (talk) 17:23, 29 July 2025 (UTC)

I'd personally include within the party changes section the founding of a new party or renaming of an existing party, the former being why this discussion came up in the first place. Eastwood Park and strabane (talk) 21:20, 29 July 2025 (UTC)

Hatnotes on state-level US presidential articles

Latest comment: 1 month ago1 comment1 person in discussion

I would like to notify members of this Wikiproject that the vast majority of state-level US presidential election articles use {{main}} as a hatnote. The template is not meant to work like that; it is meant to signify daughter articles of a subject related to a heading within the article. Also, multiple articles could claim to be the parent article of the state-level US presidential election articles. I have been chipping away at the templates for a while but over a thousand articles are affected. ✶Qux yz ✶ 21:15, 30 July 2025 (UTC)

United States House of Representatives - seat gain in state represented with an infographic instead of a table

Latest comment: 1 month ago1 comment1 person in discussion

I'm sorry if I come off as uninformed, but I really can't find anything about it on this talkpage. Is there any reason for the US HoR elections 1924-2006 not having a net seat gain per state table? Instead, the seat gain in each state is presented with this awkward infographic where the states with 1-2/3-5/6+ gained seats are being lumped together. osuh (talk) 23:43, 30 July 2025 (UTC)

2025 New York City Democratic mayoral primary has an RfC

Latest comment: 1 month ago1 comment1 person in discussion

2025 New York City Democratic mayoral primary has an RfC for determining the colors used for electoral result maps. A discussion is taking place. If you would like to participate in the discussion, you are invited to add your comments on the discussion page. Thank you. Einsof (talk) 23:44, 30 July 2025 (UTC)

MOG:GEOCOMMA in local elections

Latest comment: 1 month ago1 comment1 person in discussion

There is a proposition to amend the examples in MOS:GEOCOMMA for local elections. You'll find it here: Wikipedia talk:Manual of Style#MOS:GEOCOMMA in article titles. HandsomeFella (talk) 19:43, 31 July 2025 (UTC)