Talk:2024 Croatian parliamentary election

Polling Graphs
I'm just putting this here to stop the potential edit war. I believe there is a dispute between primarily @CoaxandBotany and @Number57 on whether we should use the inbuilt wikipedia graph, or a LOESS image version. Personally, I am in favour of the image graphs, since they work across devices and mediums, and they show more of a general trend (while the inbuilt graphs are heavily distorted with outliers). Especially if they are regularly updated, they can be an improvement to most wikipedia election/polling pages. In an attempt to build a consensus, please reply with whether you are in favour of the inbuilt graph or image graph in this case: Quinby ( talk ) 14:56, 7 January 2023 (UTC)


 * @Quinnnnnby the image graph is better because the inbuilt one seems to overfit the data and produces very poor regression lines. Also image graphs are more common place as far as I can tell. CoaxAndBotany (talk) 14:58, 7 January 2023 (UTC)
 * The graph is clearly preferable to the image, as it can be updated more easily by editors and has highlightable trendlines. Number   5  7  14:59, 7 January 2023 (UTC)
 * @Number 57 whilst I agree that it is more easily updatable, I can assure you that I will very regularly update the graphs I produce! also the highlightable trendlines are quite unhelpful since some of the party trendlines are already red CoaxAndBotany (talk) 15:02, 7 January 2023 (UTC)
 * Disagree. The standard is to use image graphs if available (see Opinion polling for the November 2019 Spanish general election, Opinion polling for the next German federal election, Opinion polling for the next Finnish parliamentary election etc), because the regression lines of the inbuilt graphs are relatively poor fits. Unless the creators of the image graphs stop updating them, I see no reason to keep the flawed inbuilt ones. --Jamaika-Koalition (talk) 15:04, 7 January 2023 (UTC)
 * If you don't like the trendlines produced by the graphs, why not request that the module is upgraded to improve them? Hard images are never a good idea where there is an alternative. And re CoaxAndBotany's point, we cannot rely on editors always being around to update graphs (I regularly hear claims from editors that they will regularly do stuff (to justify getting their way) and then they lose enthusiasm soon afterwards. Number   5  7  15:07, 7 January 2023 (UTC)
 * @Number 57 I guess we could let me update the Images, and if they aren't updated within a week of a poll, you can revert to your method? CoaxAndBotany (talk) 15:09, 7 January 2023 (UTC)
 * No, that's not appropriate. We cannot rely on a single editor doing something until they choose to stop. Number   5  7  15:10, 7 January 2023 (UTC)
 * @Number 57 this is literally how most Wikipedia polling pages work and most without any issues whatsoever! CoaxAndBotany (talk) 15:18, 7 January 2023 (UTC)
 * also as quinby points out, the wiki graphs do not appear on the Wikipedia mobile app whereas Images do CoaxAndBotany (talk) 15:09, 7 January 2023 (UTC)
 * Again, this is a bug that needs to be reported, rather than just avoiding using something. Number   5  7  15:10, 7 January 2023 (UTC)
 * i would personally prefer the image since this created chart is kinda too swingy Braganza (talk) 15:40, 7 January 2023 (UTC)
 * @Number 57 If the wiki graphs worked properly and used a better statistical method, I don't think many would be opposed to them. However, this is not the case at the moment, and the alternative produces a much more useful, readable, and statistically sound graph. If it doesn't work out because they aren't regularly updated and the author doesn't respond to contacts on their talk page, I would understand if it were to be removed. As such, taking into account both this and the precedents set in many countries, I think it is best to have the image Quinby  ( talk ) 15:46, 7 January 2023 (UTC)
 * Again: If you don't think the output from the graph module is good enough, request it be upgraded. Number   5  7  15:51, 7 January 2023 (UTC)
 * @Number 57 so whilst we wait for an upgrade...let's use the commonplace images! CoaxAndBotany (talk) 15:57, 7 January 2023 (UTC)
 * Have you requested the upgrade? Number   5  7  16:06, 7 January 2023 (UTC)
 * @Number 57 can you upgrade it. meanwhile, given the consensus from everyone here, I think I'll revert to my graph. CoaxAndBotany (talk) 17:03, 7 January 2023 (UTC)
 * No, I cannot upgrade it, as I don't know how it works. You need to request the changes at Module talk:Graph. The fact that you are unwilling to make the slightest effort to resolve the claimed issue with the graph module suggests to me that you are not approaching this matter in good faith, and simply want to use the image you created. You say the graph module is more easily updatable, so why not try and make it work the way you want it too? Number   5  7
 * @Number 57 this issue was already bought up multiple times, most recently in May 2022, with no response. It seems that there is an unwillingness to introduce this. However, I have reported the mobile app bug. Quinby  ( talk ) 17:24, 7 January 2023 (UTC)
 * There. Are we happy to revert to static images as is precedent in most major election pages now? Quinby  ( talk ) 17:37, 7 January 2023 (UTC)
 * No, not until someone confirms that the requested changes can't be done. Why the desperation to put your images in articles ahead of the graphs? Number   5  7  17:38, 7 January 2023 (UTC)
 * Considering there have been posts from 2021 asking for LOESS with no replies, we may be waiting for a while then.
 * The images are simply more useful in understanding the polling, as they are less sensitive to outliers and display overall trends. The general consensus that you asked for does seem to favour these images. How about a compromise - if we allow the images, they can be removed - no questions asked - if they aren't updated at least weekly (if there are polls that week of course). I believe that would address your concerns about any editor giving up on the graph. Quinby  ( talk ) 17:49, 7 January 2023 (UTC)
 * I updated the poll results regularly, two or three days later at the latest. If the image is not updated, I can restore the graph personally as well as other users. I agree with this compromise solution. Governor Sheng (talk) 19:11, 7 January 2023 (UTC)
 * Hello to everyone. As a user who mostly updates this graph and pools I can tell you that this graph is a real pain because of bouncing and most surely I would like to see graph as presented by @CoaxAndBotany. Anyway Wikipedia at this moment doesn't provide this. Talking about updating module is not new story and as I can see there is try for upgrading since 2019, but nobody can be sure when it will be finished. At this moment I would support creating images like it was done, with update rate of at least once per month as changes are not so drastic. Also threshold of 5 percent should be removed as it is viewed on the level of each electoral district and not a county in general.--Opatijac97 (talk) 17:44, 7 January 2023 (UTC)
 * I will remove the threshold, thank you for the clarification! CoaxAndBotany (talk) 18:55, 7 January 2023 (UTC)
 * Just a note for you, the polls are usually taken at the beginning and the end of the month, and there are 3 polls monthly on average. Governor Sheng (talk) 19:13, 7 January 2023 (UTC)
 * Just passing by as I saw this to state my opinion on this (since I'm also the author of the charts used in some of the articles that have been mentioned as an example):
 * 1) I agree that image charts are far superior to the inbuilt wikipedia graphs. At least until these can be fixed/modified to a state in which they can actually produce some useful info, I don't think they currently allow for a proper depiction of trendlines. I would rather have no chart at all than having these inbuilt graphs, frankly; we should keep in mind that we are not required to produce graph trendlines for every single set of opinion polls across Wikipedia. If image charts can be produced and updated regularly, then by all means go with these.
 * 2) That said, some of the editors pressing for the change happen to be the same ones (coincidentally or not) that are pressing for similar changes across a variety of articles. A couple days ago, I just engaged in a very similar discussion with a very similar structure at another talk page, in which the issue was not on whether to use graphs or images, but on having charts in articles lacking them (though this also involved a case in which an existing chart had been replaced with a new one based on a "lack of updating", but where such updating had not even been requested before the replacement was conducted). I'm swayed towards @'s view that maybe this is not being approached with enough good faith from some editors that, seemingly, simply want to use the images they created rather than collaborating in finding a compromise, thus creating an edit warring mess across a variety of articles that is entirely unwarranted. On this issue, I'd please ask for the people involved to have a more sensitive approach and to discuss rather than to impose. I don't think there's any profound disagreement on this issue that can't be solved in a civil and polite manner. Cheers.  Impru 20 talk 15:19, 8 January 2023 (UTC)

Few thoughts on the proposal to use images and the particular usage of the image on Opinion polling for the next Croatian parliamentary election article:

1) Making all the contents of the article viewable on mobile browsers/small screens, which is stated as a rationale for using static images, sounds like a plus. However, this seems like the only positive side of the proposal.

2) As others have stated, having a static image means that this part of the content effectively becomes "locked" for the most of the editors and we have to rely on few editors (or one) who promised to update the image regularly, which is less than ideal. Sure, there is always an option to revert to the graph if the updates stop, but it is easier to update the graph incrementally and implicit uncertainty regarding the threshold of when the updates stopped may make editors more reluctant to revert to the graph if needed (and that reluctance adds up to the one from having to build entire missing portion of the graph instead of just copying the whole thing minus one data point, which is easier to add).

3) This may be a subjective assessment, but the graph has a "cleaner" look, probably because the white background, as opposed to light gray background used in the image and also because of the better overall contrast. If the image could be made to look more like the graph (regarding design choices), that would make it much better.

4) What is the rationale for the choice of two additional minor parties (SD and Fokus) that were included in the image, but were not present in the graph? It is not clear that they deserve inclusion while other minor parties (like e.g. RF or Centar) don't, while for the top five parties it is clear that they each of them consistently receives more support than any other party.

5) As others have pointed out, the way using images is being pushed can lead one to suspect that it may not be in a good faith. It is said that it is "the standard is to use image graphs if available", but this smells like WP:OTHERSTUFF, and besides, there is no policy stating image graphs should be used. Odd choice of additional parties also contributes to the impression that the true motivation for pushing it may be hidden and different from the stated motivation. Niokog (talk) 09:53, 9 January 2023 (UTC)

Background outdated
Isn't there a problem with the Background section? In "Changes of ministers in the government", nearly all of it is about change of ministers which happened before the previous election in 2020. It doesn't belong on this page, but the 2020 one. Aréat (talk) 19:57, 11 April 2024 (UTC)


 * Hi! I wouldn't call it a problem, but yes. The point the section is trying to make is about government changes during Plenković's premiership, during both his mandates, there were a total of 30 ministers leaving. A majority of the ministers listed are from the former cabinet, but many are from the current one (Horvat, Aladrović, Marić, Tramišak, Paladina, Banožić). It is mostly written in a non-chronological way for the current cabinet, bcs predeccessors and successors on specific posts are connected: "The first minister (1st cabinet) left because of this and this, but then his replacement, the new minister, later (I think this is when it's missing dates/years) also left because of that and that." I think it should be neated up a bit, adding more years, just making it clearer, but overall I think it's fine. I do think these things should be added to the 2020 election article (all govt changes at that point, by 2020 of course). I plan on dealing with HDZ controversies, some corruption scandals and possibly adding on the Plenković Cabinets articles this weekend anyway, so I could try to improve this. CroatiaElects (talk) 20:45, 11 April 2024 (UTC)

Results breakdown
Here's the breakdown by electoral district if anyone wants to double check.

One issue is that some minor parties ran on different lists in different districts (Reformists being one), while some were split in some districts (for example, IDS–PGS ran separately to the Social Democrats–Labourists in constituency VII)... Cheers, Number   5  7  12:56, 23 April 2024 (UTC)


 * Hi! Does someone know what's the discrepancy between the DIP turnout (62.30%) and the counted turnout on here (61.89)? just DIP not updating smth or? ~14k votes are missing (according to the national results here) for it to be 62.30%
 * tried to see if i'm missing smth, but idk
 * thanks CroatiaElects (talk) 12:48, 25 April 2024 (UTC)
 * They might be including the national minority constituencies – if you do that, it increases the total votes cast to 2,215,210 and the number of registered voters to 3,558,089, giving a turnout of 62.26% (although that is still lower than 62.30%).
 * On a related note (and ccing ), the "registered voters" figures for the minority and overseas constituencies seems to actually be the number of voters who turned up at the polling stations (votes cast is only slightly lower, probably due to people taking a ballot paper and not voting). Are these the actual registered voter figures? It seems a bit unlikely that pretty much 100% of registered voters voted in each of these constituencies. Number   5  7  13:01, 25 April 2024 (UTC)
 * Ah, probably smth like that then.
 * Regarding the question, I don't know much about the Diaspora district tbh, but I'm pretty sure for the 12th Minority district they count only the voters that voted in the 12th district as registered voters, yes. It's kinda dumb, but it has a reason. Voters who are recognized as N. Minority members with the right to vote in the 12th district choose whether to vote in their regular territorial district or in the 12th. And it seems like they count only those who opt out to vote in the 12th as registered voters, instead of putting the number of all voters who have the right to vote in it, but some do and some don't. I think they do this bcs if they counted all of those NM voters instead of just 12th district voters then the total number of registered voters in the country would be bigger (when you count all 12 districts). A voter can't be registered twice if we're looking at the number of all registered voters in Croatia. If that's the reason, it makes sense, but it also creates a false perception (or rather blurs it) of how many NM voters actually choose to vote for NM candidates (a very low number actually do, if I'm correct, see the total population of Italians in CRO for example and then see the number of voters for Italian constituency in the 12th district --> this is often criticized as 900 votes can elect an MP but parties like Focus get 50k votes for the same, only 1 MP).
 * Maybe it's similar in Diaspora and you have to actually vote to be on the "registered" list (but it's dumb if they even say "registered voters" when that's not what it actually is, they should just omit that) and you have to sign in to vote & aren't automatically on the list or smth, I'm blindly guessing. But it wouldn't make sense, the Diaspora voters only can vote in their 11th district, so what's the point then? Maybe it's because of this: If the Diaspora is counted in the national data then it's to not overexagerate the total "Registered voters in Croatia" number as it is already 3.5 million (like Croatia basically has the population) and it sounds kinda ridiculous. Bcs I'm sure the DIP and State Statistics Bureau have the data of all those emmigrants, Croatian citizens in those countries with the right to vote (& it's a huge number) CroatiaElects (talk) 14:07, 25 April 2024 (UTC)
 * I hope I'm on the right track. I could check official stuff, the Election law or documents, to find the explanation but I suppose it doesn't matter... Maybe it's common knowledge but I'm not sure. Anyway, thank you! CroatiaElects (talk) 14:11, 25 April 2024 (UTC)
 * Good points. I was also thinking these voters were those who had actively registered, and so were considerably more politicaly active. The minority results had been added on the french wiki's 2020 election page with similar turnout, so it seemed usual to me. "Turnout" may be removed, then. If it stay stay, we should definitely add a sentence explaining how it work.--Aréat (talk) 14:39, 25 April 2024 (UTC)
 * Good points. I was also thinking these voters were those who had actively registered, and so were considerably more politicaly active. The minority results had been added on the french wiki's 2020 election page with similar turnout, so it seemed usual to me. "Turnout" may be removed, then. If it stay stay, we should definitely add a sentence explaining how it work.--Aréat (talk) 14:39, 25 April 2024 (UTC)