Talk:Opinion polling for the 2022 Italian general election/Archive 2

Government support
Considering the fall of the coalition, I propose to have a new table starting from polls conducted after it, in which we would go back to the practice pf showing the "lead" of the first party over thz second. What do you think?--Aréat (talk) 12:31, 10 August 2019 (UTC)
 * The government has not formally fallen yet. And we already show the lead of the first party in all tables. --Ritchie92 (talk) 12:37, 10 August 2019 (UTC)
 * I'm not particularly convinced of these "Government / Opposition" columns. While these are used for some countries (mainly from Northern Europe), there are not used in a majority of opinion polling articles, and could be misleading, i.e. M5S and the Lega are not operating as an electoral alliance, and politically their alliance has been very fragile (as events throughout the last months, particularly this week, show), thus limiting the scope and usefulness of any information these columns could provide. Do sources report on these aggregations? The columns could possibly be scrapped altogether, thus not requiring creating a new table (which would be the third table for 2019. I don't think this is sustainable). Impru 20  talk 12:42, 10 August 2019 (UTC)
 * Yes, some sources report(ed) the sum of the M5S and Lega as government backing vote. I think it was a useful data and it was a well-defined operations to be made: (vote of Lega) + (vote of M5S). So I wouldn't delete it. The operation was straightforward and simple and it would not constitute SYNTH or OR, opposed to what is currently done for the coalition table for example, where the arithmetics that is made for each polling company is not consistent among the various polls. --Ritchie92 (talk) 13:09, 10 August 2019 (UTC)
 * I don't like these kinds of operations because while the mathematical calculation in itself would be allowed per CALC, the choosing of which parties to sum, as well as the purpose of such aggregation itself, could be regarded as the SYNTH elements. In this case it may be relatively straighforward as it accounts for the parties in government, but then it means comparing a non-electoral alliance with actual vote opinion polling figures. Further, if the purpose is to present it as some sort of popular support for the current government, it could also be also interpreted as SYNTH because this is not a specific question on the government's approval, but rather an aggregation of the popular support for each of the individual parties (whose popularity is not directly comparable to an actual approval rating, specially when you aggregate it, i.e. the Lega is currently fairly popular while the M5S isn't, but then it somehow translates into a wide electoral support, which is due mostly to the Lega's strength). Considering these issues, I don't think that creating a new table is worth the effort just in order to preserve these columns, unless sources did widely report such an aggregation. But just my 2 cents. Impru 20  talk 14:39, 10 August 2019 (UTC)
 * I think that the choice of which parties to sum is quite straightforward, and poses no issues at all. The column is not titled "Government support", it's titled "Government parties", so it's clear that is the bare sum of the pollings of the two parties. There is also a footnote explaining that. I would not remove the column because it shows that the parties forming the government together have had the support of the larger part of the voters, and that basically the sum of the two did not change much, while the prevalence of one party on another changed drastically. I would actually keep the format of "Government parties" for any future table, meaning the "aggregate votes of the parties supporting the current government". What I would cancel is the "Others" column, which is just given by (100 - govt) and adds no information. --Ritchie92 (talk) 16:25, 10 August 2019 (UTC)
 * I would not remove the column because it shows that the parties forming the government together have had the support of the larger part of the voters, and that basically the sum of the two did not change much, while the prevalence of one party on another changed drastically. Yes, this is exactly the issue. The sum of both parties is straightforward, but the actual need for such a sum, and the purpose behind it, is what becomes quite obscure and SYNTHy. Why this choice of parties and not any other? Why not some columns showing parties aggregated according to their position in the ideological spectrum? Or why not showing the actual opposition parties as well, since the same purpose of "showing the drastic changes of prevalences of some parties over others" could apply to them as well, i.e. PD and FI also showing some drastic changes since the election? (Note that I'm not suggesting any one of these should be added, just bringing it as an example). Then, you would have the issue of what to do if a Monti-like cabinet is formed at some point; would you create a column full of 0.0 numbers for the government? Or would you include parties giving external support to the government even if they do not form part of it? (This later choice could raise a lot more issues, since external support can become very variable and unstable throughout time). I think this is a needless mixture of both the legislative (which is what election opinion polls do measure) and the executive (not directly measurable by these polls) that gives little information overall and instead creates more issues than it solves. Impru 20  talk 18:08, 10 August 2019 (UTC)
 * Why this choice of parties and not any other? Well, because other parties do not take part in the government, I think it's easy to see. Of course under a technocratic government the column has no meaning, but since there is a political government, why not? I don't see the problem in having multiple tables according to the different political situations. We did that when some new parties came in "play", so it's arbitrary anyway. --Ritchie92 (talk) 19:04, 10 August 2019 (UTC)
 * And why should "the government" be given any more relevance than any other combination of parties when it comes to opinion polling? "The government" is not running for election, parties do (and the current two parties in government have it since there is a political government, why not? Why yes? For this part, consider that data should not be included just because of it indiscriminately. This also does not answer the issue on what to do with parties supporting the government externally, unless this column has been planned ad hoc only for the M5S-Lega government, in which case the more useless it becomes.
 * We did that when some new parties came in "play", so it's arbitrary anyway. I don't think this is the same. New parties are added when they get reported in opinion polls, so creating new tables may be a consequence of that. Here we are discussing creating a separate table because it is argued that a user-made column consisting of automatically-agreggated figures for two parties—without it being clear whether the sources do back these sums—must be preserved somehow.
 * I think sources become very important in this context. How many opinion polls have been showing government figures aggregated and for how much time? It it has been a trend during the government's tenure then it may have a justification, if it has been done only by a few polls inconsistently then I don't think preserving the column should be prioritary over splitting the table, and if I had to choose, I would pick removing the column rather than splitting the table any further at this time. Impru 20  talk 19:41, 10 August 2019 (UTC)
 * Noto and EMG are the ones that for sure were listing the "total government" in their tables, before the European election polling started. Anyway, I just thought it was interesting data to have available and that we can be flexible and decide to show or not show data whenever we deem it interesting. However I don't have a strong preference on this, if you want you can wait for some other editor's opinion and then remove the column. I would rather spend more time discussing about the mess that is the coalition vote section, where everybody is adding polls indiscriminately by calculating stuff or by citing the source directly. --Ritchie92 (talk) 19:50, 10 August 2019 (UTC)
 * Seems fine. I don't understand that bit about the coalition section either; 2018 coalitions won't necessarily be replicated for the next election (in fact, they are usually never 100% the same ones), so assuming they will in order to justify automatically agreggating numbers for all opinion polls in such a way is SYNTH. That section should be limited to numbers actually given by opinion polls themselves, or be scrapped altogether. Impru 20  talk 21:34, 10 August 2019 (UTC)


 * Yes, but as you can read from the last entries here the matter is not clearly settled yet. At the moment that table is full of self-calculated coalitions. --Ritchie92 (talk) 21:40, 10 August 2019 (UTC)


 * I would say we should abolish the government opposition columns for polls since the cancellation of the coalition (so Aug. 9) Braganza (talk) 19:43, 11 August 2019 (UTC)
 * This is somewhat funny, since the government columns are also included since the 2018 election (4 March), despite the government coalition not being formed until 1 June 2018. That's three months full of columns on a government which was not yet formed, so if the criteria to follow is to remove the aggregated figures since the government crisis, then they should be removed as well for the period of time between March and June 2018... a further reason justifying the scrapping of these two columns altogether. Impru 20  talk 19:49, 11 August 2019 (UTC)

The government coalition has not been "cancelled" on 9 August, at least not officially. --Ritchie92 (talk) 20:03, 11 August 2019 (UTC) At this point I would say yeah, let's remove the columns altogether and forget about this. --Ritchie92 (talk) 20:03, 11 August 2019 (UTC)
 * I know that the government still exists with all their ministers and they have not been overthrown yet Actually, I am also in favor of the complete abolition since it also mixes the swing voters between M5S and Pd with those of L and FI Braganza (talk) 06:57, 12 August 2019 (UTC)
 * Which switch? I don't get it. --Ritchie92 (talk) 08:21, 12 August 2019 (UTC)
 * I mean Swing voters... Braganza (talk) 08:55, 12 August 2019 (UTC)
 * But tracking the swing voters is not the purpose of that column. It was just the bare sum of M5S + Lega, which could turn out to be handy. --Ritchie92 (talk) 09:38, 12 August 2019 (UTC)

But now MUST abolish the columns Braganza (talk)
 * In my view we can remove all the columns regarding government support. -- Nick.mon (talk) 16:13, 23 August 2019 (UTC)
 * As explained in previous comments, I support the removal of the columns as well. Impru 20  talk 16:28, 23 August 2019 (UTC)
 * Let's keep them, insead. And start a new tabe whenever the new government is formed. Why not? --Checco (talk) 14:25, 25 August 2019 (UTC)
 * Because we don't do things in Wikipedia just because for it or because we like them. Impru 20  talk 14:27, 25 August 2019 (UTC)
 * This is worth. --Checco (talk) 14:36, 25 August 2019 (UTC)

How about we just do not have a Govt parties column column between 9 August and the foundation of a potential PD-M5S government and the time before 1 June 2018?

Braganza (talk) 15:51, 25 August 2019 (UTC)


 * This would still be hampered by the common issues already pointed out above, i.e. that "the government" (be it M5S/Lega or M5S/PD) is not an electoral alliance, and that any hint that its purpose is to give readers clues about how they should interpret polling trends and swings would be SYNTH-y. Further, this proposal would also add a very POV-ish element by randomly determining that 9 August is the separation date: as far as we know, the government is still standing and is still formed by M5S and Lega ministers, and it'll still be so until replaced by a new one. Nonetheless, Checco has not given any reasoning for us to preserve the columns other than "it is worth" or other ILIKEIT arguments, so it's actually not worth preservation. Impru 20  talk 16:00, 25 August 2019 (UTC)
 * Then we just delete the columns if there is no solution Braganza (talk) 16:33, 25 August 2019 (UTC)
 * The solution is keeping the established version. Until a new government is formed, the M5S and the LN are the government, thus I would not have three tables, but just two. However, User:Braganza's proposal could be a compromise. First of all, however, we have to decide if we want to have the "government" column or not: please be all clear on that! As of now, there is no new consensus supplanting the old one. Only User:Impru20 is straighforwardly against it. User:Aréat, User:Ritchie92 and User:Braganza would keep it, if I understood well. Finally, I was a little bit surprised by User:Nick.mon: when he will read the follow-up of the discussion, I am sure he will have second thoughts. --Checco (talk) 19:14, 25 August 2019 (UTC)
 * when he will read the follow-up of the discussion, I am sure he will have second thoughts. Most people participating in this discussion have shown support for removal in one form or another; however, I won't be trying to manipulate their statements as you are attempting to do here. If your only argument for removal is that "this is the established version" (established by whom? An edit warrior?) and that all people supporting removal but myself surprise you but that they "will have second thoughts" about it, then there's no point in maintaining this just for one user's pleasure. I should also remind that these columns were unilaterally added by one IP edit warrior who kept adding and re-adding them despite being reverted multiple times by several users (frankly, seeing the page's history I don't understand why the user wasn't brought to ANEW right-away; instead, the columns were left over there out of pure boredom with the edit warring). Impru 20  talk 19:19, 25 August 2019 (UTC)
 * I do not recall anything about the IP user, but he/she did a good thing. I do not see consensus for removal. I am not even sure that User:Nick.mon is in favour of removing the column altogether: I would ask him to specify if he is for complete removal or only after the government crisis started. The other three users who participated in the discussion so far, plus me, seem to like having the column in one way or the other.
 * Everyone should make clear: 1) column yes or no; 2) if yes, what to do after the government collapsed.
 * That is me: 1) I would like to keep the column; 2) I would like to have a new table when the new government is formed.
 * Cheers, --Checco (talk) 19:30, 25 August 2019 (UTC)
 * You have a discussion in this very same talk page where that IP user was told to stop edit warring and reach a consensus here for inclusion (something they didn't). There wasn't any consensus to add the columns in the first place.
 * I am not even sure that User:Nick.mon is in favour of removing the column altogether And I'm not sure why you keep patronizing in such a way. I think everyone in here is mature enough to think of its own, as well as perfectly read what everyone has said here, without you trying to manipulate them your way because their answers did not please you. So far only you are staunchly in favour of keeping the columns. You are not even attempting to give any solution or argument to the above raised issues on the columns.  Impru 20  talk 19:48, 25 August 2019 (UTC)
 * I do not remember, sorry about that. There was not consensus then, but now it seems there is no consensus for removal. I do not understand why you need to be always so brutal. What I am asking is just open debate. People, please have a say. --Checco (talk) 19:56, 25 August 2019 (UTC)

How about we just do not have a Govt parties column column between 9 August and the foundation of a potential PD-M5S government and the time before 1 June 2018? Braganza (talk) 19:46, 25 August 2019 (UTC)
 * I think is being unresponsive enough to the above raised problems that there's not even any need to try to propose any compromise just to "please" him. Wikipedia is not owned by anyone, and frankly I'd be worried if the only purpose of the columns if for Checco's amusement (specially considering there wasn't consensus for the columns' inclusion in the first place, as they were the result of an endless edit warring with an IP user which, btw, has not come around here ever since). I'd like for him to actually try to answer to these, rather than trying to patronize people his way.  Impru 20  talk 19:52, 25 August 2019 (UTC)
 * @User:Braganza: The Conte cabinet is still in place, acting. However, I would accept your proposal as a compromise. --Checco (talk) 19:56, 25 August 2019 (UTC)
 * I'd like for you to stop commenting on contributors and, actually, start commenting on content. I'm sorry if I do seem "brutal" to you; surely other people may have some other thoughts on you, but I think it is good that these are maintained off the table and we focus on actual content. Indeed. is doing a laudable job in trying to reach a compromise so that you may be pleased, but the issue here is that there is no reason for us to please you. Ironically, you argue there is not consensus for removal but this is what turned you out in here in the first place. Seems like you'll only accept what will please you.  Impru 20  talk 20:06, 25 August 2019 (UTC)
 * I never said that there is no consensus for removing the column. Braganza (talk) 21:01, 25 August 2019 (UTC)

Dear friends, excuse me but I'm on holiday, so I don't have so much time to actively partecipate in this discussion, but I'm glad that my opinions are so requested, considering that you've mentioned me so many times ;) Anyway, I always disliked "government's columns", not only because they're not so used in other similar articles, but also because in Italy, governments change so frequently (as you can see in these days...) so it would be a bit confused. -- Nick.mon (talk) 21:06, 25 August 2019 (UTC)
 * I was addressing Checco's earlier reply there, sorry if that was misunderstood xD
 * I was thinking about suggesting an RfC on the issue if required in order to clarify everyone's stances, but then I saw your comment where you make your stance very clear. Nonetheless, an RfC could help if there are still doubts on this and if further input is required (specially considering the chaotic origin of the column issue altogether), because that would let everyone's positions (and arguments) very clear to everyone. That would depend on whether a consensus is considered to exist or not as of currently, I think, since that would be a time-consuming proceeding. Impru 20  talk 21:18, 25 August 2019 (UTC)

I would like to stress that I do not have a strong opinion on this. I had second thoughts after the discussion with Impru20, and now I actually see no problem in removing the column if you want. We should in any case remove the "Others" column, which is redundant. In case we keep the column, we should keep it empty in the transition periods (during which a government resigned) so 4 March to 1 June 2018, and 20 August 2019 to present. --Ritchie92 (talk) 08:04, 26 August 2019 (UTC) In addition, the Senate has at least rejected the motion for the motion of no confidence Braganza (talk) 14:01, 26 August 2019 (UTC)
 * Well as I said if then from the 9th of August Braganza (talk) 13:39, 26 August 2019 (UTC)
 * Conte resigned on the 20th of August. Nobody resigned on the 9th, it was just the motion of no-confidence put in the calendar by the League (and eventually the motion never happened). I think the actual day of resignation is a better separator. --Ritchie92 (talk) 13:43, 26 August 2019 (UTC)
 * But the real purpose of the Government Opposition columns is to gauge the approval of the governing parties, which, according to an announced vote of no confidence, can no longer be the case.
 * The Senate has not rejected any motion: there was no vote at all. Anyway, the announcement of a vote of no-confidence is not enough to say that there is no approval (as history taught us, since the League was still approving the government: its Ministers did not resign). And even if it was, we cannot establish that as a general rule because we do not know what will happen with future governments when they will eventually fall. I would say the resignation date is a more definite date. --Ritchie92 (talk) 14:22, 26 August 2019 (UTC)

I think there is a lot of confusion on this thread. I support the launch of a RfC to make things clear, if you still also think so. It would be great if in the RfC you put a very definite question, and list all the options available. Thank you. --Ritchie92 (talk) 14:24, 26 August 2019 (UTC)
 * Ok, so considering all the above, should the question be just about whether to preserve the columns or not and why? How many options should be offered? I'm seeing a big consensus to, at the very least, remove the "Other" column, so I don't know if this should be presented as a separate option or just be merged within the whole pack altogether. Impru 20  talk 18:46, 26 August 2019 (UTC)
 * I think it's a double question, and editors should give a reply to both: (Q1) Do we remove the "Others" column? Options: yes or no. (Q2) What do we do with the "Government" column? Options: (A) remove it altogether; (B) limit the entries to the periods during which a non-resigning government is in power (e.g. Conte cabinet only from 1 June 2018 to 20 August 2019); (C) keep it like it is, and change parties (and colour) when a new government is sworn in (e.g. Conte cabinet from 1 June 2018 until a new government is formed, possibly in the next days or weeks). Note that there is no rational reasonable option by which we can keep the Conte cabinet entries before its formation on 1 June 2018. I think these should exhaust all possible combinations. Obviously just counting the !votes is not the way WP works, the closer of the RfC will then decide about the outcome. --Ritchie92 (talk) 21:29, 26 August 2019 (UTC)
 * Hmm I agree with the essence of it but not with the specifically proposed formats for several reasons:
 * 1) in Q2 you'd be essentially greatly favouring two alternatives over any other (B and C), as well as potentially raising a conflict with sources in the question itself, i.e. giving undue prominence to 20 August as a "resigning date", despite such a date having been discarded elsewhere in Wikipedia, specially at Giuseppe Conte and Conte Cabinet (in the later of which, 20 August is merely mentioned as "the announced date of resignation", not even the date in which such a resignation was effective). Indeed, the issue about 20 August as the date would probably require a different discussion and a separate consensus to be reached globally regarding all concerned WP articles, to avoid potential inconsistencies, because if consensus globally was against using 20 August as the resignation date then it would be meaningless to use it in this article. Since RfC questions should remain brief and neutral and this would pose the issue of bringing the RfC off-topic, I'd say to leave it out.
 * 2) Q1 is problematic in that it would seem as if this was an actual dispute in the discussion, when it isn't: comments so far have shown either support for the preservation of both columns (or just the "government" one), or its removal altogether, with no one being vocal about preserving the "Other" column on its own. Indeed, preserving the "Other" column would be awkward should the "Government" column be removed, so such a possibility, even if very remotely possible, should not be allowed.
 * I'd put it in something along these lines: (Q1) What should we do with the "Government" column? Options would be (A) Keep, (B) Remove/Delete; and (Q2) In case it was preserved, what do we do with it? This one would be an open question (accepting even a "no answer") so that everyone may give out their opinions about the preferred formating of the column(s) in case the RfC resulted in a consensus for "Keep". Obviously, it is heavily implied that supporting "Delete" would automatically imply supporting the removal of the "Other" column attached to the "Government" column, as per the reasoning above. However, you'd still be able to vote "Delete" in Q1, then give out your preferred choice in Q2 in case there is any eventual consensus for "Keep". This would give us the following possible outcomes:
 * a) Q1 results in a consensus for "Removal" = The columns would be removed altogether.
 * b) Q1 results in a consensus for "Keep" and Q2 results in a consensus for status quo = Nothing changes.
 * c) Q1 results in a consensus for "Keep" and Q2 results in a choice other than status quo = The columns are preserved under the new choice consensuated at Q2.
 * d) Q1 results in a consensus for "Keep" but Q2 results in no consensus = Then this may depend on any hints given by the RfC closer. It could result in the status quo version remaining in place due to a lack of consensus or, if users are willing to keep discussing, to discuss which one of the proposed choices is the best preferred one (a drawback of this is that this could eternize the lack of consensus to the point that the status quo version remains in place by default, but at the very least the "Keep/Delete" issue would have been solved at this point).
 * I think this would encompass all possible situations while addressing the central issue at hand, while leaving the question wording neutral as well as allowing for further, more specifically-centered discussion in the future in the event of outcome d). Impru 20  talk 18:25, 27 August 2019 (UTC)
 * Ok looks good to me. --Ritchie92 (talk) 07:02, 28 August 2019 (UTC)

I much welcome what you guys are doing in order to start a RfC. I would give also the option of having the colour of the "government parties" column rightly after the formation of the new government and not having blank rows. The current government will be in office, in acting function, until the new government is installed. I like open debates and always accept the will of the majority, provided that there is a clear consensus showing it. It is not true that I only accept what pleases me. I always accept what the majority wants. Of course, I discuss and bring forward my opinions, without being rude as others, which should be a good thing, I guess. --Checco (talk) 06:31, 28 August 2019 (UTC)
 * I think your additional proposal would go in the "free" reply to (Q2). There are no options on that. --Ritchie92 (talk) 07:02, 28 August 2019 (UTC)
 * Yes, that would be within Q2's scope. Impru 20  talk 18:16, 28 August 2019 (UTC)


 * Ok, I've opened an RfC on the issue below. Out of courtesy, I'm pinging all involved editors in this discussion for providing input if wished: @ @ @ @ @ Impru 20  talk 19:02, 30 August 2019 (UTC)
 * Thank you. I feel like this discussion and the arguments involved has got me out of my depth, though, so I will remain neutral and not vote on the mentioned Rfc. Cordially.--Aréat (talk) 19:55, 30 August 2019 (UTC)

Archive bot
Hi. I think this page has become very long, and there are some closed discussions which are not really needed anymore here. Should we introduce an automated archive bot? --Ritchie92 (talk) 07:04, 28 August 2019 (UTC)
 * I support the idea. Impru 20  talk 17:39, 30 August 2019 (UTC)
 * I think nobody is strongly against this, and after more than a week I introduced an archive bot on the page. --Ritchie92 (talk) 13:32, 5 September 2019 (UTC)

Conte II govt parties
The government in charge is supported by PD, M5S, SI and A1 (addictionally, the sole EV MP voted in favour of the motion of confidence). A1 and SI are seldom polled together (as LeU); otherwise A1 is polled alone by just one pollster, and SI is included in the list LS together with the PRC - which is not a govt party. I honestly don't know how to deal with the problem. Any ideas? TommasoM (talk) 09:49, 10 September 2019 (UTC)
 * I don't know if the government column will be deleted but I think that we should make a note (like this) every time a survey polls A1 and SI separated and when LeU/Sin is the sum of SI+PRC, especially for the latter in the pre-european election opinion polling, when the column of the leftist party was called LeU. And probably we should note when parties like PAP or Cambiamo are polled in the Other column, unless the column will be dropped altogheter --Broncoviz (talk) 19:51, 10 September 2019 (UTC)
 * I agree we should note when minor parties (PaP, NcI, A1, Cambiamo and so forth) are polled in the Other column. But I don't know what to do with the govt parties column, since most pollsters tend to poll SI and PRC together. TommasoM (talk) 19:58, 10 September 2019 (UTC)
 * I think the only way to solve the problem is to get rid altogether of the government column, I posted above to say this. --Broncoviz (talk) 14:30, 11 September 2019 (UTC)
 * I just noticed that there is the same problem with More Europe: some pollsters (Ixè, Tecnè, SWG, IZI) polls them with only "More Europe", while other ones (Euromedia, Bidimedia) polled it along with Italy in Common, like in the last European Elections. There should be at least a note about this --Broncoviz (talk) 14:30, 11 September 2019 (UTC)
 * I don't oppose the proposal to get rid of the govt column. However, I don't think the More Europe - Italy in Common issue is a serious problem, at least as long as no pollster polls IC alone or with other parties (i.e. the Greens). TommasoM (talk) 19:25, 11 September 2019 (UTC)

Inclusion of events
I think a change of coalition is important enough to include as event, so the reader can more easily interpret which parties form the governement at the time the polling is done. I agree there should not be an excess of events, but this one is too important not to include.

There are some other pages which include such events (although not the majority): For most countries such an alternative coalition does not happen while there is no new election, so for most countries a mention of such an event is not needed. But for the one country I do know, there was an alternative coalition while there was not yet a new election (Belgium), it is included. --PJ Geest (talk) 10:57, 6 September 2019 (UTC)
 * Opinion polling for the 2019 Belgian federal election
 * Opinion polling for the next United Kingdom general election
 * Opinion polling for the next Scottish Parliament election
 * Opinion polling for the 2019 Canadian federal election
 * Opinion polling for the 2021 Russian legislative election
 * There was already a previous discussion on this. Firstly, other stuff may indeed exist, but that does not mean that it is "good stuff". Specifically, consensus for event inclusion was only achieved for UK election articles (and I include here both UK, Scotland and other UK-related articles, since the same criteria are applied to all of these) and Canada, and under different conditions, circumstances and application. Events for other articles are being added with absolutely no control or consistency, but rather at each editor's leisure.
 * The main issue with events are the WP:NPOV and WP:OR implications they have; i.e. some people like mixing up events with opinion polls because "they may affect opinion polling" (which is an absolutely WP:CRYSTALBALL assertion). For sure, events may affect opinion polling but it is nigh to impossible to specifically determine which event does affect an electoral trend, which parties are affected by it, what impact magnitude does it have in opinion polling and for how long does such impact last. An event may only affect some parties but not others; it may have a very strong but very short-lived impact; or whatever. Then, a joint chain of events such as a government crisis may have an impact in opinion but only because of a specific minister's resignation and not other related events. The implications of this is that either we add all and every possible event that may have an impact in opinion polling (which may lead to us having to add every event possible into the table in order to avoid POVs, which will in effect disrupt the whole table) or list none at all, at least in this article. This is an opinion polling article, not a list of events throughout Italy's XVIII Legislature. If people wish to have a comprehensive timeline of events throughout such a time period, then maybe a separate article should be created for it, without the unnecessary mix up with opinion polls. Impru 20  talk 11:21, 6 September 2019 (UTC)
 * Perfectly agree with User:Impru20. Just an addition: there is the possibility of a middle-ground solution: only list events when these events are the formation of a new government. That would make some sense, however I would still not support it because I think this is still just a list of opinion polls. --Ritchie92 (talk) 11:26, 6 September 2019 (UTC)
 * I have nothing againts the introduction of events, not just the change of government, but also others. --Checco (talk) 12:28, 6 September 2019 (UTC)
 * I support the proposal to only list events when the event is the formation of a new governement, because if you strictly include these events, according to me, you don't have the problems with WP:NPOV or WP:OR. --PJ Geest (talk) 14:45, 6 September 2019 (UTC)
 * You'd be hinting that these events have an impact in opinion polling. Otherwise, why'd you add them? so the reader can more easily interpret which parties form the governement at the time the polling is done is already a reason trying to direct the readers' attention at some direction, and thus POVish by definition. I've always been skeptical of people trying to add information where it does not belong "so that readers may understand something". You already have the Next Italian general election article to explain the changes in government; or you can create a separate timeline of events (including government formation, of course). Impru 20  talk 14:51, 6 September 2019 (UTC)
 * A change of government like this would be a good reason for a new table, let alone an event row. Onetwothreeip (talk) 10:17, 7 September 2019 (UTC)
 * Why would we have a new table? The opinion polls did not change the parties that they measure in the meantime, they are in perfect continuity with the ones before the new government. --Ritchie92 (talk) 10:33, 7 September 2019 (UTC)
 * As the former polls are measuring the parties during a previous government to the current one. As for continuity, there's certainly not continuity in the columns that sum the government and opposition parties. Onetwothreeip (talk) 10:36, 7 September 2019 (UTC)
 * This is more or less what I defined as I've always been skeptical of people trying to add information where it does not belong "so that readers may understand something". Sometimes, it goes to the point of having to be "very creative" (to label it in some way) just to try to add something that really does not belong here.
 * On the issue about the "government columns", there is already a separate RfC above on that. Impru 20  talk 10:58, 7 September 2019 (UTC)
 * In my view, on events, Opinion polling for the next United Kingdom general election should be the standard. --Checco (talk) 13:02, 8 September 2019 (UTC)
 * Firstly, this is not a Westminster system, so this is your first main flaw with using UK as "the standard" (specially without justifying why). Impru 20  talk 13:54, 10 September 2019 (UTC)
 * No flaw! It is something self-evident, I think. I do not see why elections of new leaders, changes in government, other important elections and so on should be a factor in differentiating Italy and the UK. I think we should take example from that article, as well as those on Isreal, Canada and many others. --Checco (talk) 14:40, 10 September 2019 (UTC)
 * Sincerely? Because the UK's established consensus is a flawed one. The current system there is based on a fragile compromise between those advocating for no events at all, and those wishing to add as many events as possible because "hey, this could, maybe, you know, possibly affect opinion polls yeah?", because people kept randomly adding events at a constant basis for years and years. The consensus is so tenuous that it is frequent for people to keep adding random events at leisure, then those having to be contested and disputed at the talk page. Consensus at Israel, Canada, UK and others (not "many", since it is actually a majority of opinion polling articles not using events at all) obey to specific reasons and to solve specific issues shown in those articles, not because it is good in general to have events in opinion polling tables. This is why WP:OTHERSTUFF exists. Israel opinion polls only show seat projections and I am not seeing you advocating for replicating that here. UK opinion polling articles, until very recently, only showed results for the Cons, Lab and Lib Dems, and I am not seeing you seeking to replicate that here. Canada opinion polling articles use very strict criteria for opinion polls in general (such as to be considered for inclusion), and I am not seeing you seeking to replicate that here. That something is in one way in some place does not mean it is a good idea to replicate it elsewhere, specially when, as pointed out, a majority of opinion polling articles throughout Wikipedia work differently. Impru 20  talk 15:20, 10 September 2019 (UTC)
 * The current basis for events in the UK articles is because they demonstrate changes in support, whether that is in the parliament or in by-elections. I'm not sure why you would say that only the three big parties were shown in those articles, since the Green Party and UKIP have been shown since before the 2015 election. Whether or not to use events is not relevant to the use of a Westminster system, and the consensus on the UK articles is the most we have for any election article anywhere on Wikipedia. Onetwothreeip (talk) 08:15, 14 September 2019 (UTC)
 * Precisely, I meant exactly that. UKIP and the Greens have been shown since the 2015 election, but it took sweat and blood for a consensus for adding them to be reached there even when they were already polling high in the polls throughout the 2010-2015 period. "Consensus on the UK articles" (which, btw, is limited to those articles) work very differently to consensus for other countries elsewhere in Wikipedia, as you don't typically find such level of opposition to adding parties polling in the teens. And yes, the Westminster system in general (and FPTP in particular) did have a large amount of influence in such discussions, because 1) back at the time it was argued that because UKIP wouldn't gain seats even if polling in the low teens and/or because it was not yet represented in the HOC, that it shouldn't be added in the tables, and 2) by-elections are common in Westminster FPTP systems but not elsewhere (where they may not even be legally provided), so this alone is relevant enough on the consideration on whether to add by-elections as events. Impru 20  talk 08:26, 14 September 2019 (UTC)
 * There wasn't any significant level of opposition to including UKIP once it was polling certainly more than 10%, and the Green Party never polled more than 10% but was still not much of an issue to include. "Sweat and blood" certainly wasn't my experience at the time, and I supported including both of them. Onetwothreeip (talk) 08:43, 14 September 2019 (UTC)
 * I intervened in the lengthy discussions that brought about UKIP and the Greens' inclusion, so no, I can assure you it wasn't an easy affair. For UKIP it took some time and many long discussions to add them, of which this one is just one example of it (and this despite the fact that they had been nearly tied, if not already outpolling the LDs in some polls, since early-to-mid 2012), and for the Greens, yes, it was much of an issue, actually. Impru 20  talk 21:24, 14 September 2019 (UTC)

Introduction sentences
@User:Ritchie92: You are a very careful user, but what you did this time is really contentious. Without any consensus, you removed some intro sentences from several articles, which have been there for years: 1, 2, 3, 4, 5. This was really bad. As your edits are clearly contentious and inconsistent with years of editing, not mine, please rollback them and start a discussion. If you achieve consensus good, if not let's keep the established versions. You cannot really say that I started an edit war because it was you who started to remove parts from several articles, knowing that at least one user (me) opposed it. --Checco (talk) 16:44, 13 September 2019 (UTC)
 * Please don't patronize. You don't need consensus to change anything on Wikipedia, this is a misrepresentation of what I did and of what Wikipedia is. I am allowed to be WP:BOLD and edit, otherwise nobody would edit. You added content on this page, legitimately, and I legitimately modified you edit by removing most of it. I still think what I did is right and there is no established consensus I went against. Also the fact that WP:OTHERSTUFF exists in some articles does not mean that one has to do the same errors in all other articles. And the fact that it's been there for years doesn't mean anything, if it goes against WP standards. --Ritchie92 (talk) 16:48, 13 September 2019 (UTC)

In the topic itself. This same exact section (with a meaningless title, by the way) appears in the pages Opinion polling for the 2013 Italian general election, Opinion polling for the 2008 Italian general election, Opinion polling for the 2006 Italian general election, and at the beginning of the section "Party vote" in Opinion polling for the 2018 Italian general election. It was also copied here, with the addition of the sentence "When there is a tie, no figure is shaded. The lead column on the right shows the percentage-point difference between the two parties with the highest figures." and of the sentence "Poll results use the date the fieldwork was done, as opposed to the date of publication. However, if such date is unknown, the date of publication will be given instead." which I find useful, but not limited to "Party vote", but also to "Coalition vote" etc (number one objection). My second objection is that the sentences that appear in all these pages contain useless information, since any person with the gift of eyesight can recognize that in the table the winning party is boldfaced, and that the color in the background corresponds to the color of the party. Also the title "Lead" is very clear, no need to explain what a "Lead" is. Otherwise we would also need to explain arithmetics. "When there is a tie, no figure is shaded": this sounds to me more like a guideline to the editors. In summary, in any case, all those sentences do not give any additional information to the reader, more than the numbers and layout in the table themselves. So this little "guide to the reader" on how to read the table I find really useless and also in the wrong tone for Wikipedia. --Ritchie92 (talk) 17:20, 13 September 2019 (UTC)


 * Thanks for reverting and for coming here to talk. We all need to be bold! But... we have to stop when someone else disagrees. That is what you correctly explained recently. I like the explanation (and I would have the same one in all the cited pages) and I will keep it. I added it here because it was missing, then you boldly removed it from all the articles. This is a no-big-deal issue, but hopefully other users have an opinion too. --Checco (talk) 06:16, 14 September 2019 (UTC)
 * Sorry but, as usual, the WP:ILIKEIT argument is not valid. You have to give a reason why that "manual to the table" is needed. Wikipedia is not a manual, the text should have only informative and sourced statements, not explanations about how the reader should read a table. Especially when the table is extremely visually simple! There is nothing non-trivial about it: color, bold, ordering are all self-evident. --Ritchie92 (talk) 08:19, 14 September 2019 (UTC)
 * I think that the explanation is useful, not redundant at all. It has been included for years in several articles. However, this is really a no-big-deal issue. If you decide to be bold again, I will not be an obstacle. Just make sure that all articles have the same introduction. --Checco (talk) 18:49, 14 September 2019 (UTC)

This edit by another editor looks like they also agree with my view. So would you agree if I make also the other older election pages consistent with this, as you said? --Ritchie92 (talk) 12:36, 19 September 2019 (UTC)
 * Sure. Go ahead. --Checco (talk) 06:31, 20 September 2019 (UTC)

Including Cambiamo in the table
Currently, GPF, EMG, SWG, Euromedia and Noto (the latest two are still not included in the table) are polling Cambiamo, the party founded by Giovanni Toti. Is this enough to add it in the table as a new column or should we wait for other polling firms? --Broncoviz (talk) 10:54, 12 September 2019 (UTC)
 * If we add Cambiamo, we should add Siamo Europei too. -- Nick.mon (talk) 11:37, 12 September 2019 (UTC)
 * As I said before, some pollsters include in their polls parties/lists we don't track: A1, PaP, NcI and more recently Cambiamo and Siamo Europei. To deal with this I think we could require parties to be polled at least by a certain number of pollsters (say, 3 or 5 or anything) in order for them to be included in the table and/or add side notes (as for articles about Poland, Hungary or the US Dem primary) when a party we don't include in the table is mentioned in a poll. TommasoM (talk) 12:01, 12 September 2019 (UTC)
 * Cambiamo has already been included by at least four pollsters (SWG, EMG, Euromedia, Noto), Siamo Europei by at least two (Euromedia, Noto—at 2.5%!). Thus, they should be mentioned in the table. --Checco (talk) 12:15, 12 September 2019 (UTC)
 * I would avoid creating a new table everytime some faction inside a party creates an independent movement. Let's wait a bit until things are settled, I would say. Sometimes pollsters decide to poll a certain new party just to experiment with the effects on the rest of the parties. --Ritchie92 (talk) 12:32, 12 September 2019 (UTC)
 * C! and SE are here to stay. The quicker we add them, the best (and less work for table fixers). --Checco (talk) 06:08, 13 September 2019 (UTC)
 * I think we must do things with a criterion. Some polls also measure PaP or PC, even around the same percentages to which C! appears (around 1%). What are the grounds on which we decide which parties to show in the table? As a matter of fact I think that C! has more rights to be included than SE. The latter has only one MP (Richetti probably?) and one MEP (Calenda) and is not polled by many pollsters, as instead C! is starting to be. I'm semi-serious now: a criterion could be whether it's polled by some of the most "prominent" pollsters, SWG and Ipsos for example. I say this because honestly I wouldn't base our decisions on some more "amateur" polls like GPF or TP. --Ritchie92 (talk) 08:12, 13 September 2019 (UTC)
 * You (accidentally?) tracked down a difference between CATI, CAMI and CAWI, let me explain it to those who don't really know them or doesn't speak Italian. CATI is the code name for the opinion polling made by calling people on their home phone numbers (with CAMI instead they call people on mobile phones), while CAWI means the surveys are made online. Another criterion could be if the polls are being commissioned by national and recognized organizations, both newspaper or television networks (Rai, Mediaset, La7, Il Corriere della Sera, La Repubblica). I would specify that polls must declare their sample size, as it's not the case of Noto, who's currently hosted at Porta a Porta but doesn't declare their methodological notes. Even the frequency of publications can be a criterion: poll sources who don't publish polls on a weekly or monthly frequency should be excluded. Only 9 pollsters would meet those criteria: Tecnè, SWG, EMG, Euromedia, Ipsos, Demopolis, Ixè (they're currently publishing on a weekly basis at Cartabianca), Index and Piepoli. As to create a new column for a party, I'd say that we add a new column only for parties that are polled by the 50%+1 of certified polling firm in their last opinion polling (and a threshold of 0.5%?). What do you think? --Broncoviz (talk) 15:28, 13 September 2019 (UTC)
 * I totally agree. Also, I think we should add a note when a party we don't track is included in a poll. TommasoM (talk) 16:18, 13 September 2019 (UTC)

EMG also started listing Italia Viva, at 3.4%. Lots of changes are coming and we should reflect these in the table! The RfC above should be closed ASAP so that we can hopefully abandon the government support column and have space for Cambiamo and Italia Viva soon. could you request a closing? It's been long time without any additional comments. --Ritchie92 (talk) 12:38, 19 September 2019 (UTC)


 * The "Others" column is now quite high. I would support the inclusion of at least one party not yet tracked. The Italian political life is quite dynamic, especially this year, and there might be snap elections in a few months' time, so this is important information that we should not leave aside.
 * Meanwhile, I suggest to write in a footnote[ad] (using Template:efn) the results of parties polling over 2% in the particular poll, as Wikipedians have done for the opinion polling in Poland (in fact the contributors write every other party in the footnotes). This will show which party is gaining steam and deserves attention, and perhaps a column of its own. Kahlores (talk) 20:01, 19 September 2019 (UTC)
 * Well, it's been only 4 days since the last comment in the RfC. While I don't think we need to wait for a full 30-day period to expire before requesting for a close, I think it would be nice to keep it open for any last-hour opinion, out of courtesy. Impru 20  talk 20:33, 19 September 2019 (UTC)
 * IV is currently polled by Euromedia, Noto, Ixé, EMG, Index and Tecné (most of these are still not published in the main page). We know already that there will be two parliamentary groups both in the Chamber of the Deputies and in the Senate of the Republic. I would probably start to report this party (and then decide whether to add Cambiamo or not). --Broncoviz (talk) 22:32, 19 September 2019 (UTC)


 * I agree with all those who support the inclusion of IV in the table. I also would like to see C! and SE. It is easier to remove them in case their polling numbers drop than adding them as time passes. --Checco (talk) 06:31, 20 September 2019 (UTC)
 * I didn't see SE in many polls actually. --Ritchie92 (talk) 08:07, 20 September 2019 (UTC)
 * In the end these discussions look really useless since anyway other editors come here and do whatever they think it's right without participating in the discussion, and I'm not blaming them. Now since consensus is in the process of being built, the C! and IV columns should be removed, in principle. But I guess that would be silly, so let's keep them. --Ritchie92 (talk) 13:25, 20 September 2019 (UTC)
 * True, but SE has polled higher than C! in a Noto poll. I would include it. --Checco (talk) 14:08, 20 September 2019 (UTC)
 * I think Noto would not meet the criteria about pollsters that was explained above. Anyway I would say let's take a decision anyway on when to add new parties to the table. I think that what User:Broncoviz was proposing was quite reasonable. --Ritchie92 (talk) 10:02, 21 September 2019 (UTC)
 * I almost gave up since almost nobody checked my proposal, I can do a table hoping that everyone else will understand better my proposal. --Broncoviz (talk) 12:15, 22 September 2019 (UTC)

Just wanted to summarize the situation of the current polls. I list the main pollsters first, and then add the less authoritative ones afterwards. Cambiamo! is currently polled by EMG, SWG and Euromedia, with the addition of GPF, Noto, TP, Quorum and Bidimedia; Italia Viva is polled by Euromedia, EMG, Demopolis and Tecnè (and probably today also the SWG poll will have it), with the addition of TP, Scenari Politici (Winpoll), Quorum and IZI; Siamo Europei is only polled by Euromedia and EMG, with the addition of Noto and TP. the latest Ixè poll does not show IV actually (next one might), and Index does not publish new polls since June 2019, AFAIK. --Ritchie92 (talk) 13:19, 23 September 2019 (UTC)
 * A new Index poll just came out. --Ritchie92 (talk) 18:23, 23 September 2019 (UTC)

Currently polled parties in real time
I made a table of the main pollsters with the parties that they poll. It's transcluded here from this subpage (earlier it was taken from my sandbox, but I think this is better; the reason why I don't want to have the table directly here is that it's a very long code and it makes reading the talk page difficult).

Feel free to edit and update it (or add missing information) as you wish. My idea is that this should be kept up-to-date also with new parties, so that we always have a clear picture of what's the status. If some editors have additional information, please put it in. Also if there are suggestions for the layout and the legend please share.

The main criterion is: I list in the tables the main pollsters which publish new polls periodically. The separation is between the "authoritative" pollsters above, and the others below, including the ones that are not very transparent like Noto, and the online ones like TP and Bidimedia. Feel free to add new lines.

The parties I chose are the ones which appear in at least two pollsters (PaP has been in all latest Noto polls except the very latest one). I use yellow (Yes/No) for some cases like the latter I mentioned and the latest Demopolis poll which excludes LS and includes IV (might be they include both in the next one?). Please correct mistakes if there are any! --Ritchie92 (talk) 11:18, 25 September 2019 (UTC)


 * Wonderful job! Can you just tell me why you put Noto (Porta a Porta, RAI) and Demos&Pi (la Repubblica), authoritative and renowned pollsters, in the second category? Many thanks, --Checco (talk) 12:34, 25 September 2019 (UTC)
 * I could say the same for Quorum–YouTrend (Sky Tg 24), isn't it authoritative? Anyway, great job! -- Nick.mon (talk) 13:17, 25 September 2019 (UTC)


 * Thank you. Noto, although it was commissioned by RAI, almost never publishes the technical details of their polls (sample size, actual fieldwork dates, polling method, etc) so it sounds a bit fishy to me to start with; Demos&Pi I just moved up.  Quorum is not as "historical" as the other pollsters, I appreciate a lot the work that they do on YouTrend (they're all young and very professional) but I didn't feel like they should be put together with the other more experienced pollsters. However we can discuss the prerequisites that the various agencies should have in order to be put in our "authoritative" list, I'm not the most expert. Asking  for their opinion on this. --Ritchie92 (talk) 18:27, 25 September 2019 (UTC)


 * That's a great work, congratulations! I will make a table of the pollsters, based on the criteria I proposed before (sorry, I was busy these days). In a short answer, as far as I remember Quorum does not publish a survey regularly, like every month, but they publish a new one when the client (Sky Tg24) request it. I didn't include at first Demos & Pi for the same reason, but it seems like La Repubblica has started to publish it more regularly. Let me know if you have any questions. --Broncoviz (talk) 09:16, 26 September 2019 (UTC)
 * Oh, and Bidimedia didn't poll SE or IV yet, so we should probably fix the table. Everyone can fix it or should we wait? --Broncoviz (talk) 09:29, 26 September 2019 (UTC)
 * Just fixed it. But you are free to edit the table, it's a subpage of this talk page. If you want to add another table for the specifics of each pollster you could either just add columns on the table that I made, or make a new one. I think the first option could be more practical. --Ritchie92 (talk) 10:09, 26 September 2019 (UTC)
 * Alright, I edited the table and added the columns. There are some things I didn't notice at first. EMG apparently makes their polls only with an online panel. They usually refer to it as a "telematic survey", and in some cases, it's reported they use only CAWI. Another thing is that apparently Il Sole 24 Ore hired Winpoll as their pollster and they published already 3 polls (25 August, 14 and 21 September). Last, but not least, I don't really know if the "Bi-weekly" frequency is correct. Porta a Porta does publish sometimes once a week, once every two or three weeks, depending on who knows what. About GPF I'm almost certain they publish bi-weekly (latest ones: 20-21 August, 4-5 September, so the next one will come in the next days?), and I don't know either if Piepoli will publish regularly with Povera Patria, since the last year they were on Porta a Porta and was replaced by Noto. They did publish some polls on La Stampa, so we'll see, since Povera Patria started three or four days ago. --Broncoviz (talk) 20:17, 26 September 2019 (UTC)
 * That's very nice! I added the typical numbers for the sample sizes, just to be more detailed. I don't know about the frequency of those in yellow, I guess we can wait and see, since this year some agencies changed their client. The question now is: how do we determine which ones are to be considered authoritative enough such that we can count on their result to decide which parties to include or not in our table? I guess EMG should go down now that you found that it only uses web interviews, for example. --Ritchie92 (talk) 08:54, 27 September 2019 (UTC)
 * I would like to wait for the opinion of other users about this, but I would be in favor. We should discuss about Winpoll too, and about introducing certain rules for adding a party in the main table, since, as I'm speaking, Cambiamo! would not meet the criteria I listed originally (the 50%+1 rule in certified pollsters), so we need to think about it carefully. I'm reading the Pollster Ratings of FiveThirtyEight and the one made by HuffPost in order to evaluate better the pollsters, but I don't want to introduce complicated things as we're just trying to determine when we should add a new party. --Broncoviz (talk) 18:22, 27 September 2019 (UTC)
 * One simple criterion could be that the pollster should have no "red" entries in the first four columns of the table above. This way we should exclude EMG, and include Quorum and Winpoll in the list of "higher-quality" polls, I guess? --Ritchie92 (talk) 12:23, 6 October 2019 (UTC)
 * However, in the article all opinion polls should be listed, do you agree? --Checco (talk) 14:26, 6 October 2019 (UTC)
 * @Ritchie92 I wouldn't add for now Quorum and Winpoll in the higher-quality polls, because the former doesn't publish regularly opinion polls, they publish the so-called supermedia instead, which is another thing. For Winpoll (and Piepoli too) the problem is that we don't have a clue about when they publish a new poll. Piepoli is part of another tv program and they published one poll so far, while for Winpoll I would wait a bit, but if you say otherwise it won't be a problem to add them in the top pollsters. Anyway, we should start to think whether to add a new party, since now we have this detailed table. --Broncoviz (talk) 12:02, 7 October 2019 (UTC)
 * So I guess "Not regularly" is a red, not a yellow. I would make the table such that a red entry means "this poll is not entirely high-quality". That's why we need to agree ASAP on a more "universal" criterion to decide on the polls quality. I guess for now yes, selecting which polls have to be inserted in the table is not the discussion topic here. If someone wants to limit the table to "higher-quality" pollsters only, then they should start a separate discussion. --Ritchie92 (talk) 14:37, 7 October 2019 (UTC)
 * Thanks! I just wanted to make sure. I oppose selecting polls. --Checco (talk) 06:35, 8 October 2019 (UTC)