Talk:Opinion polling for the 2020 Israeli legislative election

ADD FINAL PROJECTION

LINK : https://www.youtube.com/watch?v=ilmlnxu6q3c

Date	Polling firm	Publisher	Blue & White	Likud	Joint List	Labor- Gesher -Meretz	Shas	Yisrael Beiteinu	UTJ	Yamina	Otzma	Gov.
01 Mar 2020	Globe Elections UN	Globe Elections UN	33 27.15	34 27.66	14 11.11	9 7.33	8 6.80	7 5.49	8 6.16	7 5.91	0 1.42	56 46.53

• Do not add anything related to Globe Elections UN, this is an already reported case of spam advertising, with the main site having been blacklisted due to the repeated disruptive edits conducted on multiple articles through the last months. This page is protected, so the IP editor can't add it themselves on their own as is instead asking for someone else to do it for them. Check the main case threads here, here and here. Impru20^talk 18:40, 2 March 2020 (UTC)

References

Scenarios

There is no reason to have two sections, one for "regular" polling and one for "scenario" polling. We should instead have one section, with the date/pollster/publisher columns spread on multiple rows (using rowspan=x for x scenarios). Here's why:

1. Special polls are most often published together with the regular one, with the intent to see what the flows are. The quickest way to provide that information to the reader is to show both scenarios next to each other.
2. Pollsters often have large discrepancies with one another. Logically, they can poll the same scenario but have discrepancies too; scenarios make sense only when compared to the same pollster's other hypotheses. In case the reader wants to see what else has been published with this scenario, he or she can still track down the common footnote letter ^[d] or the merged cells.
3. In the event of a political realignment, the number of scenarios can be very large, at the risk of terribly inflating the scenario section. This is what happened on the article about the opinion polling for the 22nd Knesset.
4. All polling is hypothetical anyway, the "scenario" poll is just a less likely hypothesis than the regular polls.

On a technical level, the sorting function (by date and pollster) is perfectly compatible with rowspan.

Kahlores (talk) 22:57, 7 October 2019 (UTC)

I prefer it the way it was. The scenario description clearly appeared in bold at the title of the table, and the readers saw it clearly before they even reached the numbers. In your layout it's in a footnote, which is more difficult to find.

It gets even worse in case of split and merger scenarios, especially in a merger of non-adjacent parties. The table would look like a mess. ערן117 (talk) 04:24, 8 October 2019 (UTC)

I'll try to see if there's a more visible way to stress that a poll is a scenario. We could highlight the rows, like the exit polls last time. We could also add a small "Note" column.

The second issue can be resolved by arranging columns differently. You can see here how we can handle hypothetical polling seamlessly. Thanks to the left-right arrangement, it was possible to show all the competing scenarios on the scope of merger lists, whether with the right-wing parties (Zehut, New Right, URWP and Otzma) or the left-wing parties (Meretz, IDP, Labor and Gesher). There were cases spanning 2, 3 or all 4 parties. All you need to do is use colspan. If a list spans two parties that aren't next to each other, we can still add a footnote in both columns as is often done on other poliing pages.

The example also reminds us that polls that are very hypothetical at date N can become regular polls at date N+1. For instance, polls during the month of July were full of speculations about a hypothetical Yamina list. Since one scenario eventually became reality, it is meaningless to separate them. Political reality is also speculation.

Kahlores (talk) 11:37, 8 October 2019 (UTC)

1. I like the highlight of the scenario row.
2. I'll restate the scenario description issue: Footnotes are not suitable here, they are meant for additional information that is not crucial for understanding. The scenario description is important information that the reader can't understand the poll without. It should appear in the body of the article before the numbers. Maybe we can add the description in another row above the numbers.
3. I strongly object to adding any more columns, either Note or separation or whatever. We should minimize the number of columns, the table is already too wide, and we struggle to shorten column titles to decrease the width.
4. I personally support sorting the party columns in a left-right order regardless of the scenario issue, but this was very controversial in the articles about the last two elections, and we could never agree on it. If you insist to reopen this discussion, I suggest we create a new section for that rat hole.
5. The table should be sortable by any column. For example if I want to see when the government got the best results, I can sort by the Gov column.

ערן117 (talk) 04:53, 9 October 2019 (UTC)

Your point 5. makes sense, I didn't see it that way.

However, I'm not convinced by the long row describing the scenario (2.). Right now, with just one double poll, it is tolerable to our eyes. But have you imagined the kind of table it will be, once we have half a dozen times the same scenario? triple-scenario polls? When, during the next campaign, there is one month and a half of speculations by pollsters about various merger lists, as happened last July/August on both the left-wing and the right-wing? There would be table-wide rows all over, making the table illegible, the opposite of our goal. It is not a sustainable solution.

Maybe the problem is on your side for I don't see the table as too wide. It is much smaller than the Dutch page (13 parties, 17 columns) or the Italy page (10 parties, 15/18 columns). If it really becomes too big, we can still use a smaller font. So let's not limit ourselves too much.

A half-way scenario could be to reinstate the "Note" column and the footnote, but adding a one-word description, here "Sa'ar".

Kahlores (talk) 20:26, 9 October 2019 (UTC)

Yes, I agree that the table will become illegible over time. It's also incomprehensible after you sort it, because the scenario description line is no longer above the scenario numbers. The best solution is to revert to the separate section for Scenarios. ערן117 (talk) 04:31, 10 October 2019 (UTC)

I agree that there should be a separate section. Take a look at precedent: the September 2019 and April 2019 articles both have separate scenario polls sections. David O. Johnson (talk) 03:02, 22 October 2019 (UTC)

Precedents are not a sufficient reason. I made the case (at the top of this section) that the scenario sections should be removed, for 5 different reasons. These reasons apply for all articles and those previous articles as well (I have already prepared a version for September and could easily do one for April).

So I would like you to offer counter-arguments to the points I brought up. Kahlores (talk) 02:14, 23 October 2019 (UTC)

It's been a month, and no answer. To be sure, here's how it would look if we merged the two tables:

Israel: election polls, all in one table

Color key:   Speculative poll Date Polling firm Publisher Note Joint List Labor- Gesher-Meretz Green Blue & White Yisrael Beiteinu Sa'ar party Likud New Right Nat. Union United J. Home Shas UTJ L R Current seats 13 6 3 1 33 8 32 3 2 2 – 9 7 57 55 28 Feb 2020 Election silence starting end of Friday before election day 15 Jan 2020 Deadline to submit the candidate lists 13 Jan 2020 Midgam Channel 12 13 9 – 34 7 31 6 5 8 7 56 57 Right 13 9 – 34 7 32 10 8 7 56 57 13 Jan 2020 Labor-Gesher and Democratic Union merger Date Polling firm Publisher Note Joint List Labor- Gesher Dem. Union Green Blue & White Yisrael Beiteinu Sa'ar party Likud New Right Nat. Union United J. Home Shas UTJ L R 10 Jan 2020 Maagar Mochot Israel Hayom 15 5 5 (0.2%) 34 7 30 4 (1%) 5 7 8 59 54 Left 13 11 – 33 7 30 6 (1%) 6 7 7 57 56 URWP 14 5 5 – 33 7 29 5 8 7 7 57 56 Right 14 5 5 – 34 7 29 11 7 8 58 55 9 Jan 2020 Midgam Arutz Sheva URWP 13 5 4 – 34 7 32 6 4 8 7 56 57 NR+NU 13 5 4 – 34 7 32 9 (2.7%) 8 8 56 57 8–9 Jan 2020 Panels Politics Makor Rishon 14 5 (2.4%) – 35 8 31 7 – 4 8 8 54 58 URWP 14 5 (2.4%) – 35 8 31 7 6 7 7 54 58 Yamina 14 5 (2.4%) – 35 8 32 10 (2.1%) 8 8 54 58 Right 14 5 (2.4%) – 35 8 33 10 8 7 54 58 8 Jan 2020 Midgam Walla! 13 5 4 (0.2%) 35 8 33 6 (1.1%) (2.8%) 8 8 57 55 Left, Right 13 9 – 34 8 32 8 8 8 56 56 7 Jan 2020 Meretz and Democratic Choice agree to run together as the Democratic Union list 2 Jan 2020 Panel Project HaMidgam Channel 13 URWP 13 5 4 – 36 6 34 8 (2.9%) 7 7 58 56 Left, URWP 13 10 – 35 7 33 8 – 7 7 58 55 2 Jan 2020 Kantar Kan 11 URWP 13 6 5 – 35 7 32 5 (3%) 7 8 59 54 2 Jan 2020 Midgam Channel 12 13 5 4 (0.0%) 35 8 33 6 (0.9%) (2.6%) 8 8 57 55 URWP 13 5 4 (0.0%) 34 7 32 6 4 8 7 56 57 1 Jan 2020 Panels Politics Maariv URWP 13 5 (2.9%) – 36 8 30 6 7 8 7 54 58 Sa'ar, URWP 13 4 – – 34 7 6 28 6 7 8 7 51 62 29 Dec 2019 Midgam Channel 12 13 5 4 – 34 8 32 5 4 8 7 56 56 26 Dec 2019 Likud reelects Benjamin Netanyahu as its leader 20 Dec 2019 The Jewish Home and Otzma Yehudit agree to run together 19 Dec 2019 Stav Shaffir announces a run as part of the Green Movement Date Polling firm Publisher Note Joint List Labor- Gesher Dem. Union Blue & White Yisrael Beiteinu Likud New Right Nat. Union J. Home Otzma Yehudit Shas UTJ L R 19 Dec 2019 Smith 103fm URWP 13 5 4 35 7 32 5 4 8 7 57 56 Sa'ar, URWP 13 5 4 33 6 30 8 5 9 7 55 59 18 Dec 2019 Panels Politics Makor Rishon 14 5 (2.1%) 38 7 34 7 (2.7%) (1.6%) 9 7 57 56 Sa'ar 14 5 (2.6%) 34 7 30 10 4 – 8 8 53 60 15 Dec 2019 Kantar Kan 11 13 6 5 35 8 31 6 – – 9 7 59 53 Sa'ar 13 5 5 34 7 27 8 4 – 10 7 57 56 12 Dec 2019 Maagar Mochot Israel Hayom 14 6 4 37 8 31 5 (2%) (1%) 8 7 61 51 Sa'ar 15 5 4 33 7 31 7 (2%) – 11 7 57 56 12 Dec 2019 Midgam Channel 12 13 5 5 35 8 32 6 (2.8%) (1.3%) 8 8 58 54 Sa'ar 13 5 5 35 8 26 9 4 – 8 7 58 54 12 Dec 2019 Smith Maariv URWP 12 5 4 35 7 33 5 4 8 7 56 57 12 Dec 2019 The Knesset is dissolved for lack of a new government 10 Dec 2019 Panel Project HaMidgam Channel 13 13 5 5 37 8 33 6 – – 6 7 60 52 Sa'ar 13 4 5 35 8 29 7 5 – 7 7 57 55 9–10 Dec 2019 Midgam Walla! 13 5 4 35 8 33 6 (2.1%) (1.3%) 8 8 57 55 8 Dec 2019 Kantar Kan Radio 13 5 4 35 7 34 6 (2.4%) (1.9%) 8 8 57 56 8 Dec 2019 Kantar Kan 11 Sa'ar 13 5 4 33 8 32 8 – – 9 8 55 57 5 Dec 2019 Midgam Channel 12 13 5 5 34 8 33 6 (2.2%) (1.9%) 8 8 57 55 5 Dec 2019 Maagar Mochot 103fm 12 4 4 34 8 32 6 4 – 8 8 54 58 28 Nov 2019 Maagar Mochot Israel Hayom 13 4 4 35 8 33 6 (2%) (1%) 9 8 56 56 28 Nov 2019 Maagar Mochot Maariv 13 4 4 35 8 33 6 – – 9 8 56 56 Sa'ar 14 4 – 33 9 28 12 – – 11 9 51 60 26–27 Nov 2019 Smith Maariv 13 5 5 35 8 33 6 – – 8 7 58 54 Sa'ar 13 5 4 35 7 31 6 4 – 8 7 57 56 26 Nov 2019 Midgam Channel 12 13 5 4 34 9 33 6 (2.3%) (1.2%) 8 8 56 55 Sa'ar 13 6 4 35 9 26 9 (2.9%) – 10 8 58 53 25–26 Nov 2019 Midgam Walla! 13 5 4 34 8 33 7 (2.3%) (1.2%) 9 7 56 56 25 Nov 2019 Panels Politics Channel 12 12 5 4 37 7 30 6 4 – 8 7 58 55 22 Nov 2019 Panel Project HaMidgam & Statnet Channel 13 13 4 4 36 8 33 6 4 – 6 6 57 55 21–22 Nov 2019 Maagar Mochot Israel Hayom 13 4 6 34 7 33 7 – 8 8 57 56 21 Nov 2019 Kantar Kan 11 12 5 5 33 8 33 5 4 – 7 8 55 57 21 Nov 2019 Attorney General indicts incumbent PM Benjamin Netanyahu on corruption charges 21 Nov 2019 Midgam Channel 12 13 5 5 33 9 33 7 – 8 7 56 55 15 Nov 2019 Maagar Mochot Israel Hayom 13 6 6 33 9 31 5 – 10 7 58 53 10 Nov 2019 Panel Project HaMidgam Channel 13 13 5 4 35 9 34 4 4 – 6 6 57 54 31 Oct 2019 Maagar Mochot 103fm 12 7 5 34 8 33 6 – 9 6 58 54 22 Oct 2019 Panel Project HaMidgam Channel 13 13 5 4 34 8 33 4 5 – 7 7 56 56 10 Oct 2019 The Yamina alliance splits up. Date Polling firm Publisher Note Joint List Labor- Gesher Dem. Union Blue & White Yisrael Beiteinu Likud Yamina Otzma Yehudit Shas UTJ L R 3–6 Oct 2019 Midgam Walla! 13 6 5 33 7 33 7 – 9 7 57 56 Sa'ar 13 6 5 33 7 26 12 – 10 8 57 56 17 Sep 2019 Election results 13 6 5 33 8 32 7 (1.9%) 9 7 57 55

Kahlores (talk) 21:38, 27 November 2019 (UTC)

I'm not so sure if I like this unified section. I took a look at your sandbox for the Sep. election to see how such a graph would end up, and it seems that there are some weird situations (like the empty section before Democratic Israel was created and when you have hypothetical unions between two ideologically distant parties), however I do agree we should look into ways to clean up the scenarios section, which looked extremely hectic and confusing in the Sep. elections page. Gibzit (talk) 19:25, 28 November 2019 (UTC)

Thanks for your feedback.

The first problem has been solved: IDP's empty column was removed (it was a leftover from a previous table).

The second problem—hypothetical unions— can't be fixed in 100% of cases if, indeed, they unify parties from all over the place. But we can maximize this number by arranging columns very wisely. Before the September election, dozens of merger hypotheses were tested (I counted 28), but just 1 poll (on 19 July) in my table shows a merger between non-neighboring columns. How good is that?

I am still open to other remarks, since my goal is to find a consensus layout that solves our issues. Kahlores (talk) 23:21, 29 November 2019 (UTC)

Once again, the "scenario" section is becoming bloated and illegible, whereas this table is well streamlined. All potential mergers and scenarios are managed perfectly.

So what are your other concerns? If you have none, can I move it into the article? and likewise for the two previous elections? Kahlores (talk) 15:34, 14 December 2019 (UTC)

why hasnt any one updated the polls in a weak 213.8.151.40 (talk) 21:41, 1 January 2020 (UTC)

You should probably start a new section; to my knowledge, no polls have been released in the last week. David O. Johnson (talk) 23:22, 1 January 2020 (UTC)

I've updated the template, but there was just one poll missing. Kahlores (talk) 16:15, 2 January 2020 (UTC)

Seven new polls, and the table is still unbeaten at showing the various hypotheses in one look. Kahlores (talk) 01:03, 4 January 2020 (UTC)

Sources

New poll line graphing method

So as Israel hurls towards a third election, it seems that is page will once again become important. I believe we could learn some less from September elections' page. One thing that I believe we could significantly improve is the line graph that appears at the top of the page. The graph of the September elections poll is not bad, but it's certainly not good either. Compared to other polling graphs, it is chaotic and confusing, and so I believe we should improve it.

I believe the problem lies with the fact that unlike other poll graphs, our graph does not show a "polling average", instead it shows the polling of the day. What this means is that any outlier poll causes the graph to go jagged and confusing. If the Likud gets 27 in a poll on sunday and 33 in a poll on tuesday, the result on the graph will look quite bad.

How do we solve this? Well I believe that we should take a page from a professional poll graphing org: realclearpolitics.org.

Here is their graph for the recent 2020 USA democratic party primaries: https://www.realclearpolitics.com/epolls/2020/president/us/2020_democratic_presidential_nomination-6730.html

You can see that despite the fact that the primaries have quite a few outlier polls, their graph remains quite stable and isn't as confusing as ours. The method that they use for the graph is quite simple, every time a new point on the graph is added when a new poll is released they average the last 4 polls (now including the new poll of course), instead of only taking into account the polls released on that same day.

So in short, how does this system work? (for future reference for people who want to edit the graph). Once a new poll is released and you want to add it to the line graph, average it with the three previous polls that were released and put that number in the line graph. For example if the new poll has the Likud getting 33, and the last 3 polls show the Likud getting 30, then you should 33+30+30+30/4 = 30.75 and put that in the line graph.

As a proof of concept, here is this method applied on the poll graph from the September election (without Zehut or Otzma):

As you can see, this method makes the graph much smoother, clearer and in my opinion, prettier. I hope that you accept this proposal for the graph for the new election and that we can implement it in the next line graph. Gibzit (talk) 18:51, 25 November 2019 (UTC)

Thanks for making this, I like it a lot. If we are doing it this way for this page, could we go back and do the same for the opinion polling pages for April and September? Jacoby531 (talk) 19:18, 25 November 2019 (UTC)

Using the "basis" parameter instead of "monotone" produces a smoother graph as well. I think that would be easier and just as good. --Gbuvn (talk) 18:10, 14 December 2019 (UTC)

Question about the latest poll

Why is Otzma not included? (Jan 2 2020) Idan (talk) 12:33, 3 January 2020 (UTC)

Otzma is part of UJH, it is included. ערן117 (talk) 13:25, 3 January 2020 (UTC)

Right-Wing Scenarios

I'm not sure anymore what is real and what is a scenario in the right-wing. URWP is a scanario beacuse the NU refuses to join it. Yamina is a scenario because the New Right refuses to join it. But NU independent run is also a scenario, because the NU has never announced any such thing, and it doesn't seem to want that; it wishes to unite the whole religious Zionism. The NU has threatened to withdraw, but it hasn't announced that officially, and this threat doesn't sound serious. So until the NU makes up its mind, it seems like everything is a scenario... ערן117 (talk) 05:20, 4 January 2020 (UTC)

https://hamodia.com/2019/12/31/smotrich-breaks-silence-otzma-yehudit-deal/

It seems like Smotrich definitely won't join Otzma and Jewish Home. I added refs on the article page for the election about Tkuma being in talks for a joint run with the New Right. David O. Johnson (talk) 05:44, 4 January 2020 (UTC)

So just to clarify, once we find out which scenario will actually happen, will we add the polls for whichever scenario ends up happening to the main table? I ask because it seems like all of the recent polls are included under scenarios and there hasn't been a new poll in the main table in some time. It would be strange if there was a gap of two or three weeks in the main table even though there have been plenty of polls. Jacoby531 (talk) 15:34, 4 January 2020 (UTC)

no polls?

I'm wondering why there are no polls added to the official poll list. Everything, it seems, is below, in a "scenario" list. This is ridiculous. If you are going to have a page about Israeli election polls, you should put Israeli election polls on it in a visible place. If your Israeli election polls are "special" then put it in a "special" place. If ALL your Israeli election polls are "special" then they are not special at all. There have been 6 polls since the last one on the list. ALL six are marked as "special". Stop looking at the "special" section (IE: well this is a scenario, so it goes with scenarios) and look at the article as a whole; every poll that comes out that does not fit into the section of "real" polls only makes the quality of the article drop. This is harming the article as a whole. Nickjbor (talk) 22:56, 4 January 2020 (UTC)

I totally agree with you. I have already prepared a unified table in my own sandbox (or also see the section above called #Scenarios).

I need your support to bring it on the article because there are always people complaining for X or Y foolish reason, and reverting. Kahlores (talk) 23:34, 4 January 2020 (UTC)

I see why we should include the unified table . I'd support it if you start an RFC. David O. Johnson (talk) 23:57, 4 January 2020 (UTC)

I support including split and merger scenarios in the main table if we can move the parties involved to adjacent columns. This way it is clear by the split/merged cells what the scenario is, and we don't need any further explanations. Other types of scenarios (i.e. Sa'ar leading Likud) I prefer leaving in a separate section as it is now, and I don't want the "note" column.

Another question is which scenarios, if any, we can include in the graph without making it look like a mess. ערן117 (talk) 01:59, 5 January 2020 (UTC)

Should we integrate the scenarios into the main table?

It's been a few months since the idea was first raised.

I am for integrating the scenarios. This way we wouldn't have an apparent gap in polls. David O. Johnson (talk) 22:44, 7 January 2020 (UTC)

for integrating merger & split scenarios only.

against integrating "Sa'ar leading Likud" or any other types of scenarios. ערן117 (talk) 04:44, 8 January 2020 (UTC)

@Kahlores, Jacoby531, and Nickjbor: Ongoing discussion you all may want to contribute to. David O. Johnson (talk) 07:38, 8 January 2020 (UTC)

I agree with ערן117; for integrating merger and split polls, but not alternative leadership polls. Jacoby531 (talk) 14:22, 8 January 2020 (UTC)

I am against integrating the scenario polls, especially with Sa'ar, but for integrating "nu runs separate" Braganza (talk) 17:42, 9 January 2020 (UTC)

The main problem is with the right wing parties jewish home, new right, national union and jewish power. The solution is to stick to official anouncements. Right now, according to official anouncements, the new right runs separately and so does the national union, while the jewish home and jewish power signed an agreement to run together on 20.12. Therefore, the two polls under "National Union independent run" should be integrated into the main part. Otherwise, we should regard all previous polls assuming that jewish home and national union run together without official anouncement as scenarios too. אראלי44 (talk) 17:51, 9 January 2020 (UTC)

The National Union never announced that it runs separately. That's why I put it in the Scenarios section. On the contrary, Smotrich said he wants to unify the entire religious Zionism, and even threatened to withdraw rather than risk falling under the threshold. The problem right now is that the official announcements of the different parties are inconsistent with each other. ערן117 (talk) 18:35, 9 January 2020 (UTC)

The surveys were only carried out in January not December/November or something. Braganza (talk) 17:54, 9 January 2020 (UTC)

I am for integrating mergers and splits

I am against integrating new leadership polls (Saar leading Likud) and X forms new party polls (Saar forms new party) which should be kept in a scenario section. Gibzit (talk) 08:18, 11 January 2020 (UTC)

If we do integrate the scenarios, then we should make sure that once lists are closed (15 Jan) a new table is created (similarly to how a new table is created with every merger right now) so it is clear what parties and alliances are finally running. Gibzit (talk) 08:23, 11 January 2020 (UTC)

How is "Sa'ar forms new party" any different than other splits? ערן117 (talk) 09:23, 11 January 2020 (UTC)

The other splits involve already existing parties which have had many different splits and mergers with each other. Saar forming a new party is completely speculative and would make the graph much more confusing, especially if we get more polls for other personalities who could form parties, are we going to add a column for each one? Gibzit (talk) 11:46, 11 January 2020 (UTC)

I agree that the Sa'ar party scenario doesn't justify bloating the table with an additional column, as it appears in one poll only. I suggest that if any new party scenario becomes real, we add it to the main table. For example think about the April election and the "Gantz forms a new party" scenario. ערן117 (talk) 07:13, 12 January 2020 (UTC)

It is good to see people finally realize the need for a unified table.

So it seems all of us (5) agree to include merger and splits scenarios (Braganza excepted but I don't see his point).

As for Sa'ar, I suggest we keep the practice started by Spanish Wikipedian Impru20 who created an "Alternative scenarios" section, including polls with alternative leaders, or with a new party.

However, there is one thing we shouldn't be missing: most scenario polls are double-polls, that is, the same panel of respondents is asked two questions in a row, the second of which is more hypothetical than the other.

So the very meaning of these double polls is to allow for comparisons. Now, given the fact that you don't want to include "Sa'ar as Likud leader" in the main table, or "Sa'ar forms a new party", I suggest to provide comparisons like this:

double-poll as it is right now in the sandbox

Date	Polling firm	Publisher	Note		Joint List	Labor- Gesher	Dem. Union	Green	Blue & White	Yisrael Beiteinu	Sa'ar party	Likud	New Right	Nat. Union	United J. Home		Shas	UTJ		L	R
1 Jan 2020	Panels Politics	Maariv	URWP		13	5	(2.9%)	–	36	8	30		6	7			8	7		54	58
			Sa'ar, URWP		13	4	–	–	34	7	6	28	6	7			8	7		51	62

double-poll as it could be shown in the "scenarios section"

Date	Polling firm	Publisher	Note		Joint List	Labor- Gesher	Dem. Union	Green	Blue & White	Yisrael Beiteinu	Sa'ar party	Likud	New Right	Nat. Union	United J. Home		Shas	UTJ		L	R
1 Jan 2020	Panels Politics	Maariv	Sa'ar, URWP		13 ±0	4 –1	–	–	34 –2	7 –1	6 +6	28 –2	6 ±0	7 ±0			8 ±0	7 ±0		51 –4	62 +4

In any case, I'll start to work to add the 10 new polls of these past few days, and I'll bring the table - leadership scenarios excepted - on to the article. Kahlores (talk) 14:29, 11 January 2020 (UTC)

The second alternative is difficult to understand, the first is much clearer. But please remove the Note column, it is unnecessary because it is clear that if the NU and UJH cells are merged, then it means that they run together, and if the Sa'ar and Likud cells are split, then it means that they run separately. ערן117 (talk) 07:13, 12 January 2020 (UTC)

Im against those edits, this makes the table unreadable. Also in the previous article the notion of sorting the table according the left-right political spectrum was rejected because the spectrum was incorrect and changed subjectively. (The religious parties aren't the most right wing parties, its Otzma Yehudit and Yisrael Beitenu) right-left wing spectrum changes regarding the issues, its different by social, security, Israeli-palestinain conflict etc. Scenario polls shouldn't be included in the main table because they just a media gimmick and don't have a real value, they are hypothetical. The main table should be as formal as can. Also the Left wing/Right wing column is wrong because Lieberman isn't left-wing but he does against Netanyahu. This all was discussed before in previous articles. The best option is to sort by number of seats because in that way readers can notice the gain/loss of power in each poll. Sokuya (talk) 07:37, 14 January 2020 (UTC)

There seems to be a majority here for integrating the scenarios except the Sa'ar related ones. I'm counting 5 for and 2 (including you) against. If you want to wait some more time for more comments, I'm OK with that. Regarding the column order, we only need the NR-NU-JH-Otzma and Labor-Meretz-Green next to each other, because they are the only parties that split and merge in this election. UTJ, Shas, YB and all the others can be placed wherever you want, but they can't be sorted by number of seats if we are to have the splitting and merging parties next to each other . ערן117 (talk) 10:41, 14 January 2020 (UTC)

I'm against that sorting, its too confusing. Every merge or split should be reflected in the table by starting new header. So it will be easier to read each time. It's hard to read the table with all that information combined. It is better to split the table. Order should be by number of seats. It's the most important thing in the table. Sokuya (talk) 15:28, 14 January 2020 (UTC)

Edit

The new table has been kept up-to-date on my sandbox over the last 2 months, while the previous system was in place. However, I could not keep track of each and every edit and mostly copied-and-pasted the references of the last polls.

So please check whether your past edits have been incorporated, particularly regarding:

1. new references added days after the first edit
2. mistakes corrected days after the first edit (numbers, names, etc.)

Kahlores (talk) 00:15, 14 January 2020 (UTC)

---

@ערן117, David O. Johnson, Jacoby531, and Gibzit:

As I said last week, "I need your support to bring it on the article because there are always people complaining for X or Y foolish reason, and reverting."

Here we are . . . Kahlores (talk) 13:40, 14 January 2020 (UTC)

@Impru20:, can you please explain why you reverted? You didn't even vote or comment in the previous section.

I understand Sokuya's arguments, I just have to disagree with him, except for the L/R columns where he does have a point. The L column can't turn into an alternative government because MKs Hendel and Hauser won't sit with the Joint List, so the info in that column is useless. The R column is identical to the Gov column we have now, so we might as well keep the Gov column. ערן117 (talk) 19:04, 14 January 2020 (UTC)

@Eran117: Why? Has the consensus at Talk:Opinion_polling_for_the_September_2019_Israeli_legislative_election#Coalitions changed now on the issues of left-to-wing ordering of parties and column sums? Because it doesn't seem so. I'm in favour of integration of the polls from the various scenarios (and there is seemingly a consensus here for that), but I think Kahlores has been maybe a little too bold by attempting to re-insert a scheme which was already discarded for previous elections for similar reasons, and for which there does not seem to be any new reason for adding those specifically for this article. I don't think it is fair play to try to game the currently established consensus on those issues by taking an advantage of the integration issue; integration does not need neither the ideology sorting from left to right neither the addition of the "Left" and "Right" sums in columns, something which was already brought down in the past and which has little to do with the main issue of scenario integration. Impru20^talk 20:15, 14 January 2020 (UTC)

My edit yesterday was not a bold edit, as I've been patiently working on this for months, and garnering support.

Sadly, the issues most people bring up are minor, compared with the current chaos of the "scenarios" section.

You may partly or fully disagree with the left-right arrangement, but please show us another axis that never fails to manage mergers, splits and the battle for prime ministership?

For the first two, you can still have an ad hoc arrangement. For example, here's how the Israeli Wikipedia does it. Parties are arranged by seat count, except for alliances (right-wing and left-wing), that are explained with a special row in the header (below, Hebrew with English links) and the footer.

תוצאות סקרים מהכרזת הבחירות (12 בדצמבר 2019) ועד להגשת הרשימות (15 בינואר 2020)
תאריך מפרסם מכון מחקר מושבים בכנסת כחול לבן הליכוד הרשימה המשותפת ש"ס ישראל ביתנו יהדות התורה עוצמה יהודית ימינה העבודה-גשר המחנה הדמוקרטי הבית היהודי האיחוד הלאומי הימין החדש                         תוצאות הבחירות לכנסת העשרים ושתיים 33 32 13 9 8 7 X 7 6 5 תוצאות הבחירות לכנסת ה-22 - לאחר פיצולים 33 32 13 9 8 7 X 2 2 3 6 5 15 בינואר הסתיימה הגשת הרשימות לבחירות לכנסת ה-23 תאריך מפרסם מכון מחקר כחול לבן הליכוד הרשימה המשותפת ש"ס ישראל ביתנו יהדות התורה הבית היהודי - עוצמה יהודית הימין החדש - האיחוד הלאומי העבודה-גשר-המחנה הדמוקרטי
מופיעים רק סקרים המתחשבים באחוז החסימה בחישוב המנדטים

For the English Wikipedia, this would give us this (let's call it plan B):

Date	Polling firm	Publisher	Blue & White	Likud		Joint List	Labor- Gesher-Meretz		Green	Shas	Yisrael Beiteinu	UTJ	New Right	Nat. Union	United J. Home		L	R
			33	32		13	10		1	9	8	7	3	2	2

However, plan B doesn't show what all polls or poll pundits calculate, namely the split for Prime ministership, which is the reason why there have been 3 elections in a year. It's like moving heaven and earth to avoid the risk of being slightly editorializing. We're not at all editorizalizing. Every poll which gave one camp over 60 seats, was published in the Israeli press with that news in its title!

Proof, if you needed one, that Plan A isn't pushing any POV. It's just following mainstream Israeli media, showing the fault lines of Israeli party politics.

So let's choose. Plan A or B?

• Plan A:
  • Kahlores (talk)
  • I prefer plan A, but I can live with plan B as a compromise, and without the L/R columns please. ערן117 (talk) 05:48, 15 January 2020 (UTC)
• Plan B:

---

Minor issues that could be talked elsewhere and later:

• The L count. Of course, the L count cannot assess the chances of an L government, but you could call it G as it can assess Gantz's future recommendations (reminder, Odeh claimed that Balad's 3 MKs did not recommend Gantz only for tactical reasons).
• Liberman's position. Of course, Liberman is not a historical centrist, but Liberman re-positioned himself as the champion of a 'liberal unity government' with B&W and Likud, campaining against both religious and Arab parties. With just one dimension, where else would you put him?

Kahlores (talk) 22:01, 14 January 2020 (UTC)

The fact is that you have been slowly building up your preferred version of the table in your sandbox, awaiting until the right time to enforce it. That you now try to mix the left-right axis and the L/R columns with the scenario issue is absurd: you were already building it up before the current chaos as you describe it arose, so you were obviously not taking such an issue into consideration for the table's structure (unless you could foresee the future back at the time, which I hardly think so). The fact is that neither the left-right axis nor the L/R columns have anything to do with the scenario integration issue, so I do not know why should we be pressed on a seemingly "take-it-or-leave-it" position.

please show us another axis that never fails to manage mergers, splits and the battle for prime ministership? I got mentioned in the discussion above as a reference for "Alternative scenarios" schemes which I applied in other articles, and I've never used a left-to-right arrangement of any sort (much less any L/R column scheme), so yes, it is very possible to manage mergers, splits and other such issues without it, but the key is to consider that not all alternative scenario polls should merit inclusion into the main tables. This is because some scenarios are about alternatives that will never materialize (not before the 2020 election, at least). They may be of interest for readers, but do not need to be included within the main row of polls because their importance is secondary. For example: all opinion polls asking on Gideon Sa'ar as Likud leader should not be included in the main table, because it is a certainty as of now that he won't be leading Likud into the election.

On the alliances and mergers, this is more tricky due to the high number of them, but I'd call for awaiting until the final lists are confirmed (this seems to be tomorrow at 10pm, right?), because some of the hypothesized alliances may not come into fruition and, as such, could be separated from the main table as mere speculation. As a must, the main table should include: 1) Hypotheses reporting on mergers/alliances/splits that are confirmed for the 2020 election; this is, party lists that will be put to voters on 2 March 2020. 2) Hypotheses that rely on previous election results, though some leeway for adjustment can be taken if required (for example, as an extinct alliance, Yamina could live without a full column for practical purposes, instead using a merger of JH-NU and NR). Footnotes could also be used to help, in a similar fashion as in Opinion polling for the 2016 Spanish general election#Nationwide polling. Once full lists are confirmed, I can work out a model table. Impru20^talk 00:49, 15 January 2020 (UTC)

This is because some scenarios are about alternatives that will never materialize (not before the 2020 election, at least). They may be of interest for readers, but do not need to be included within the main row of polls because their importance is secondary. – well said. Some scenarios polls are only for media stunts, to bring more readers to the news site and engage readers, so there will be something to talk about everyday. It's the media trying to fill the time until the "showtime" of the elections. That is another reason why I think is should be separate from the main table, its make it harder to separate the wheat from the chaff. Sokuya (talk) 18:48, 15 January 2020 (UTC)

OK, I've integrated the scenarios I could with the existing column order by number of current seats. I had to leave out all URWP polls and one Yamina poll, for which we need another column order. Can people please vote on Kahlores' suggestions for the column order? ערן117 (talk) 17:24, 18 January 2020 (UTC)

No, because somehow those who don't agree with either because of the reasons already exposed seem to be repeteadly ignored. We are not required to chose between any one of these. It has been already said that both proposals (but specially A) go against currently established consensus both on the left/right arrangement and the L/R columns, reasons for which have not changed. You and Kahlores were both in favour of these in the past discussion only a couple months ago and lost the case for it: get on with it and move on. Impru20^talk 18:23, 18 January 2020 (UTC)

The current arrangement is not looking good. At least some table were left unchanged, thankfully. It's not readable, you can't understand what is the scenario from the table, it's just a mess. Sokuya (talk) 20:11, 19 January 2020 (UTC)

Left Alliance

Hebrew wikipedia is still fighting if the two small parties (IDP and Golans party) made the alliance with labor. So I am waiting for the official lists tmr then we can update. Idan (talk) 07:40, 14 January 2020 (UTC)

Semi-protected edit request on 16 January 2020

In the table that shows the polling results since 15 January 2020 (the newest table) Likud and Blue&White have been mixed up. According to the table Likud holds 33 seats in the current Knesset while Blue&White holds 32. In reality it is the other way round. This needs to be changed Urgently! Yackshack (talk) 06:12, 16 January 2020 (UTC)

I assume this was done because MK Gadi Yevarkan defected from B&W to Likud. But I think he might be forced to resign in order to be included in the Likud list, then he will be replaced by another B&W MK, and the seat count will revert to what it was before. I suggest we wait for further developments. ערן117 (talk) 06:31, 16 January 2020 (UTC)

As I thought, Yevarkan has resigned, and Yorai Lahav-Hertzano from B&W will replace him. So we're back to the same seat count as after the last election. ערן117 (talk) 03:38, 17 January 2020 (UTC)

Why would he resign tho...? Idan (talk) 18:31, 18 January 2020 (UTC)

He is forced to resign. He can't be an MK for one faction and run with another existing faction. That was the problem of Orly Levy in the first round: she was elected as part of YB and she didn't resign, so she couldn't run with B&W which includes Yesh Atid, and she had to run separately. ערן117 (talk) 05:13, 19 January 2020 (UTC)

References

Green Party

Why is it still even listed? 82.166.134.110 (talk) 16:49, 29 January 2020 (UTC)

Stav Shaffir is an MK for the Green Party, so it gets listed. David O. Johnson (talk) 16:52, 29 January 2020 (UTC)

I too think we should remove it after it withdrew. ערן117 (talk) 16:29, 30 January 2020 (UTC)

Agreed. The Green Party isn't running so there's no need to include it.--Rxtreme (talk) 02:42, 3 February 2020 (UTC)

But Stav Shafir still holding a seat, so the party have a seat in the 'current seats'. Sokuya (talk) 20:47, 4 February 2020 (UTC)

This article is about polls. The current seats are there only for comparison. If the polls don't include the GP, is shouldn't be in the table. ערן117 (talk) 13:19, 6 February 2020 (UTC)

New polls after the election

What do we do with this poll? https://www.bhol.co.il/news/1087458 Sokuya (talk) 08:41, 13 March 2020 (UTC)

An article named Opinion polling for the next Israeli legislative election seems to be the best fit, I think. David O. Johnson (talk) 13:30, 13 March 2020 (UTC)