User talk:Psephomancy

From Electowiki
Jump to navigation Jump to search

Ask about Wikipedia cite templates[edit source]

Could you import more cite templates from Wikipedia? I was thinking of cite:arXiv in particular (so I could properly cite the PAMSAC article), but others may be useful too. I'm not familiar with the nitty-gritty of Wikipedia programming, so I don't know what would have to be copied over. Kristomun (talk) 14:46, 29 January 2020 (UTC)

Yes, I've been meaning to import some more. Do you want to start a list?
Actually Special:WantedTemplates has the ones that people have tried to use but don't currently exist. I was going to add Template:Clarify‏‎ Template:Cleanup‏‎ Template:Redundant Template:WhoPsephomancy (talk) 17:29, 29 January 2020 (UTC)
I don't want to start using the templates before they're there. So I'd say Template:Cite_arXiv and a general stub template (Wikipedia has lots, but we could probably make do with a single one for everything for now). Could you also change the "doi:" link on cites with a doi tag to go to Wikipedia's doi article? It currently points to a nonexistent article here. Kristomun (talk) 16:48, 4 February 2020 (UTC)
Template:Stub already exists, no?
doi should work now. You can make interwiki redirects to Wikipedia. — Psephomancy (talk) 02:44, 5 February 2020 (UTC)
I thought that was just part of the page because it didn't have any additional formatting (e.g. icon like Wikipedia's stub templates). I've made a suggestion for a stub template with an icon here: User:Kristomun/Template:Stub, but it appears it needs a script, Module:Asbox, to work - I can't just copy over Template:Asbox from Wikipedia. Could you import Module:Asbox, or is it a simple matter of just copying the source from Wikipedia's module article over to a module article with the same name here? Kristomun (talk) 11:44, 5 February 2020 (UTC)

Oh I added Template:Cite arXiv, but it requires a bot to complete the citation? That won't work here. — Psephomancy (talk) 03:07, 5 February 2020 (UTC)

Whoops, didn't know that. Kristomun (talk) 11:44, 5 February 2020 (UTC)

Importing templates is finicky. I basically followed these instructions:

  1. Go to https://en.wikipedia.org/wiki/Special:Export
  2. In "Add pages manually:", add the Template:Name (Probably best to only do one template at a time)
  3. Select all of "Include only the current revision, not the full history", "Include templates", and "Save as file"
  4. An XML file will be downloaded
  5. Go to https://electowiki.org/wiki/Special:Import
  6. Click "Browse" and find the .xml file
  7. Enter an interwiki prefix from https://electowiki.org/wiki/Special:Interwiki for rewriting links to point to the original wiki. (I could swear there was a wikipediawikipedia prefix yesterday, which would point to en.wikipedia's Wikipedia: namespace, but now I don't see it?) (Actually this should probably be wikipedia because it shows up in edit history like https://electowiki.org/w/index.php?title=Help:CS1_errors&action=history, while the actual links always need manual modification anyway)
  8. Add an edit comment, which will show up in the Import log and the edit history of the page.
  9. Select "Import to default locations"
  10. Press "Upload file"
  11. It will list "Importing pages..." and each imported page, explaining whether and why it was imported, and then "Import finished!" (hopefully)
  12. 'After it has been imported open the template for editing, check the list below the edit box "Pages transcluded onto the current version of this page:". If there are any red links repeat the above steps for these templates as well. You might also need to copy styles in MediaWiki:Common.css page and Javascript in MediaWiki:Common.js before everything works. Finally edit the templates to customize for your wiki.'

I just tried this with Template:Stub

Psephomancy (talk) 15:31, 5 February 2020 (UTC)

... and Template:Stub doesn't even have any text in it, it just depends on Template:Asbox, which has "Wikipedia" hardcoded into it somewhere and I don't have time to do this right now so I just reverted it. Wikipedia templates have a rat's nest of dependencies which make them far more complicated than we need them to be. It's not as simple as copying the HTML and CSS of the template and having it work fine here. — Psephomancy (talk) 15:40, 5 February 2020 (UTC)

That seems to be w:Module:Asbox under function p.main Kristomun (talk) 17:23, 5 February 2020 (UTC)

I don't have permission to import things with https://electowiki.org/wiki/Special:Import. Could you import w:Template:db-user? Kristomun (talk) 16:30, 6 February 2020 (UTC)

No longer as important as I found out how one's supposed to move pages. Kristomun (talk) 21:18, 6 February 2020 (UTC)

Links and references[edit source]

I've found https://tools.wmflabs.org/citer/citer.fcgi quite useful for creating Wiki-type references. Could we somehow draw new editors' attention to this so that links are more likely to be references/citations than just plain URL links? Is there any way to automatically find stale/dead links and flag them, as well? Presumably Wikipedia has something like the latter to keep links current. Kristomun (talk) 17:04, 4 February 2020 (UTC)

That's great! I've been using Wikipedia's sandbox to generate the links in VisualEditor and then copy and paste them into Electowiki's VisualEditor. I think we need to make a EW-specific Help page... — Psephomancy (talk) 02:44, 5 February 2020 (UTC)

User:Kristomun, I got Citoid partially working, so we can do references like Wikipedia. It isn't working for web links, though. — Psephomancy (talk) 01:30, 19 February 2020 (UTC)

What's keeping Citoid from working with web links? I've been trying to refine some references in Uncovered_set, and a few journal articles lack DOI/ISSN numbers and Citer happens not to work on them. Would it be possible to get web to work, or is it currently impractical to fix? (Alternatively, can Citoid be modified to handle other citation formats, e.g. BibTeX?) Kristomun (talk) 22:09, 14 March 2020 (UTC)
I don't know. I suspect it's something we can fix locally, but I haven't figure out how yet. I think there are several types of IDs it doesn't work for. Making a list of what works and doesn't work would probably help figuring it out, or finding who to contact to ask about it. — Psephomancy (talk) 23:43, 17 March 2020 (UTC)
User:Kristomun See https://phabricator.miraheze.org/T5357Psephomancy (talk) 15:13, 27 March 2020 (UTC)

Electowidget[edit source]

Is Electowidget still part of this wiki? Kristomun (talk) 20:39, 4 February 2020 (UTC)

You can ask User:RobLa about that. — Psephomancy (talk) 02:44, 5 February 2020 (UTC)
Thanks for the ping, User:Psephomancy. My answer for User:Kristomun: Electowidget in it's old form was effectively mothballed around 2013-ish. I fiddled around with creating a Lua-based replacement, but my Lua skills aren't up to the the task yet. Still, I think a Lua-based rewrite has the most long term viability. -- RobLa (talk) 14:50, 5 February 2020 (UTC)
Should we then remove or move the Electowidget pages until such a rewrite happens? Kristomun (talk) 15:05, 5 February 2020 (UTC)
User:Kristomun: yes, that would be fine. I redirected many of the pages over to User:RobLa/Electowidget. It'd probably be best to take conversation about the Electowidget over to User_talk:RobLa/Electowidget. -- RobLa (talk) 04:33, 6 February 2020 (UTC)

Not noticing there's a move option[edit source]

Looks like I did the FAB->MCAB rename incorrectly. Could you move User:Kristomun/Floating Approval Bucklin to Maximum Constrained Approval Bucklin‎, delete User:Kristomun/MCAB duplicate, and delete this particular revision: https://electowiki.org/w/index.php?title=User:Kristomun/Floating_Approval_Bucklin&oldid=7450 ? Kristomun (talk) 21:17, 6 February 2020 (UTC)

I moved/deleted the pages. The revision can't be deleted, though parts of it can be hidden, but it's not normal to do that on Wikipedia unless there's a serious problem like copyright violation, doxxing, etc. I assume we'll do the same here? — Psephomancy (talk) 06:14, 7 February 2020 (UTC)
I agree, following Wikipedia's guidelines is a good thing to do. Disregard my request to delete the revision :-) Kristomun (talk) 11:55, 7 February 2020 (UTC)

Criterion compliance logic[edit source]

Talk:Distributed_Score_Voting led me to think about criterion compliance tables. Would it be possible to add logic to a wiki so that e.g. if we know that monotonicity, LNHelp, LNHarm, and mutual majority are incompatible, then if someone adds to a method's page that this method passes the latter three, the system automatically adds that the method is not monotone?

Or would it be possible to add tags that say "this method is monotone" and then have a big comparison of methods page/table that automatically updates itself for every method, filling in whether the method is monotone, not monotone, or its monotonicity status is unknown? Then if we have a policy that criterion compliance claims must be referenced or proven, we'd get a better comparison of methods page than the one on Wikipedia.

But perhaps this requires additional software that would be hard to combine with the existing Miraheze platform, e.g. https://www.semantic-mediawiki.org/wiki/Semantic_MediaWiki or https://stackoverflow.com/questions/1854794/. Hence my question. Kristomun (talk) 10:17, 8 February 2020 (UTC)

I was hoping categories could serve some purposes like this (Electowiki:Help#Dynamic_page_Lists), but I think they're not powerful enough.
meta:Extensions says "Semantic MediaWiki – alters MediaWiki too much, use Cargo instead"
mw:Extension:Cargo
mw:Extension:Wikibase_Client is also available. Would either of those do what you want? (We can turn on any of the extensions on meta:Extensions with a few clicks.) — Psephomancy (talk) 19:37, 8 February 2020 (UTC)
Quite a lot could be done with such an idea. Consider for example that if a voting method passes the majority criterion and ordinal cloneproofness, it must also pass mutual majority, since a mutual majority's preferred set of candidates are essentially a clone set where if you eliminate all but any one of them, the remaining candidate is the majority's 1st choice, and must win. So an automatic criterion checker could suggest this for voting methods that are proven to pass the first two criteria, for example. BetterVotingAdvocacy (talk) 19:58, 21 February 2020 (UTC)
I don't think those extensions have any automatic capabilities, they would just allow for structured information. I think the automatic part would have to be done by a bot.
Not sure if the amount of work to write a bot counteracts the amount of work to just make changes manually. — Psephomancy (talk) 21:06, 21 February 2020 (UTC)
A mutual majority set isn't a clone set. Suppose voters constituting a bare majority rank some permutation of {A,B,C,D} above everybody else, and the remaining 49% voters rank every candidate in a random order. Then {A, B, C, D} is a mutual majority set but not a clone set. So a method of the type "determine what candidates are ranked consecutively in some order by absolutely everybody, replace them with a single candidate, and run FPTP on the result" would be cloneproof and pass majority but would fail mutual majority.
I get what you mean, though. The absolute deluxe option would be to have something like a theorem prover and then being able to specify implications (e.g. "Smith criterion" and "plurality criterion" => not "Mono-add-top criterion", or max two of "Mutual majority criterion", "Later-no-harm", "Monotonicity"), and have the system automatically fill criteria it can infer from already specified pass/fail data. Then Electowiki could become a theorem prover library for criterion compliances. But I have no idea how you'd integrate such a thing with the Mediawiki software, and it may well be overkill. Kristomun (talk) 20:52, 21 February 2020 (UTC)

User:Kristomun,

mw:Extension:Wikibase_Client is probably what we want? Essentially a local version of Wikidata?

Wikidata already has voting system criteria and voting systems, but not many relationships between them yet. Looks like User:RobLa has been looking into this, too:

Wikidata itself has the same notability rules as Wikipedia, so we can't put newly-invented voting systems on there, but it seems like we could access the existing info on Wikidata as well as the local database? I'm not sure how it all works yet. — Psephomancy (talk) 04:15, 9 February 2020 (UTC)

Does Wikidata support implications or tagging data properties from the page itself? My thinking is that if users have to go to a separate page to fill in criterion compliances then they're probably not going to do it, and even less so if they manually have to keep the wiki page in sync with the data attributes. That's why Semantic Wiki seemed a better solution, but I see that it modifies Mediawiki too much. I'll investigate Cargo later. Kristomun (talk) 12:56, 10 February 2020 (UTC)

On another note, there's been a bit of back-and-forth at Sequentially Spent Score. I've been thinking about marking claims without evidence with some sort of "evidence needed" tag, in the vein of Wikipedia's citation needed tag. Can I go ahead and make such a template, and then make that the policy for claims without evidence? Kristomun (talk) 12:56, 10 February 2020 (UTC)

That's fine with me. You mean evidence that is not a citation, though? What form would that take? — Psephomancy (talk) 05:44, 12 February 2020 (UTC)
Things like the proofs I've given on the weighted positional method page, where the proof is given directly on the page. It might come back to bite us if people start submitting very involved or very subtly wrong proofs, but we could try to deal with that if/when it happens. Kristomun (talk) 20:52, 21 February 2020 (UTC)

Set theoretic article a good idea?[edit source]

Given how often we discuss set-theoretic concepts (not just things like the Smith set, but even most criteria; for example, if you have 5 candidates and a Pareto preference for the first over the second, then in some sense the "Pareto-compliant set of candidates" is all candidates except the 2nd), should we have a dedicated article for it? Reading things like Schulze's article on the Schulze method might be a lot easier if we have a voting theory-specific guide for understanding set theory. BetterVotingAdvocacy (talk) 19:29, 22 February 2020 (UTC)

That's probably fine. What would it say that's different from Set theory? — Psephomancy (talk) 19:32, 22 February 2020 (UTC)
I'd think stuff like the Nakamura number, discussion on the relation of various criteria to each other, and in particular, discussion of how set theory can be used in the context of PSC to find which winner sets are PSC-compliant or not could be very useful. Basically, it could be a glossary of the intersection between voting theory concepts and set theory. I'll make an article, but I'm not too sure how to write it so that it can maximally encompass all of the different voting theory concepts in its various sections. BetterVotingAdvocacy (talk) 19:54, 22 February 2020 (UTC)

OR/notability criteria[edit source]

As I've understood it, this wiki is more relaxed than Wikipedia about OR/notability concerns, because it covers a lot of methods that are "original" or not academically published anywhere, much less in use by governments or organizations. But there should probably be some kind of limit, so that new users can't just come up with a ton of methods or method modifications (e.g. Landau//Minmax, "MJ then a tiebreak by Range", "Ranked pairs but only admit A>B if A covers B" etc.).

What should the threshold be, and what should its logic be? It seems difficult to find a balance that admits interesting methods while not opening for the possibility that the wiki will get too cluttered. Kristomun (talk) 22:04, 14 March 2020 (UTC)

Personally, I think a lot of the current articles are pretty much in the sweet spot. Maybe something like "Smith winner" doesn't make much sense (or at least, is better off as a subsection to Smith set rather than its own article), but for the most part, a lot of these articles really do deserve their own space. For example, Condorcet//FPTP is notable because it's one of the simplest Condorcet methods possible, Smith//Score because several ideas can be thought of for mixing the cardinal and pairwise information together in finding the winner, etc. It may help if you can point to some articles that you find to be particularly bad in this respect, then I might be able to give some more ideas. BetterVotingAdvocacy (talk) 23:12, 14 March 2020 (UTC)
User:Kristomun, I don't know. I'm pretty inclusionist about such things. My threshold might be lower than anything currently on the wiki. Can you find the worst-case example, from your perspective, and we'll see how I feel about it? :D — Psephomancy (talk) 23:51, 17 March 2020 (UTC)
I was mainly thinking of Smith-Schulze. Being included as part of the Schulze article might give the impression that it is a variant of the Schulze method also proposed by Schulze. Kristomun (talk) 11:03, 18 March 2020 (UTC)
In more general terms, though, there could be a situation where we have some very rigorous proofs (with lots of references) and others that someone just thought of at that moment (and that has a greater chance of being false). Since I allow proofs without references as long as the proof is given, it could be difficult to distinguish them. I haven't happened upon such things yet, but if I do, I'll let you know.
On the other hand, Wikipedia might have such problems itself. E.g. the Schulze STV "proof" that it is monotone (and that has propagated into its election methods comparison table), that Schulze said he doesn't know whether is true. Kristomun (talk) 11:03, 18 March 2020 (UTC)
Kristomun, I've edited the Schulze article to make it clear that "Smith-Schulze" (which I proposed) was not proposed or analyzed by Schulze. I did email him about it by the way, and he seemed okay with it being used as a teaching tool for people to understand Schulze. In the future, I suggest that it's a good idea to just make any concerns that you have known on the discussion page, or simply add disclaimers to the article. BetterVotingAdvocacy (talk) 17:01, 18 March 2020 (UTC)
One way to address this concern is to create a wiki article with links to different users' voting methods (hosted on their own page) with the disclaimer that there may be errors. That way, users can create as many methods as they want, and then plop a link to a category of them on their user page from this article. BetterVotingAdvocacy (talk) 19:43, 19 March 2020 (UTC)

Majority and utilitarianism[edit source]

User:Psephomancy, what would be a good name for an article that encompasses both majority rule and utilitarianism? I think it'd be interesting to make the overarching point that majority rule only counts the number of people who prefer one outcome over another, while utilitarianism also accounts for their strength of preference. BetterVotingAdvocacy (talk) 03:43, 18 March 2020 (UTC)

Would that be an advocacy page then? We've talked about those a few times but I don't think any exist yet except in userspace. — Psephomancy (talk) 05:34, 20 March 2020 (UTC)
No, I was thinking more of what the page should be called that encompasses both as a category. I ultimately settled on using the Utility page for this, with cardinal utility for utilitarianism and ordinal utility for majoritarianism. BetterVotingAdvocacy (talk) 06:15, 20 March 2020 (UTC)

I'd like to make a page regarding "allocation methods". Specifically, it should serve as a category for highest-averages and largest remainder methods, and talk a little about some of the impossibilities of creating a perfect allocation method. I think "allocation method", "seat allocation method", "apportionment method" would be good names for such an article, but what do you think? BetterVotingAdvocacy (talk) 05:29, 23 March 2020 (UTC)

Would it be a good idea to have an article explaining some notations often used in voting theory? For example, I think it'd be good to have some redirects such that someone reading Schulze's PDF on his voting method could quickly learn what all the symbols means, as well as find any notations specific to voting theory, such as the formatting of the ballots. BetterVotingAdvocacy (talk) 05:52, 28 March 2020 (UTC)

Advocacy for gov formation reform[edit source]

User:Psephomancy, You recently said "I don't think advocacy movements for government formation really makes sense. " Here is a list of some movement groups

There are a ton of different movements. These all have to do with how you go from votes to a governing body and have nothing to do with the voting method itself. Maybe there is a better generic term for it than "Government Formation" or maybe we should deal with each independently. I do not really know or care. My goal it to try and get some of this content onto electowiki. --Dr. Edmonds (talk) 03:47, 30 March 2020 (UTC)