Talk:Ranked voting/Archive 3

All my work edits

I spent the better part of 2.5 days, and he/you undid all of them without discussing any of the changes. I fixed spelling errors. I ran things through Grammarly. I discussed significant changes. You don't get to undo everything all at once without talking about them. You can't just make a blanket change to undo all of it. I was told to be bold, but I wasn't. I discussed everything significant. I justified almost every change with a description.

Starting at the bottom, and working my way up:

'1) The categories at the bottom of the page'

I think it was good to keep these:

1. Category:Psephology
2. Category:Public choice theory
3. Category:Social choice theory
4. Category:Cardinal electoral systems
5. Category:Electoral systems

'2) "See also" ' I fixed the "See Also" so that it was in a column. I used the 'colbegin' and 'colend' command, that helps with formatting. I added a bunch of cross references to relevant information. Please explain why you deleted them. What didn't you think was relevant? Myclob (talk) 17:03, 30 December 2021 (UTC)

I agree that your changes were pretty reasonable and moreoever made in discrete, explained edits. Colin's bulk revert was inappropriate and he should give specific concerns rather than rejecting it all at once. Reywas92^Talk 17:15, 30 December 2021 (UTC)

Thanks. I'm a very bad writer. I'm an electrical engineer. But my wife is a technical writer. She got a masters in it, and graduated at the top of her class. She works as a Sr. Technical writer. I respect the field, but need help. I use Microsoft Word, and bought the premium version of Grammarly. I spent a lot of time on this article, trying to make it better. I understand that other people spent a lot of time also, writing it. I respect the work that went into writing it. I get my feelings hurt when I spent lots of time writing things, and then my wife edits them. But I swallow my bride, and almost always accept her suggestions. Anyways, I have kids and I didn't spend much time with them. I took Christmas break, and could be doing things that are more fun, but I read the article and figured I would spend my time trying to make suggestions. I almost didn't check to see if my changes stuck, but when I checked this morning every single one of my changes over the past few days was un-done. I probably need to go for a walk and calm down, but it is very frustrating. Anyways, I am trying to be respectful of other people, and if other people disagree with my changes, I want to give my best explanation, but I will accept the majority rule... Myclob (talk) 17:55, 30 December 2021 (UTC)

'3) "Drawbacks"'

I discussed changes to "Drawbacks of ranked voting. Why would you undo those changes without participating in the discussion? Myclob (talk) 17:08, 30 December 2021 (UTC)

This is how it read before:

• Ranked voting elicits more information about voter preferences than is revealed through an first-past-the-post ballot, but this comes with certain costs. Voters are confronted with more complicated ballot slips to complete, and the counting procedure – depending on the nature of the voting method – is more complicated and slower, often requiring mechanical support.

Before, it wasn't very clear. It sounded like it was concerned about sharing of too much information like it was a privacy issue. It may have sounded better for someone with a master's in political science. Still, Wikipedia is written for the general public. Please make changes one at a time instead of reverting dozens of hours of work all at once.

This is what I changed it to:

• Ranked voting allows voter to select candidates that more accurately represent their preferences than first-past-the-post ballots. However, they are more complicated and the counting procedure, if done manually, is more complicated and slower.

Please be respectful of other people's efforts and work and make changes one at a time, re-wording things if you can make them sound better or improve them. Myclob (talk) 16:57, 30 December 2021 (UTC)

I propose this is "vague" and "unnecessarily complex" (see the style guide):

• "Ranked voting elicits more information about voter preferences than is revealed through an first-past-the-post ballot."

Isn't this better?

• "Ranked voting allows voter to select candidates that more accurately represent their preferences."

I would have been fine with you proposing any of the following:

• "Ranked voting communicates more information from the voter than a first-past-the-post ballot"
• "Ranked voting allows the voter to communicate more information than a first-past-the-post ballot"

However, to just undo all of my work from the last few days is very rude and disrespectful. Myclob (talk) 18:04, 30 December 2021 (UTC)

'4) "mathematical properties"'

It used to say: 

• "The mathematical properties of 'a voting method' need to be balanced against its pragmatic features, such as its intelligibility to 'the average voter'."

I changed it to:

• "Mathematical properties of 'voting methods' must be balanced against pragmatic features, such as intelligibility to 'average voters'."

I am not a great writer, but this is very poorly written. Grammarly helped me realize that it was very wordy. If you put it in Microsoft Word it will also tell you it is overly wordy. I changed it to this, but it wasn't easy. It took me time to wordsmith all these changes. Why did you come along and just delete all my work? Myclob (talk) 16:57, 30 December 2021 (UTC)

'5) "Empirical comparisons" Run on sentence '

It used to have this very long run-on sentence. made it two sentences. Why did you undo my work? Myclob (talk) 17:00, 30 December 2021 (UTC)

Before:

• Empirical comparisons can be performed using simulated elections. Populations of voters and candidates are constructed under a spatial (or other) model', and the' accuracy of each voting method – defined as the frequency with which it elects the candidate closest to the centre of the voter distribution – can be estimated by random trials. Condorcet methods (and Coombs’ method) give the best results, followed by the Borda count, with IRV some way behind and first-past-the-post worst of all.

After:

• Empirical comparisons can be performed using simulated elections. Populations of voters and candidates are constructed under a spatial (or other) model'. The accuracy of each voting method – defined as the frequency with which it elects the candidate closest to the centre of the voter distribution – can be estimated by random trials. Condorcet methods (and Coombs' method) give the best results, followed by the Borda count, with IRV some way behind and first-past-the-post worst of all.

'6) The sentence that begins with "Logical voting criteria..."'

"extrapolating" is better than "extrapolations of"

", reasonable measures" is better than "plausible criteria" without a comma. This article is used for the general public. I have a paid subscription to Grammarly, and this article should be written for the general public. Please don't bulk und-my changes, if what I had sounded better. Myclob (talk) 17:27, 30 December 2021 (UTC)

"each other" is not required. The word "contradict" implies the comparison. It is redundant.

Here is the before:

• "Logical voting criteria can be thought of as extrapolations of the salient features of examples into infinite spaces of elections. The consequences are often hard to predict: initially plausible criteria contradict each other and reject otherwise satisfactory voting methods."

Here is the after:

• "Logical voting criteria can be thought of as extrapolating the salient features of examples into infinite spaces of elections. The consequences are often hard to predict: initially, reasonable measures contradict and reject otherwise satisfactory voting methods."

This is still very poorly written, but lets make sure we don't go backwards, before we go forwards: "extrapolating the salient features of examples into infinite spaces of elections." This either doesn't belong, or it needs to be put in plain English. This is ridiculous. "The consequences are often hard to predict: initially plausible criteria contradict each other and reject otherwise satisfactory voting methods." What is that saying? I have multiple professional certifications, but I have no idea what this is saying and neither does anyone else, unless we can put this in plain English. Anyways, lets start by doing the simple fixes. Reasonable is not jorgon. Plausible is. Wikipedia says not to use jargon. Criteria is jargon. Wikipedia says not to use jargon. See the manual of Style (https://wiki.riteme.site/wiki/Wikipedia:Manual_of_Style). It says: "editors should avoid ambiguity, jargon, and vague or unnecessarily complex wording."

I agree that many of your edits were sound and belong in the article. But the process of making a lot of edits, having them reverted, and then discussing them on the talk page is a very normal one. We're now at step 3 in WP:Bold, revert, discuss, and I'm very glad that you're actively using the talk page and opening discussions on each of the main edits under dispute. I would like to ask, though, that we please confine the discussion to one place and keep it organized. A bullet point for each edit you're considering is a good idea. But when there are multiple new talk page sections every day it becomes incredibly hard to navigate the page and hold a conversation. Let's have a straightforward organized discussion within this section about the edits that you want to keep restored. I hope to come back and consider things more fully when I have a bit more time (hopefully soon), but for now I've posted notices on a few relevant talk pages to try to attract a critical mass of editors here to make sure that the outcome of these discussions is verifiable and well sourced. - Astrophobe (talk) 17:49, 30 December 2021 (UTC)

User:Astrophobe re: "the process of making a lot of edits, having them reverted, and then discussing them on the talk page is a very normal one." Thanks again for the response. However, is it typical to have 60 or 70 edits undone all at once, without any discussion? Doesn't that seem more like a personal attack or insult?

re: "I would like to ask, though, that we please confine the discussion to one place and keep it organized." I got rid of the headings, and subheadings. I was thinking of using a "page break|label=" between each subject. I wasn't sure how to do bullets for each topic, and so I just started numbering them manually. Tell me if there is anything I can do to help, or feel free to re-organize things how you prefer. Myclob (talk) 18:20, 30 December 2021 (UTC)

Thank you! I think it's readable now, and I think it's fair to put the onus on people responding point by point and not reverting all at once. I know that it can be harsh to have your work undone, and I can't pretend to know anyone else's motivations for editing, but I believe in assuming good faith. I don't think it's healthy to believe that other volunteers are personally attacking you unless they really overtly say something nasty, and I don't think that a revert is evidence of an intent to personally insult. It just means that someone didn't like the changes you made. There are some good tips on how to deal with this sort of situation at WP:LEMONADE. - Astrophobe (talk) 20:08, 30 December 2021 (UTC)

'7) The sentence that begins with "The simplest form of comparison..."'

• I am a bad writer, but even I know that "what many people would consider to be a weakness of IR" is poor writing.
• "what many people would consider a weakness of IRV" is much better, and should not have been undone.

Also, this: "other examples show purported weaknesses in other methods" is worse than: "other examples show asserted flaws in different methods."

• Before: "The simplest form of comparison is through argument by example. The example in the present article illustrates what many people would consider 'to be (to be is not required) a weakness' of IRV; other examples show 'purported weaknesses' in 'other' methods."
• My changes: "The simplest form of comparison is through argument by example. The example in the present article illustrates what many people would consider 'a weakness' of IRV; other examples show asserted flaws in different methods."

Again, this breaks the Wikipedia Style guide requirement for "editors should avoid ambiguity, jargon, and vague or unnecessarily complex wording." Asserted is more common the purported, weaknesses is actually over used, and is less specific. Also "other" is overused, and Grammarly or Microsoft word suggest "different" instead of other. I think it reads better, but don't care about this so much. However, "purported" is not written at the general public. Myclob (talk) 18:34, 30 December 2021 (UTC)

'8) The sentence that begins with "Smith's method reduces..."'

Please tell me why it was wrong for me to remove ",which is." Grammarly said it was unnecessarily wordy.

'9) The sentence that begins with "Coombs' method..."'

Please tell me why you like the run-on sentenses better.

Before:

• Coombs' method is a simple modification of IRV 'in which the candidate' eliminated in each round is the one with (too wordy) most last-place preferences rather than with fewest (LOL. This is just wrong) first-place preferences (so C rather than B is eliminated in the first round of the example and B is the winner). Coombs’ method is not Condorcet-consistent but nonetheless satisfies the median voter theorem. REF:"B. Grofman and S. L. Feld, “If you like the alternative vote (a.k.a. the instant runoff), then you ought to know about the Coombs rule” (2004)." It has the drawback that it relies particularly on voters’ last-place preferences which may be chosen with less care than their first places.

My wording:

• Coombs' method is a simple modification of IRV. The candidate eliminated in each round has the most last-place preferences rather than the fewest first-place choices (so C rather than B is eliminated in the first round of the example and B is the winner). Coombs' method is not Condorcet-consistent but satisfies the median voter theorem. Ref:"B. Grofman and S. L. Feld, "If you like the alternative vote (a.k.a. the instant runoff), then you ought to know about the Coombs rule" (2004)" It has the drawback that it relies mainly on voters' last-place preferences',' (this comma is needed) which may be chosen with less care than their first places.

'10) The sentence that begins with "A Condorcet completion elects ..."'

Please explain why this:

• A Condorcet completion elects the Condorcet winner if there is one, and otherwise falls back on a separate procedure for determining the result. If the Borda count is the fallback we get Black's method; if we use IRV we get Tideman's ‘Condorcet-Hare‘.

... is better than this:

• A Condorcet completion elects the Condorcet winner if there is one and otherwise falls back on a separate procedure for determining the result. If the Borda count is the fallback, (I believe we need this comma) we get Black's method; if we use IRV, (I believe we need this comma) we get Tideman's 'Condorcet-Hare'.

I believe the grammar changes I made were valid. Myclob (talk) 19:38, 30 December 2021 (UTC)

'11) Moving the paragraph that begins with "Ranked voting is different from cardinal voting" under "other methods" instead of the intro to this topic'

I don't think we should talk about other voting methods under the intro to this voting method. That is the whole purpose of the "Other Methods" category. That is why I moved it. Please explain why you undid this change. Thanks. Myclob (talk) 19:43, 30 December 2021 (UTC) 

'12) The paragraph that begins with "Copeland's method...."'

Please explain why you undid my edits that used consistent spaces after the period at the end of the sentence. Also, "therefore" needs commas. Also, Grammarly says "has the drawback that" is poor writing and unnecessary. You just explain what the drawback is. It is too wordy.

This is how it was before

• "Copeland's method is simple and Condorcet-consistent but has the drawback that for certain patterns of voter preferences with no Condorcet winner it will yield a tie however large the electorate. Its advocates therefore..."

this is the wording that I did (with the help of Grammarly) and that you reverted:

• Copeland's method is Condorcet-consistent and straightforward but, for specific patterns of voter preferences (with no Condorcet winner), it will yield a tie however large the electorate. Its advocates, therefore,

'13) The paragraph that begins with "Determining the minimax winner...."'

User:Colin.champion I replaced "particularly simple" with "straightforward". I also replaced "as electing" with "selecting. I also deleted an unnecessary comma and added a necessary one. Please justify reverting my edits if you want to discard them.

'14) The paragraph that begins with "The minimax system determines a result by constructing a results table...."'

Grammarly helped me realize that this was a particularly poorly written paragraph. I'm not saying who wrote this paragraph, I just don't think my grammar corrections should have been rejected. Let me start with a before and after, and then justify each change: Before:

The minimax system determines a result by constructing a results table 'in which there is an' entry for every pair of distinct candidates 'showing' how often the first is preferred to the second. Thus since 51 voters prefer A to C and 49 have the opposite preference, the (A,C) entry reads ‘51:49’. In each row we identify the least favourable (i.e. minimal) result for the first candidate (shown in bold), and the winning candidate is the one whose least favourable result is most favourable (i.e. maximal). In the example the winner is B, whose least favourable result is a win while the other candidates’ least favourable results are slightly different losses. After" The minimax system determines a result by constructing a results table'. An' entry for every pair of distinct candidates 'shows' how often the first is preferred to the second. Thus since 51 voters prefer A to C and 49 have the opposite preference, the (A,C) entry reads '51:49'. In each row, (the comma is required) we identify the least satisfactory (i.e., minimal) result for the first candidate (shown in bold)'. The (this was a run-on sentence and needed to be broken up) winning candidate 'has' the fewest least preferred results and is most favourable (i.e., maximal). In the example, the winner is B, whose least Preferred result is winning. In contrast, the other candidates' least favourable results are slightly different losses.

• "favourable" is used 6 times in 2 sentences. That is not good writing. Also, it uses British spelling. We don't have to use only geographic specific synonyms. That is why I replaced 4 of them with synonyms including "satisfactory" and "preffered."
• "in which there is an" should just be replaced by "an"
• "showing" in this context should be replaced by "shows"
• ". In each row" needs a comma
• I replaced a very long run-on sentence with two sentences.
• this is a little complex, but I replaced this confusing sentence: " and the winning candidate is the one whose least favourable result is most favourable (i.e. maximal)" with ". The winning candidate 'has' the fewest least preferred results and is most favourable (i.e., maximal)
• "In the example" needs a comma
• Finally I replaced this complex sentence: "whose least 'favourable' result is 'a win while' the other candidates’' least favourable results are slightly different losses" with these two simpler ones: "whose least Preferred result is winning. In contrast, the other candidates' least favourable results are slightly different losses."

I still think it needs some work, but I ask that you explain why any of my grammar corrections should be reverted. Thanks. Myclob (talk) 20:15, 30 December 2021 (UTC)

I was a bit harsh in saying that I wasn’t willing to spend the time checking for errors. Having closely compared a few paras, the main difference I can see is a persistent breaking of sentences into shorter ones. There’s something very close to an error in the words ‘Llull is believed to have supported Copeland's method that used a sequence of two-way elections rather than ranked-choice ballots’ which has Llull agreeing with what Copeland said 650 years later and tells us that Copeland’s method used a sequence of two-way elections – well that simply is an error.

My sentence length is that of normal writing on a subject like this. Using an automated tool to tell you how long sentences should be has no validity as an editing principle, and it is none of your business to try to impose American spelling. My original was better. Colin.champion (talk) 06:56, 31 December 2021 (UTC)

re: "My sentence length is that of normal writing on a subject like this." What do you base this conclusion on, other than your own preference? Myclob (talk) 19:15, 1 January 2022 (UTC)

re: "Using an automated tool to tell you how long sentences should be has no validity as an editing principle." I am not just using "an automated tool." Have you ever heard of Google? Try searching "recommended maximum sentence length." Do you need me to tell you all the reputable sources that come up? How about this one: https://readabilityguidelines.co.uk/clear-language/simple-sentences/#:~:text=Oxford%20Guide%20to%20plain%20English,are%20hard%20to%20comprehend%20easily. "Oxford Guide to plain English, GOV.UK and linguists agree: 15 word sentences are more likely to be comprehensible. 25 words is a good maximum sentence length limit. Above 40 words sentences are hard to comprehend easily." Maybe you have read this article from the Atlantic: https://www.theatlantic.com/education/archive/2015/10/complex-academic-writing/412255/. "The Needless Complexity of Academic Writing, A new movement strives for simplicity." By Victoria Clayton. I am not naturally a good writer, and I am being humble. I am not attacking your writing personally, but this software that people pay for to help them write better, suggested some changes. Grammarly is Worth $13 Billion. Are you saying you know more than them? Have people given you $13 Billion dollars to fix their writing? https://venturebeat.com/2021/11/17/grammarly-ai-writing-suggestions-startup-raises-200m-at-13b-valuation/. Do you disagree with this: ""Oxford Guide to plain English, GOV.UK and linguists agree: 15 word sentences are more likely to be comprehensible. 25 words is a good maximum sentence length limit. Above 40 words sentences are hard to comprehend easily?"Myclob (talk) 19:15, 1 January 2022 (UTC)

re: "and it is none of your business to try to impose American spelling." You believe using "favourable" six times in two sentences is better? I would like to vote on this. How do I call for a vote? The Wikipedia style guide says to use local dialects, when the article is about a local subject. Are you saying this is a British specific article? Why is "Favourable" "better" than "prefer"? Other than your opinion, why is British English Better than American English? Perhaps you are from Britain? However, that shouldn't be a logical justification. I think there are 225 Million American English users on this planet https://wiki.riteme.site/wiki/American_English. There might be 60 million British English speakers. Is every British person worth 3.75 American English speakers? I would like to see some argument, other than the statement of the conclusion that your version is "better." Myclob (talk) 19:15, 1 January 2022 (UTC)

re: "My original was better." You don't have to take things so personally. This is not "your" website. This is written by the crowd, and you really should apologize of undoing all of my edits, without discussing them. Even when you explicitly went against the Wikipedia manual of style, you undid my changes. Myclob (talk) 19:15, 1 January 2022 (UTC)

But my replies were personally inconsiderate towards @Myclob:, for which I apologise. Colin.champion (talk) 11:32, 1 January 2022 (UTC)

[Unindent] @Myclob: raises some fair points.

On sentence length (and vocabulary) there is a guideline Wikipedia:Writing_better_articles#Tone here

Encyclopedic writing has a fairly academic approach, while remaining clear and understandable.

All I can say in addition is that I’m fairly familiar with academic writing, and that where I’ve fallen down on clarity and intelligibility you’re right to point it out. The references you cite do not refer to technical communication.

On national varieties Wikipedia is quite explicit (see Wikipedia:Manual_of_Style#Retaining_the_existing_variety):

With few exceptions (e.g., when a topic has strong national ties or the change reduces ambiguity), there is no valid reason for changing from one acceptable option to another.

This is surely far too strict, but the aim is to avoid unconstructive edit-warring. For my part I have no strong preference for British spelling (I’d be hard-pushed to justify ‘enter/centre’ over ‘enter/center’) but it’s what I’m used to. I’m happy to adapt to American spelling where it’s established. If one day someone else takes over the main burden of maintaining this page, I’d be happy for them to write in the way they find natural. Colin.champion (talk) 10:26, 2 January 2022 (UTC)

I agree. My goal wasn't to change it from British to American English. There were just the two sentences that used a British English word. Grammarly said that it was a monotonous tone. I can find other things that say not to use the same synonym over and and over, besides Grammarly. I was trying to justify the use of other words, to tone down on the British approach. To be honest, I did change a couple of the British spellings to English, before I stopped. I would be 100% fine if you wanted to change them back to the British spelling. Myclob (talk)

... and looking over the above, I see that my sentence

All I can say in addition is that I’m fairly familiar with academic writing, and that where I’ve fallen down on clarity and intelligibility you’re right to point it out.

is 20% longer than the suggested maximum of 25 words while being both short and simple by the standards of academic prose. Colin.champion (talk) 10:34, 2 January 2022 (UTC)

I'm not sure I totally understand, but I appreciate your tone. Also, I can tell you are experienced with academic writing. I have been out of college for a number of years, and when I went to college, I was doing engineering and not anything related to this subject. I can tell you are an expert in it, and I am very grateful for what you have written. Like I said, my wife is a technical writer. They often tell her she needs to explain things like she is explaining things to her mom. She worked in computer manufacturing, and knew nothing about computers. They told her that was prefect. I don't want to change the content. I just want to make sure I understand it. When I read the article, there is still a lot that I don't understand. My only contribution can be that I can read the article as a true outsider, and let you know when it is hard to understand. Thanks again! Myclob (talk) 16:16, 2 January 2022 (UTC)

A supporting reference:

Conversational sentences tend to be short. In fiction, sentences are on average about 20 words long. In academic writing, they are about 40 words long.

Colin.champion (talk) 11:43, 2 January 2022 (UTC)

I do agree that historically academic writing has been harder to understand, and has not been written with the general public in mind. I do think we have a different audience than pure academic writting, and this article talks about a trend towards simplicity in acedimic writing: "The Needless Complexity of Academic Writing, A new movement strives for simplicity.

@Myclob: – thanks for your reply, I think we’re getting closer to each other.

While I wouldn’t accept software tools as a yardstick, they can be valuable for detecting bad writing, and you seem to have caught me out with 6 ‘favourable’s.

I actually agree 100% with Victoria Clayton’s article, but so (I suspect) does Wikipedia in its heart of hearts, which is why it says “fairly academic”. There used to be a learned writing style practised chiefly outside universities which used long words and sentences but aimed for clarity, elegance, and precision; the learned style has been eaten whole by academia, which often cares more about defending its output against peer reviewers. Macaulay, an early contributor to the Encyclopedia Britannica, is a good example of a non-academic learned author (not that anyone would write like that today).

Even when articles are written in the style of academics whose style is learned rather than academic (Milton Friedman being an example), they will still impose a barrier to many readers. If Wikipedia is often hard reading for native English speakers, then imagine what it must be like for readers whose native language is Czech or Cantonese – yet English Wikipedia is effectively global and has to answer their needs. Presumably Simple English Wikipedia is intended to bridge the gap, but it doesn’t yet have enough critical mass to save readers from dependence on the main version. Colin.champion (talk) 17:31, 3 January 2022 (UTC)

Thanks for understanding my approach. I take pride in my contributions, and all I did was a very small spelling and grammar check. I can't imagine all the time you put into it, and it is very big of you to be open to some edits, and suggestions. Cheers! Myclob (talk) 04:40, 9 January 2022 (UTC)

Wiki Education Foundation-supported course assignment

This article was the subject of a Wiki Education Foundation-supported course assignment, between 26 August 2020 and 2 December 2020. Further details are available on the course page. Student editor(s): Komalbadesha. Peer reviewers: Bryankjh, VillusionV.

Above undated message substituted from Template:Dashboard.wikiedu.org assignment by PrimeBOT (talk) 07:48, 17 January 2022 (UTC)

re: "Ranked voting elicits more information about voter preferences than is revealed through an first-past-the-post ballot,"

Do we have any evidence of this? In most modern societies we have anonymous voting, and I am unaware of any example of voter preferences being harvested or used. This seems like a scare tactic that doesn't really hold up to analysis.

Can I suggest a change, and let other make it if they agree? Wouldn't it be better to say "Ranked voting could be used to gather more specific information about voter preferences than is revealed through an first-past-the-post voting systems." myclob (talk) 23:56, 24 December 2021 (UTC)

I didn't write this material but I think you are misunderstanding it. The idea is not that by observing anonymized ballots we can associate the ballot with an individual and make inferences about that specific individual's preferences. The idea is that you see how many voters had each ranking of the candidates, and not just how many had each candidate in first place. If there are 3 candidates in a single-vote election, you only know how many people voted for each candidate. But if they are competing in a ranked choice election, then after the ballots are counted you will know how many people ranked each candidate first, how many ranked each candidate second, and how many ranked each candidate third. So you have more information about peoples' preferences (or their strategic choices) than you do in a single-vote election. It's absolutely true (and sourcing won't be hard to find on this point) that you are collecting more information from voters in a ranked choice election than in a single-vote election. - Astrophobe (talk) 00:44, 25 December 2021 (UTC)

It seems totally logical to me without "could be" wording.

Key point: It eliminated strategic voting: People don't have to vote for the lesser of evils, fearing that their vote would be wasted voting for Ralph Nader, to name only one example: We don't know (1) how many people who voted for Gore or Bush in the 2000 United States presidential election would have voted for Nader with ranked voting, nor (2) how many Nader voters in that election might not have voted at all if he had not been on the ballot. Ranked voting would answer the first question and make the second one moot. ??? DavidMCEddy (talk) 00:44, 25 December 2021 (UTC)

I'm impressed and confused that we somehow avoided an edit conflict! I agree with nearly all of that, but I want to note (for the sake of making sure that it doesn't feed a consensus that might affect the contents of the article) that ranked choice voting does not eliminate strategic voting. There are lots of sources on types of strategic voting in ranked choice rulesets. - Astrophobe (talk)

How does it not eliminate strategic voting?

Should we add a section on something like "Strategic voting when ranking choices", citing a few of the "lots of sources" you say exist on this? DavidMCEddy (talk) 03:11, 25 December 2021 (UTC)

Ranked voting doesn't eliminate tactical voting, e.g., Arrow's impossibility theorem. I would support the addition of such a section, yes. - Astrophobe (talk) 04:04, 25 December 2021 (UTC)

Thanks for the search in Google Scholar and the link to the Wikipedia article on Tactical voting. I don't see how Arrow's impossibility theorem is relevant here: It says "that when voters have three or more distinct alternatives (options), no ranked voting electoral system can convert the ranked preferences of individuals into" community ranks that meet reasonable criteria. I haven't studied this enough to know, but it seems like the Median voter theorem would imply that violations of the Condorcet criterion are probably quite rare. If so, that would imply that Arrow's impossibility theorem is a theoretical curiosity but largely irrelevant in reality.

As of 2021-12-25 the Wikipedia article in the Gibbard–Satterthwaite theorem includes an example of strategic voting impacting the results of ranked voting. That should also be mentioned. However, I would naively expect that the Median voter theorem logic would imply that strategic voting would on average have a much smaller impact on any ranked voting scheme than it would on traditional first past the post rules.

Whether I'm correct or not about this, it would be good to have a discussion of this in the present article, I think. DavidMCEddy (talk) 07:14, 25 December 2021 (UTC)

I agree that there's very good reason to believe (and that the balance of sources will support) that the practical opportunities for strategic voting are less than they are in single-vote plurality voting. But there is guaranteed to be some combination of expected vote totals and preference ordering in which the highest expected utility ballot is not a sincere ballot. Worse, there are also pathologies of some ranked choice systems that make totally different types of strategic voting possible. The fact that instant runoff voting fails the monotonicity criterion, for example, means that sometimes you should lower where you rank a candidate in your ballot if you want to make them more likely to win. That's a sort of strategic voting that exists in (that particular type of) ranked choice voting which does not exist in single-vote plurality. When I have a bit more time I'll take a crack at adding some of this to the article, or if you move over some discussion first I'll gladly build on it. I actually just defended my phd dissertation on voting systems 8 days ago so the sourcing won't take much extra work! It's just about writing about it in a comprehensible/NPOV way. :) - Astrophobe (talk) 21:49, 25 December 2021 (UTC)

You "just defended" your PhD dissertation 8 days ago? I trust you passed? Congratulations. This article can use your expertise. DavidMCEddy (talk) 22:46, 25 December 2021 (UTC)

I changed it to: "Ranked voting allows voter to select candidates that more accurately represent their preferences than first-past-the-post ballots. However, they are more complicated [15]: §8.1   and the counting procedure, if done manually, is more complicated and slower." I think it is more straightforward, but feel free to change it if you don't like it. Thanks! I have read through the article 3 or 4 times today, and don't think I can contribute anything else. myclob (talk) 18:15, 26 December 2021 (UTC)

My own method of counting STV (which uses an incremental Bucklin count, and when reduced to single-winner is independent of covered alternatives…which is interesting) has amusing pathologies. Despite being ICA as a single-winner system (I still hate runoffs; Ranked Pairs has superior properties anyway), I'm able to contrive sets of ballots where a quota coalition can have one ballot with X above Y and the rest with Y above X and Y is eliminated, X is elected (it's not easy but I can). They're all broken, in some way, as you note (the pathology I describe in my own STV system is its failure of IIA, so I didn't break Arrow's theorem by a long shot).

What's fascinating about STV is it's immune to strategy in practice. To implement strategy against STV, you need to know the content of all ballots (i.e. a lot of information); this happens by…counting the ballots after they're cast. To know what changes to made (add, drop, alter), uh…Traveling salesman problem. It's NP-Complete. Seriously, good luck with that. (Amusingly, my own STV method has a pathology…but successfully using it strategically seems to be equivalent to the strategic approach of not ranking Y at all, i.e. you're trying to get Y far enough down your ballot that Y is judged to essentially not be on your ballot at all at some point.) John Moser (talk) 22:50, 7 August 2022 (UTC)

It's not just about ranking first, second, third, either. Consider if 1/3 ranks X>Y>Z, 1/3 ranks Y>Z>X, and 1/3 ranks Z>X>Y. Well, X has a 2/3 majority over Y, Y has a 2/3 majority over Z, and Z has a 2/3 majority over X. That's incoherent: X>Y>Z>X what?

Here's the thing: individual preferences are transitive; group preferences are not necessarily transitive, and thus are irrational.

Because individual preferences are transitive, we not only know you ranked X first, Y second, and Z third; but we know the pairs. If X>Y and Y>Z, then X>Z, for an individual voter (only in a single-dimensional, single-peaked preference model is the group's decision necessarily also transitive; but in any number of dimensions, the individual's preferences are transitive).

That means in an election between, say, 5 candidates, we have information about 10 different 2-candidate elections. Between 10 candidates, we have information about 45 elections.

Voters may not rank all candidates. That restricts our information; however, if candidate C is the Condorcet winner and we are looking for the Condorcet winner, then as long as a sufficient number of voters rank as far as to C on their ballot, C will win. That means if all voters rank up to C, C wins. C also wins given any ballots wherein any combination of voters rank any combination of distances further than C. We may be missing some information, but we have enough in these cases to determine who the winner would be if we had that information.

Voters may also have variable information. You know your first 2 or 3 candidates well by policy and otherwise; you have a general understanding of the next few; the remainder you vaguely understand as more "moderate" or "extreme" or however you're scaling them. For example, maybe you like Bernie. You know the difference between Bernie, Warren, and Castro; you kind of understand Harris and Biden; you know nothing about Kaische, Rubio, Cruz, and Trump except that Kaische seems the most moderate (something something he supports abortion rights), Cruz is crazy, and Trump is a no-go, so the bottom of your ballot is Kaische>Rubio>Cruz>Trump. Close enough, and not necessarily any worse than if you were fully informed (your top choices matter a lot more than your bottom choices, generally, in practice with large and diverse sets of candidates).

There are limits of course. STV may be able to reliably select 5 or 7 or 9 winners with a sizable number of candidates (20 or 30 or even 60) given just your top 4 or 6 preferences (voting power is eroded from a ballot each time it donates some portion of a vote to a winning candidate, reducing its influence rapidly with high probability if the top candidates are popular and concentrated among an ideological grouping, i.e. top favorites of a group of similar-ideal voters); whereas a Smith-efficient method such as Ranked Pairs, given 60 candidates, needs you to at least rank on average over 30 (if your first choice is the Condorcet winner, then that's all the information we need from you if all other ballots are sufficiently complete to elect the Condorcet winner; the further you get, the more rankings we need from you). That gets into debates over what kind of primary election cycles we should have, if we need primaries at all, and so forth. (I'm sure you can guess what I've suggested as to that.)

First-past-the-post single-mark, on the other hand, uses a ranked ballot truncated at rank 1, with no equal given rankings (all others are ranked 2). It discards most of that information. Multiple non-transferable vote similarly squashes the top few ranks into rank 1 and everyone else into rank 2 (although it's considered a scored system, as is approval, which can be described in the same manner). You can see where information is lost I'm sure. John Moser (talk) 23:05, 7 August 2022 (UTC)

Disambiguation

Why doesn't this have a disambig page? --Simetrical 02:44, 19 Jan 2005 (UTC)

related pages

Should this be merged with the article on Score Voting (https://wiki.riteme.site/wiki/Score_voting)? There's actually a few articles related to alternative voting methods focused on related concepts, and most of them are borderline-stub articles. 137.118.200.159 (talk) 18:32, 18 January 2023 (UTC)

Almost certainly not. It might make sense to merge the "Score voting" article with the "Cardinal voting" article (since Score and Cardinal voting use the same ballot type), and it might make sense to merge the "Ranked voting" and "Preferential voting" articles (since Ranked and Preferential voting use the same ballot type). -- RobLa (talk) 22:50, 19 January 2023 (UTC)

Comparison and Evaluation of Ranked Voting Systems

This is a big change, and so I want to discuss it before I make it: Myclob (talk) 17:58, 21 May 2023 (UTC)

Analysis Using Examples

Michael Gallagher points out that the strengths and weaknesses of different ranked voting methods can be best elucidated through specific examples. A practical example can be provided by considering a voter's ranking of candidates: "If A is eliminated, the voter's vote is transferred to B. If B is then eliminated, the voter's vote is transferred to C. This process continues until one candidate has a majority of the votes," explains Steven Brams and Peter Fishburn.

Applying Logical Voting Criteria

Logical voting criteria play a crucial role in evaluating voting methods. The Condorcet criterion, as defined by Duncan Black, states that "the candidate who would win in a head-to-head race against all other candidates should be the winner of the election." Moreover, the independence of irrelevant alternatives (IIA) criterion, explained by Kenneth Arrow, implies that "the ranking of candidates should not change if the order of the candidates is changed."

Empirical Approach: Simulated Elections

Steven Brams and Peter Fishburn highlight the utility of simulated elections as a means to study the properties of ranked voting methods. These simulations can estimate the 'efficacy' of each voting method, which is defined by how frequently it elects the candidate closest to the center of the voter distribution. In terms of these simulations, Andrew Reynolds indicates that "Condorcet methods and Coombs' method tend to deliver optimal outcomes, followed by the Borda count. Instant-runoff voting comes further behind, with first-past-the-post voting trailing." Myclob (talk) 17:58, 21 May 2023 (UTC)

References

Regarding the section of “Other methods“

May I please receive the green light to fix the first point there, considering that it refers to ranked voting as such an other method, despite the whole article being about ranked voting in the first place, while it's description is also incoherent and seems to mix things up with elements of yet another voting method?

For now I am thinking that the point there was meant to be about score voting instead. Wulfstrex (talk) 08:20, 11 July 2023 (UTC)