This article is rated C-class on Wikipedia's
content assessment scale. It is of interest to the following WikiProjects: | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
As I read the green papers, California is winner take all. Will the unpledged delegates also go the the winner? -- Charles Edwin Shipp ( talk) 05:02, 20 March 2016 (UTC)
From the Green Papers: http://www.thegreenpapers.com/P16/CA-R
This article says/reports that Democrat delegates (especially Super Delegates) are not 'bound' and could jump from Hillary Diane Rodham Clinton to Joe Biden, John Kerry, or even Bernie Sanders.
Linked from the DrudgeReport.com "What if Hillary is forced out?" http://thehill.com/blogs/ballot-box/presidential-races/273626-the-chaos-scenario-for-democrats
Charles Edwin Shipp ( talk) 13:13, 20 March 2016 (UTC)
I see that there is some conflict concerning which map is displayed in the infobox. I’ll add my reasoning for reverting to the old map.
The overall popular vote as listed above whichever map is used should be enough to show who the winner of a given election (or in this case, just the state) is. The pie chart does give the margin, but when the populations of the counties are factored in, it becomes nearly impossible to understand what is happening in the smaller counties. There are at least 13 counties that are impossible to read (even when I look at the file on Commons) and there are 9 more that are difficult to read. The map is rather cluttered overall, with the terrain (forests?), county names, pie charts, and the surrounding ocean/land that isn’t part of California.
The original map is easy to read with the key supplied in the info box. I think that it’s assumed that not all counties have the same population.
I like the idea of showing both the total votes and the margin by which the given candidate won, but it definitely needs some refinement. Perhaps an improved version could be put in the section with the table county results (speaking of the table, it and the rest of the numbers on this page will have to be updated once finalized results with write-ins are posted by the Sec. of State, but that’s for another time). Teak the Kiwi ( talk) 05:03, 14 December 2016 (UTC)
What information are you giving the reader when you present them with a graphic that tells them Trump won 50-60% of the vote in Inyo County? What does that mean? 50-60% of what? Is it compraable to teh 50-60% that Clinton won in neighboring San Bernadion County? Without the population, you're going to great effort to convey utterly meaningless information. The number of counties "won" has no bearing on anything. Counties are not won, only states. The land area of the counties is meaningless. What percentage of a county a candidate got means nothing without the number of votes. Your map means literally nothing. Zero. Might as well delete it altogether.
The reason the smaller counties are hard to understand on the pie map is that there is no physical way to display both to scale without LA county filling the entire display. It's like trying to make a scale model of the solar system with both the Sun and Pluto side by side, and Pluto visible.
What exactly is of so much interest in Alpine county, where 575 votes were cast? Compared to LA county, with over 3,000,000, and a couple more with over 1,0000,0000 and several with more than 500,000? This is a drastic violation of WP:UNDUE to minimize the importance of so many votes in favor of several counties which have a lot of empty land in them.
Inyo County has an area of 10,227 sq mi, which is 6% of California's 163,696 square miles. The total number of votes in Inyo is under 8,000, which is 0.05% of the over 14,000,000 votes cast. So the importance of Inyo is inflated by a factor of more than 100! This is undue weight: giving excessive attention to an objectively small thing. You want to devote 6% of the map to a county that contains 0.05% of the votes. Why?
The pie maps don't eliminate the small counties; they have marks that exist. But the fact that you can't tell much about them is exactly the point. Complaining that the graph doesn't tell you how they voted in a tiny county is like complaining that some tiny neighborhood in Los Angeles doesn't appear on the map. Why would it? -- Dennis Bratland ( talk) 05:43, 14 December 2016 (UTC)
You do make a fair point about Alpine County not having nearly as much weight on the overall election results as Los Angeles County. But I would argue that ignoring regional trends tends towards calling for essentially the complete removal of geography from any election map. County (and state) maps give people an idea of the results in relation to the land. While a cartogram would introduce the least amount of bias, there are many issues with them, especially for foreign readers who may not be as familiar with American geography as Americans are. Cartograms are relatively easy to make for something like the Electoral College. There’s no perfect way to represent results, but the present pie charts hide a lot about smaller regions. I’m not arguing that land determines election results, but there’s a good chunk of information lost in the present situation (the pie chart map). A reader can still tell that the coastal regions lean heavily Democratic, but the results of smaller inland counties are lost. The pie chart map is not very scalable and can be hard to read on small screens (mobile) as well as desktop-sized screens. I still stand by my point that it’s safe to assume that a reader would know that more land shaded for a given candidate does not make that candidate the overall winner.
Perhaps a good compromise map/graphic would show the margins that a candidate won by in a county. Rather than showing the percentage that a candidate got, it would show how many raw votes a candidate beat their nearest competitor. That would keep the county maps relatively easy to read while still giving an idea about how important a given county is to the overall statewide margin. Basically shade the counties in a color that represents how many more votes the county winner got than the runner up. That would still readily show the geographical aspect while also representing the amount of votes. Teak the Kiwi ( talk) 06:45, 14 December 2016 (UTC)
It is very much a matter of undue weight. You're defending a map that places extreme emphasis (over 100 times!) on the point of view of people in desolate areas. There is a significant point of view of people who think an American majority is being dominated by urban elites, based on maps that are covered with wide swaths of red. The New York Times, for example said there is a "common complaint about shaded-area 'choropleth' maps like this: They are misleading because they seem to suggest that the vast majority of America votes Republican." How is that not an issue of point of view?
It is especially misleading to imply that "number of counties won" is a thing. It creates a false impression that the election was closer than it was. In Nevada, the effect is profound: it creates exactly the opposite impression on in every way.
It is not neutral to take a set of data and create a map that distorts what it means, by the arbitrary choice to divide it up by counties, and ignore how many votes were cast in each county, in favor of the meaningless statistic of how much a county was "won" by. That is a biased, and fantastical, point of view. It is a falsehood to say anyone "won" a county. These filled-area maps are terrible in every way imaginable.
Cartograms are great and all, but I invite you to find out how difficult it is to make one on the state-county level. If you succeed, hats off to you, such a map would be somewhat better. Though still ignoring third parties, and so on. When the final numbers are released, you have to struggle through the whole painful cartogram generation process again to update the maps. Pies on a map is by far a better choice, and it's practical, since such maps are not difficult to produce, and can be quickly updated and corrected as needed. -- Dennis Bratland ( talk) 07:39, 14 December 2016 (UTC)
Counties are not just arbitrary lines on maps; they reflect historical patterns and over time may reflect new patterns (i.e. Ski resort dominated counties (such as Alpine and Mono) tend to vote Democratic). Voting patterns can reflect many underlying demographic patterns. This information for smaller counties is lost in the current form of the pie chart map. I'm not sure of what program you're using, but my experience with ArcMap suggests that it shouldn't be too difficult to color the counties by the percentage for the winner as well as including labels and pie charts. The current map appears too cluttered while not clearly showing useful information. Teak the Kiwi ( talk) 08:49, 14 December 2016 (UTC)
"For example, on the Utah pie charts map Clinton has big sections of the larger pies which is very misleading because she also lost the state in a landslide." WTF? Clinton has "big sections" of what pies? The pie for Salt Lake County, for example? The blue piece of that pie comprises 42% of the total for Salt Lake. Clinton received 42% of the vote in Salt Lake. It's exactly correct. The size of the pie for Salt Lake is exactly in proportion to the number of votes cast in that county. How is that misleading? It's true that it's not visually obvious that Trump won the state, because the aggregate visual impression is not dominated by red, but in fact, he only got 45% of the vote, a plurality. Trump won, but did not dominate the state, and that is illustrated accurately. His lack of a majority is reflected visually by showing that other candidates, notably McMullin, took significant chunk. The filled map utterly fails to display this unique fact about the Utah election.
Trump did dominate, with a real majority, states like Idaho and Montana, and that fact jumps out at you looking at the pie maps for those states. You also note that Johnson played a significant, though not decisive, role in those states, where McMullin was a non-factor, as in almost every state. All this rich detail is present in the pie maps, and utterly missing from the filled-area maps.
The claim keeps being repeated that without the filled-area map (the choropleth), "information is lost". What information? Both maps tell you which counties were "won" (though that's not a thing, but both tell you that). Both tell you what proportion the "winner" got of the county got. And that right there is the end of the information contained in the choropleth. You have no idea how well the second place candidate did. Or if any third party candidates were a factor, and by how much. You have no idea how many votes were cast in each county. If you delete the choropleth map, no information is removed. It's true the pie map places great emphasis on the largest counties, since those essentially determine the outcome in every state, while making the micro-counties hard to see. They give the impression that the micro-counties are of little importance. Which is a fact. They have a negligible impact on the outcome. About 0.05% in many cases.
I'm sorry to see this utter disregard for facts. The only "information" that is lost without the choropleth is false information. The appearance that large land-area counties dominate the election in the same way they dominate the map. The choropleth multiplies the proportion of the vote of the county "winner" by the area. It would make as much sense to multiply their vote by the average height of the voters, or the number of miles of roads in the county, or any other arbitrary number. When you multiply the visual impact of a statistic by an utterly arbitrary number, you create a false image. That's the only "information" we "lose" when we take away this naive representation of the election data. -- Dennis Bratland ( talk) 23:30, 14 December 2016 (UTC)
The assertion "The conventional map is also there for people to see counties that may not be able to see the victor or the margins in all counties." is an example of the violation of WP:Undue weight happening here. We don't devote space to realtively unimportant data. By insisting on giving prominence to a few hundred voters that is equal, and even greater, than a few million voters is exactly the kind of imbalanced presentation that WP:UNDUE forbids. The Utah example is instructive, in a way exactly the opposite of what Travis McGeehan claimed. The choropleth gives excess prominence to Clintoin's majority win in Summit county; it gives the appearance that her plurality win in Salt lake was of lesser weight. The pie presents the data correctly: yes, she did win more than half the vote in Summit, 15 points ahead of Trump, but that was out of only 20,000 votes, one 20th the number cast in Salt Lake County. The large red swaths in Mallard and Emry counties are similarly misleading: there were only 4,000, or 5,000 votes cast there. You're touting the fact that the choropleth lets readers see these miro-counties better. Wikipedia has policies that say we don't want to spotlight insignificant information. And we certainly don't want to be emphasizing the "victor" in a county. Candidates don't win counties. Coloring in the whole county as if all that matters is who "won" that county is factually incorrect. Racking up votes in a county, even if it was part of a minority in that county, is still important in the overall statewide contest. Those votes count, because they add to the total that determines who wins the state, the only winning that matters.
The only good thing to be said of sandbagging the infobox with two maps, one of them being misleading and nearly useless, is that it stops edit warring. That's the only good that comes of it. -- Dennis Bratland ( talk) 01:02, 15 December 2016 (UTC)
I will also repeat that WP:UNDUE does not apply. There is not a question of whether or not the information is from a reliable source. This data is (presumably) the same as what is posted on the CA Secretary of State’s website (the info will be slightly outdated when write in tallies are added). Hillary Clinton receiving 50-60% of the vote in Alpine County is not a fringe theory, it is a fact supported by the county clerk’s website as well as the state Secretary of State. The percentages for all counties are in the table. If we are to follow the idea that results for smaller counties don’t matter, should those counties be removed from the table? The choropleth map is not about what people in a region think, it is about how people in a region acted. Displaying only the pie chart map would mask these facts. I really have to disagree with the idea that smaller counties don’t matter. Yes, their results do not have as large of an impact as say Los Angeles, but they are still displayed and reported on by the state Secretary of State.
From WP:BALASP “An article should not give undue weight to minor aspects of its subject, but should strive to treat each aspect with a weight proportional to its treatment in the body of reliable, published material on the subject.”
I’ll use the 2008 Results from the CA Secretary of State as an example because the 2012 results seem to link to the primary results for 2012. These results, labeled as “Statewide Summary” under the “President” heading are listed by county. All counties are displayed equally, as they are equal in the eyes of the SoS. The fact that some counties have far lower populations does not affect how they are reported by the SoS. For this year’s results, the SoS is still showing this map on their website, showing that the SoS, the definitive source for California state election results, discusses the results by county in equal amounts no matter how large the county’s population is.
The choropleth map shows which candidates did best in a given county; this information can help readers understand long term trends. For example, Lassen County used to be a Democratic stronghold. Jerry Brown, CA’s incumbent governor, won both counties during both of his gubernatorial elections in the 1970’s; now, Lassen County is one of the worst counties for Democrats; it was one of two counties that Brown lost in the blanket primary in 2014. Without the choropleth map, it would take a careful reading of the full results chart to see such a change. The pie chart map completely removes this aspect. While it does bring some new information to the table, it is rather cluttered and lacks the information provided by the choropleth.
I’m really struggling to understand why smaller counties shouldn’t matter. Are you suggesting that all choropleth maps involving counties should be eliminated? Just because a county is smaller doesn’t mean that it doesn’t matter. Many politically oriented websites use choropleths to display election results, from uselectionatlas.org (which is a fun website if you have time to kill), to the New York Times (virtually every news site reporting election results uses choropleths or at least has them as an option), and even the California Secretary of State. The choropleth map that is in dispute on this page provides more information than at least the latter two websites due to the varying shades (as opposed to simply red or blue).
I would like to close this post by asking that all participants keep WP:CIVIL in mind in all parts of their posts including the edit summaries of the talk pages. Teak the Kiwi ( talk) 06:31, 15 December 2016 (UTC)
In the closing discussion, there is a discussion involving "ignore the rules" to reach a rough-but not perfect consensus. On the WA discussion page, such discussion such as multiple articles describing why pie charts are not necessarily the best method of representing statistics either. This should roughly cancel out-for lack of a better term, the articles presented in your portion of the discussion page. PalmerTheGolfer ( talk) 01:38, 15 December 2016 (UTC)
I changed the sentence to read that California's contribution was decisive in winning Clinton the national popular vote. If you ignore the California totals? Trump wins the national popular vote by 1.5 million. GoodDay ( talk) 04:54, 23 December 2016 (UTC)
It seems like original research, and it represents a POV, when you start spinning scenarios: "If you take away this state, here's what would have happened!" Do we take away the 11 states of the Confederacy and talk about what the outcome would have been in that alternative universe? To we construct scenarios group counties by their contribution to the GNP? These types of artificial "what-if" constructs are the province of pundits and analysis. They can be included in articles, but they must be attributed to a reasonably well-known or respected analyst.
Lacking a source, as well as text outside the footnote attributing by name, the commentary like "contributed decisively in Clinton's winning of the national popular vote despite losing the electoral college" should be removed. Add attribution, and then move it down to an ==Analysis== section for appropriate context and alternate interpretations. -- Dennis Bratland ( talk) 05:15, 23 December 2016 (UTC)
Once again, sentences like "California has voted Democratic in every presidential election after Republican George H.W. Bush won the state in 1988" or "Hillary Clinton carried California by the largest margin of any Democratic candidate since Franklin D. Roosevelt swept the state by 35.25% in his 1936 re-election landslide" are not just "basic math" as @ Master of Time: has asserted. We have examples of this kind of cherry-picked factoid on many articles and it needs to stop. Basic math doesn't tell you to that California's history of voting Democratic is the fact you have to lead with. Basic math doesn't tell you that Clinton's winning margin compared to other Democrats is that important. Basic math says yes, it's true that FDR's margin was 35.25%, and Clinton's is the largest since then.
The problem is that basic math can also support an infinite number of factoids. "First Democrat who has seen Time Bandits more than twice (in a theater not home video) to win California by more than 15 points!" or "Only left handed Republican whose mother is from New Jersey to win Orange County!" Why are these superlatives important? You can sit here all day and make them up.
The thing is, it's easy in most cases to find more than one credible, independent expert who has said "California has voted Democratic in every presidential election after Republican George H.W. Bush won the state in 1988." Cite the name of experts who said it, and now it's not original research or synthesis any more. Then again, if you search and search and you really can't find one single reliable source who actually said something like, "Clinton carried California by the largest margin of any Democratic candidate since Franklin D. Roosevelt swept the state by 35.25% in his 1936 re-election", then maybe it's about as significant as left handed Republicans whose mother is from New Jersey.
The beauty of attribution is that some random anonymous Wikipedia editor is not in charge of deciding if it matters if you're the first left handed Republican who did this or did that. Nobody needs debate it. If the sources highlight it, then we know it matters. It's easy. -- Dennis Bratland ( talk) 20:35, 31 March 2017 (UTC)
I'm confused by the fact that some of the write-ins were official nominees of parties and yet are listed as independents in the Statewide Results box near the top (specifically, Mike Maturen - American Solidarity Party & Jerry White - Socialist Equality Party). I'm guessing this is some kind of technicality...if parties don't make the ballot, their nominees aren't officially recognized as being affiliated with the party. I see the same pattern for 2012 results, so I assume it's intentional. I think readers would be interested to know what write-in candidates were running as nominees of parties, if there are any established criteria for what a party is (I understand that someone might create a "party of 1" for their write-in candidacy that is basically a party in name only).
EDIT: I see that basically a political party exists in the eyes of the state when they have enough signatures or registered voters to get on the ballot, and there is probably no other form of official recognition. It might still be useful to identify parties that independents associate themselves with, even if they are parties in name only. If a candidate is officially recognized as a write-in and gets enough votes to make it on the wikipedia page, I'd think readers would be interested to know under what party banner (if any) that person was running. It could be an additional column on the table at the far right, or a footnote paragraph below the box.
EDIT: California's Secretary of State Office made one major error (although inconsequential in the present instance). There were separate slates of electors for the Republican Party candidates (Trump and Pence) and for the American Independent Party candidates (Trump and Pence), but no way for voters to distinguish these electors on California ballots.
California Republican Party Electors Pledged to Donald J. Trump for President of the United States and Michael R. Pence for Vice President of the United States:
Joel Anderson, Marilyn Barke, Jennifer Beall, Robert Bernosky, Arun Bhumitra, Jim Brulte, Nachhattar Chandi, Claire Chiara, Tim Clark, Greg Conlon, Matthew Del Carlo, Harmeet Dhillon, Elizabeth Emken, Jean Fuller, Ted Gaines, Ron Gold, Lisa Grace-Kellogg, Barbara Grimm Marshall, Howard Hakes, Diane Harkey, Matthew Harmon, Noel Irwin Hentschel, Mark Herrick, Tom Hudson, Kenneth Korbin, Kevin Krick, Linda Lopez-Alvarez, Robin Lowe, Papa Doug Manchester, Shirley Mark, Chuck McDougald, John Musella, Ron Nehring, Mike Osborn, Douglas Ose, John Peck, Pete Petrovich, Donna Porter, Dennis Revell, Scott Robertson, Carla Sands, Truong Si, Robert Smittcamp, Mike Spence, Jeff Lalloway, Shawn Steel, Mark Vafiades, Marcelino Valdez, Errol Valladares, Cyndi Vanderhorst, Megan Vincent, Elissa Wadleigh, Deborah Wilder, Dave Willmon, John Young
American Independent Party Electors Pledged to Donald J. Trump for President of the United States and Michael R. Pence for Vice President of the United States:
Linda Lea Alsbury, Merwyn Alsbury, The Honorable Steve Baldwin, Gary Brown, Ruth Brown, Mark Brownlee, William C. Cardoza, Joseph J. Cocchi, Julie Colglazier, Kayla Colglazier, Patrick Colglazier, Dr. J. Steven Davis, Sallie Hansen Doman, The Honorable Robert K. Doman, Wiley Drake, Sally S. Easter, The Honorable Virgil Goode, Ron Gold, Jeff Grage, Charles Edward Harrison, Jr., Thomas Nowlen Hudson, The Honorable John LeBoutillier, The Honorable Robert Marc Levy, Mary Parker Lewis, Gaudencio Gene Lopez, Judy Lopez, Raul Lopez, Sheila Schultz Lopez, Leonard Luna, Kim McDermott, Eric McDermott, Arthur Loyal Morgan, Matthew Justin Morgan, Richard Matthew Nettleton, Sr., Julie Marie Nettleton, Marc Nettleton, Jaycob Andrew Ornelas, Melissa Ornelas, Robert Ornelas, Marilyn Plumb, Jamie Rangel, Jeffrey Rangel, John Daniel Robertson, Markham Robinson, Mary Robinson, Stephanie Roundy, Terrance Arthur Rust, Dustin Paul Salsi, Richard Scott, Andrew Schalo, David James Scholl, Mark J. Seidenberg, Chris Smentech, Glenn Smentech, Michael Warnken, Jack Warren
Both slates of electors are listed in the Certificate of Ascertainment has each having received 4,483,810 votes.
Fortunately, the Democratic Party electors received 8,753,788 votes and therefore won. However, if the Trump-Pence ticket had won in California, the Secretary of State would have been extremely embarrassed when both sets of electors showed up in Sacramento on December 19, 2016 to cast their ballots for President and Vice President, each claiming to have been duly elected as the state's electors. — Preceding unsigned comment added by 173.228.6.44 ( talk) 19:19, 21 March 2017 (UTC)
If the Certificate is "NOT a mere formality", then can you explain why this supposed error made no difference? Seems like the definition of "formality" if there can be an error like this yet it has no effect. -- Dennis Bratland ( talk) 18:27, 31 March 2017 (UTC)
Very well! You and I disagree on the interpretation of what constitutes original research and the proper way to cite primary references (such as public documents and statutory law sections) to meet wikipedia requirements, but since we can agree on at least certain minimal content, we'll leave the text per your reduction. — Preceding unsigned comment added by 173.228.6.44 ( talk) 19:37, 3 April 2017 (UTC)
This article is rated C-class on Wikipedia's
content assessment scale. It is of interest to the following WikiProjects: | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
As I read the green papers, California is winner take all. Will the unpledged delegates also go the the winner? -- Charles Edwin Shipp ( talk) 05:02, 20 March 2016 (UTC)
From the Green Papers: http://www.thegreenpapers.com/P16/CA-R
This article says/reports that Democrat delegates (especially Super Delegates) are not 'bound' and could jump from Hillary Diane Rodham Clinton to Joe Biden, John Kerry, or even Bernie Sanders.
Linked from the DrudgeReport.com "What if Hillary is forced out?" http://thehill.com/blogs/ballot-box/presidential-races/273626-the-chaos-scenario-for-democrats
Charles Edwin Shipp ( talk) 13:13, 20 March 2016 (UTC)
I see that there is some conflict concerning which map is displayed in the infobox. I’ll add my reasoning for reverting to the old map.
The overall popular vote as listed above whichever map is used should be enough to show who the winner of a given election (or in this case, just the state) is. The pie chart does give the margin, but when the populations of the counties are factored in, it becomes nearly impossible to understand what is happening in the smaller counties. There are at least 13 counties that are impossible to read (even when I look at the file on Commons) and there are 9 more that are difficult to read. The map is rather cluttered overall, with the terrain (forests?), county names, pie charts, and the surrounding ocean/land that isn’t part of California.
The original map is easy to read with the key supplied in the info box. I think that it’s assumed that not all counties have the same population.
I like the idea of showing both the total votes and the margin by which the given candidate won, but it definitely needs some refinement. Perhaps an improved version could be put in the section with the table county results (speaking of the table, it and the rest of the numbers on this page will have to be updated once finalized results with write-ins are posted by the Sec. of State, but that’s for another time). Teak the Kiwi ( talk) 05:03, 14 December 2016 (UTC)
What information are you giving the reader when you present them with a graphic that tells them Trump won 50-60% of the vote in Inyo County? What does that mean? 50-60% of what? Is it compraable to teh 50-60% that Clinton won in neighboring San Bernadion County? Without the population, you're going to great effort to convey utterly meaningless information. The number of counties "won" has no bearing on anything. Counties are not won, only states. The land area of the counties is meaningless. What percentage of a county a candidate got means nothing without the number of votes. Your map means literally nothing. Zero. Might as well delete it altogether.
The reason the smaller counties are hard to understand on the pie map is that there is no physical way to display both to scale without LA county filling the entire display. It's like trying to make a scale model of the solar system with both the Sun and Pluto side by side, and Pluto visible.
What exactly is of so much interest in Alpine county, where 575 votes were cast? Compared to LA county, with over 3,000,000, and a couple more with over 1,0000,0000 and several with more than 500,000? This is a drastic violation of WP:UNDUE to minimize the importance of so many votes in favor of several counties which have a lot of empty land in them.
Inyo County has an area of 10,227 sq mi, which is 6% of California's 163,696 square miles. The total number of votes in Inyo is under 8,000, which is 0.05% of the over 14,000,000 votes cast. So the importance of Inyo is inflated by a factor of more than 100! This is undue weight: giving excessive attention to an objectively small thing. You want to devote 6% of the map to a county that contains 0.05% of the votes. Why?
The pie maps don't eliminate the small counties; they have marks that exist. But the fact that you can't tell much about them is exactly the point. Complaining that the graph doesn't tell you how they voted in a tiny county is like complaining that some tiny neighborhood in Los Angeles doesn't appear on the map. Why would it? -- Dennis Bratland ( talk) 05:43, 14 December 2016 (UTC)
You do make a fair point about Alpine County not having nearly as much weight on the overall election results as Los Angeles County. But I would argue that ignoring regional trends tends towards calling for essentially the complete removal of geography from any election map. County (and state) maps give people an idea of the results in relation to the land. While a cartogram would introduce the least amount of bias, there are many issues with them, especially for foreign readers who may not be as familiar with American geography as Americans are. Cartograms are relatively easy to make for something like the Electoral College. There’s no perfect way to represent results, but the present pie charts hide a lot about smaller regions. I’m not arguing that land determines election results, but there’s a good chunk of information lost in the present situation (the pie chart map). A reader can still tell that the coastal regions lean heavily Democratic, but the results of smaller inland counties are lost. The pie chart map is not very scalable and can be hard to read on small screens (mobile) as well as desktop-sized screens. I still stand by my point that it’s safe to assume that a reader would know that more land shaded for a given candidate does not make that candidate the overall winner.
Perhaps a good compromise map/graphic would show the margins that a candidate won by in a county. Rather than showing the percentage that a candidate got, it would show how many raw votes a candidate beat their nearest competitor. That would keep the county maps relatively easy to read while still giving an idea about how important a given county is to the overall statewide margin. Basically shade the counties in a color that represents how many more votes the county winner got than the runner up. That would still readily show the geographical aspect while also representing the amount of votes. Teak the Kiwi ( talk) 06:45, 14 December 2016 (UTC)
It is very much a matter of undue weight. You're defending a map that places extreme emphasis (over 100 times!) on the point of view of people in desolate areas. There is a significant point of view of people who think an American majority is being dominated by urban elites, based on maps that are covered with wide swaths of red. The New York Times, for example said there is a "common complaint about shaded-area 'choropleth' maps like this: They are misleading because they seem to suggest that the vast majority of America votes Republican." How is that not an issue of point of view?
It is especially misleading to imply that "number of counties won" is a thing. It creates a false impression that the election was closer than it was. In Nevada, the effect is profound: it creates exactly the opposite impression on in every way.
It is not neutral to take a set of data and create a map that distorts what it means, by the arbitrary choice to divide it up by counties, and ignore how many votes were cast in each county, in favor of the meaningless statistic of how much a county was "won" by. That is a biased, and fantastical, point of view. It is a falsehood to say anyone "won" a county. These filled-area maps are terrible in every way imaginable.
Cartograms are great and all, but I invite you to find out how difficult it is to make one on the state-county level. If you succeed, hats off to you, such a map would be somewhat better. Though still ignoring third parties, and so on. When the final numbers are released, you have to struggle through the whole painful cartogram generation process again to update the maps. Pies on a map is by far a better choice, and it's practical, since such maps are not difficult to produce, and can be quickly updated and corrected as needed. -- Dennis Bratland ( talk) 07:39, 14 December 2016 (UTC)
Counties are not just arbitrary lines on maps; they reflect historical patterns and over time may reflect new patterns (i.e. Ski resort dominated counties (such as Alpine and Mono) tend to vote Democratic). Voting patterns can reflect many underlying demographic patterns. This information for smaller counties is lost in the current form of the pie chart map. I'm not sure of what program you're using, but my experience with ArcMap suggests that it shouldn't be too difficult to color the counties by the percentage for the winner as well as including labels and pie charts. The current map appears too cluttered while not clearly showing useful information. Teak the Kiwi ( talk) 08:49, 14 December 2016 (UTC)
"For example, on the Utah pie charts map Clinton has big sections of the larger pies which is very misleading because she also lost the state in a landslide." WTF? Clinton has "big sections" of what pies? The pie for Salt Lake County, for example? The blue piece of that pie comprises 42% of the total for Salt Lake. Clinton received 42% of the vote in Salt Lake. It's exactly correct. The size of the pie for Salt Lake is exactly in proportion to the number of votes cast in that county. How is that misleading? It's true that it's not visually obvious that Trump won the state, because the aggregate visual impression is not dominated by red, but in fact, he only got 45% of the vote, a plurality. Trump won, but did not dominate the state, and that is illustrated accurately. His lack of a majority is reflected visually by showing that other candidates, notably McMullin, took significant chunk. The filled map utterly fails to display this unique fact about the Utah election.
Trump did dominate, with a real majority, states like Idaho and Montana, and that fact jumps out at you looking at the pie maps for those states. You also note that Johnson played a significant, though not decisive, role in those states, where McMullin was a non-factor, as in almost every state. All this rich detail is present in the pie maps, and utterly missing from the filled-area maps.
The claim keeps being repeated that without the filled-area map (the choropleth), "information is lost". What information? Both maps tell you which counties were "won" (though that's not a thing, but both tell you that). Both tell you what proportion the "winner" got of the county got. And that right there is the end of the information contained in the choropleth. You have no idea how well the second place candidate did. Or if any third party candidates were a factor, and by how much. You have no idea how many votes were cast in each county. If you delete the choropleth map, no information is removed. It's true the pie map places great emphasis on the largest counties, since those essentially determine the outcome in every state, while making the micro-counties hard to see. They give the impression that the micro-counties are of little importance. Which is a fact. They have a negligible impact on the outcome. About 0.05% in many cases.
I'm sorry to see this utter disregard for facts. The only "information" that is lost without the choropleth is false information. The appearance that large land-area counties dominate the election in the same way they dominate the map. The choropleth multiplies the proportion of the vote of the county "winner" by the area. It would make as much sense to multiply their vote by the average height of the voters, or the number of miles of roads in the county, or any other arbitrary number. When you multiply the visual impact of a statistic by an utterly arbitrary number, you create a false image. That's the only "information" we "lose" when we take away this naive representation of the election data. -- Dennis Bratland ( talk) 23:30, 14 December 2016 (UTC)
The assertion "The conventional map is also there for people to see counties that may not be able to see the victor or the margins in all counties." is an example of the violation of WP:Undue weight happening here. We don't devote space to realtively unimportant data. By insisting on giving prominence to a few hundred voters that is equal, and even greater, than a few million voters is exactly the kind of imbalanced presentation that WP:UNDUE forbids. The Utah example is instructive, in a way exactly the opposite of what Travis McGeehan claimed. The choropleth gives excess prominence to Clintoin's majority win in Summit county; it gives the appearance that her plurality win in Salt lake was of lesser weight. The pie presents the data correctly: yes, she did win more than half the vote in Summit, 15 points ahead of Trump, but that was out of only 20,000 votes, one 20th the number cast in Salt Lake County. The large red swaths in Mallard and Emry counties are similarly misleading: there were only 4,000, or 5,000 votes cast there. You're touting the fact that the choropleth lets readers see these miro-counties better. Wikipedia has policies that say we don't want to spotlight insignificant information. And we certainly don't want to be emphasizing the "victor" in a county. Candidates don't win counties. Coloring in the whole county as if all that matters is who "won" that county is factually incorrect. Racking up votes in a county, even if it was part of a minority in that county, is still important in the overall statewide contest. Those votes count, because they add to the total that determines who wins the state, the only winning that matters.
The only good thing to be said of sandbagging the infobox with two maps, one of them being misleading and nearly useless, is that it stops edit warring. That's the only good that comes of it. -- Dennis Bratland ( talk) 01:02, 15 December 2016 (UTC)
I will also repeat that WP:UNDUE does not apply. There is not a question of whether or not the information is from a reliable source. This data is (presumably) the same as what is posted on the CA Secretary of State’s website (the info will be slightly outdated when write in tallies are added). Hillary Clinton receiving 50-60% of the vote in Alpine County is not a fringe theory, it is a fact supported by the county clerk’s website as well as the state Secretary of State. The percentages for all counties are in the table. If we are to follow the idea that results for smaller counties don’t matter, should those counties be removed from the table? The choropleth map is not about what people in a region think, it is about how people in a region acted. Displaying only the pie chart map would mask these facts. I really have to disagree with the idea that smaller counties don’t matter. Yes, their results do not have as large of an impact as say Los Angeles, but they are still displayed and reported on by the state Secretary of State.
From WP:BALASP “An article should not give undue weight to minor aspects of its subject, but should strive to treat each aspect with a weight proportional to its treatment in the body of reliable, published material on the subject.”
I’ll use the 2008 Results from the CA Secretary of State as an example because the 2012 results seem to link to the primary results for 2012. These results, labeled as “Statewide Summary” under the “President” heading are listed by county. All counties are displayed equally, as they are equal in the eyes of the SoS. The fact that some counties have far lower populations does not affect how they are reported by the SoS. For this year’s results, the SoS is still showing this map on their website, showing that the SoS, the definitive source for California state election results, discusses the results by county in equal amounts no matter how large the county’s population is.
The choropleth map shows which candidates did best in a given county; this information can help readers understand long term trends. For example, Lassen County used to be a Democratic stronghold. Jerry Brown, CA’s incumbent governor, won both counties during both of his gubernatorial elections in the 1970’s; now, Lassen County is one of the worst counties for Democrats; it was one of two counties that Brown lost in the blanket primary in 2014. Without the choropleth map, it would take a careful reading of the full results chart to see such a change. The pie chart map completely removes this aspect. While it does bring some new information to the table, it is rather cluttered and lacks the information provided by the choropleth.
I’m really struggling to understand why smaller counties shouldn’t matter. Are you suggesting that all choropleth maps involving counties should be eliminated? Just because a county is smaller doesn’t mean that it doesn’t matter. Many politically oriented websites use choropleths to display election results, from uselectionatlas.org (which is a fun website if you have time to kill), to the New York Times (virtually every news site reporting election results uses choropleths or at least has them as an option), and even the California Secretary of State. The choropleth map that is in dispute on this page provides more information than at least the latter two websites due to the varying shades (as opposed to simply red or blue).
I would like to close this post by asking that all participants keep WP:CIVIL in mind in all parts of their posts including the edit summaries of the talk pages. Teak the Kiwi ( talk) 06:31, 15 December 2016 (UTC)
In the closing discussion, there is a discussion involving "ignore the rules" to reach a rough-but not perfect consensus. On the WA discussion page, such discussion such as multiple articles describing why pie charts are not necessarily the best method of representing statistics either. This should roughly cancel out-for lack of a better term, the articles presented in your portion of the discussion page. PalmerTheGolfer ( talk) 01:38, 15 December 2016 (UTC)
I changed the sentence to read that California's contribution was decisive in winning Clinton the national popular vote. If you ignore the California totals? Trump wins the national popular vote by 1.5 million. GoodDay ( talk) 04:54, 23 December 2016 (UTC)
It seems like original research, and it represents a POV, when you start spinning scenarios: "If you take away this state, here's what would have happened!" Do we take away the 11 states of the Confederacy and talk about what the outcome would have been in that alternative universe? To we construct scenarios group counties by their contribution to the GNP? These types of artificial "what-if" constructs are the province of pundits and analysis. They can be included in articles, but they must be attributed to a reasonably well-known or respected analyst.
Lacking a source, as well as text outside the footnote attributing by name, the commentary like "contributed decisively in Clinton's winning of the national popular vote despite losing the electoral college" should be removed. Add attribution, and then move it down to an ==Analysis== section for appropriate context and alternate interpretations. -- Dennis Bratland ( talk) 05:15, 23 December 2016 (UTC)
Once again, sentences like "California has voted Democratic in every presidential election after Republican George H.W. Bush won the state in 1988" or "Hillary Clinton carried California by the largest margin of any Democratic candidate since Franklin D. Roosevelt swept the state by 35.25% in his 1936 re-election landslide" are not just "basic math" as @ Master of Time: has asserted. We have examples of this kind of cherry-picked factoid on many articles and it needs to stop. Basic math doesn't tell you to that California's history of voting Democratic is the fact you have to lead with. Basic math doesn't tell you that Clinton's winning margin compared to other Democrats is that important. Basic math says yes, it's true that FDR's margin was 35.25%, and Clinton's is the largest since then.
The problem is that basic math can also support an infinite number of factoids. "First Democrat who has seen Time Bandits more than twice (in a theater not home video) to win California by more than 15 points!" or "Only left handed Republican whose mother is from New Jersey to win Orange County!" Why are these superlatives important? You can sit here all day and make them up.
The thing is, it's easy in most cases to find more than one credible, independent expert who has said "California has voted Democratic in every presidential election after Republican George H.W. Bush won the state in 1988." Cite the name of experts who said it, and now it's not original research or synthesis any more. Then again, if you search and search and you really can't find one single reliable source who actually said something like, "Clinton carried California by the largest margin of any Democratic candidate since Franklin D. Roosevelt swept the state by 35.25% in his 1936 re-election", then maybe it's about as significant as left handed Republicans whose mother is from New Jersey.
The beauty of attribution is that some random anonymous Wikipedia editor is not in charge of deciding if it matters if you're the first left handed Republican who did this or did that. Nobody needs debate it. If the sources highlight it, then we know it matters. It's easy. -- Dennis Bratland ( talk) 20:35, 31 March 2017 (UTC)
I'm confused by the fact that some of the write-ins were official nominees of parties and yet are listed as independents in the Statewide Results box near the top (specifically, Mike Maturen - American Solidarity Party & Jerry White - Socialist Equality Party). I'm guessing this is some kind of technicality...if parties don't make the ballot, their nominees aren't officially recognized as being affiliated with the party. I see the same pattern for 2012 results, so I assume it's intentional. I think readers would be interested to know what write-in candidates were running as nominees of parties, if there are any established criteria for what a party is (I understand that someone might create a "party of 1" for their write-in candidacy that is basically a party in name only).
EDIT: I see that basically a political party exists in the eyes of the state when they have enough signatures or registered voters to get on the ballot, and there is probably no other form of official recognition. It might still be useful to identify parties that independents associate themselves with, even if they are parties in name only. If a candidate is officially recognized as a write-in and gets enough votes to make it on the wikipedia page, I'd think readers would be interested to know under what party banner (if any) that person was running. It could be an additional column on the table at the far right, or a footnote paragraph below the box.
EDIT: California's Secretary of State Office made one major error (although inconsequential in the present instance). There were separate slates of electors for the Republican Party candidates (Trump and Pence) and for the American Independent Party candidates (Trump and Pence), but no way for voters to distinguish these electors on California ballots.
California Republican Party Electors Pledged to Donald J. Trump for President of the United States and Michael R. Pence for Vice President of the United States:
Joel Anderson, Marilyn Barke, Jennifer Beall, Robert Bernosky, Arun Bhumitra, Jim Brulte, Nachhattar Chandi, Claire Chiara, Tim Clark, Greg Conlon, Matthew Del Carlo, Harmeet Dhillon, Elizabeth Emken, Jean Fuller, Ted Gaines, Ron Gold, Lisa Grace-Kellogg, Barbara Grimm Marshall, Howard Hakes, Diane Harkey, Matthew Harmon, Noel Irwin Hentschel, Mark Herrick, Tom Hudson, Kenneth Korbin, Kevin Krick, Linda Lopez-Alvarez, Robin Lowe, Papa Doug Manchester, Shirley Mark, Chuck McDougald, John Musella, Ron Nehring, Mike Osborn, Douglas Ose, John Peck, Pete Petrovich, Donna Porter, Dennis Revell, Scott Robertson, Carla Sands, Truong Si, Robert Smittcamp, Mike Spence, Jeff Lalloway, Shawn Steel, Mark Vafiades, Marcelino Valdez, Errol Valladares, Cyndi Vanderhorst, Megan Vincent, Elissa Wadleigh, Deborah Wilder, Dave Willmon, John Young
American Independent Party Electors Pledged to Donald J. Trump for President of the United States and Michael R. Pence for Vice President of the United States:
Linda Lea Alsbury, Merwyn Alsbury, The Honorable Steve Baldwin, Gary Brown, Ruth Brown, Mark Brownlee, William C. Cardoza, Joseph J. Cocchi, Julie Colglazier, Kayla Colglazier, Patrick Colglazier, Dr. J. Steven Davis, Sallie Hansen Doman, The Honorable Robert K. Doman, Wiley Drake, Sally S. Easter, The Honorable Virgil Goode, Ron Gold, Jeff Grage, Charles Edward Harrison, Jr., Thomas Nowlen Hudson, The Honorable John LeBoutillier, The Honorable Robert Marc Levy, Mary Parker Lewis, Gaudencio Gene Lopez, Judy Lopez, Raul Lopez, Sheila Schultz Lopez, Leonard Luna, Kim McDermott, Eric McDermott, Arthur Loyal Morgan, Matthew Justin Morgan, Richard Matthew Nettleton, Sr., Julie Marie Nettleton, Marc Nettleton, Jaycob Andrew Ornelas, Melissa Ornelas, Robert Ornelas, Marilyn Plumb, Jamie Rangel, Jeffrey Rangel, John Daniel Robertson, Markham Robinson, Mary Robinson, Stephanie Roundy, Terrance Arthur Rust, Dustin Paul Salsi, Richard Scott, Andrew Schalo, David James Scholl, Mark J. Seidenberg, Chris Smentech, Glenn Smentech, Michael Warnken, Jack Warren
Both slates of electors are listed in the Certificate of Ascertainment has each having received 4,483,810 votes.
Fortunately, the Democratic Party electors received 8,753,788 votes and therefore won. However, if the Trump-Pence ticket had won in California, the Secretary of State would have been extremely embarrassed when both sets of electors showed up in Sacramento on December 19, 2016 to cast their ballots for President and Vice President, each claiming to have been duly elected as the state's electors. — Preceding unsigned comment added by 173.228.6.44 ( talk) 19:19, 21 March 2017 (UTC)
If the Certificate is "NOT a mere formality", then can you explain why this supposed error made no difference? Seems like the definition of "formality" if there can be an error like this yet it has no effect. -- Dennis Bratland ( talk) 18:27, 31 March 2017 (UTC)
Very well! You and I disagree on the interpretation of what constitutes original research and the proper way to cite primary references (such as public documents and statutory law sections) to meet wikipedia requirements, but since we can agree on at least certain minimal content, we'll leave the text per your reduction. — Preceding unsigned comment added by 173.228.6.44 ( talk) 19:37, 3 April 2017 (UTC)