Voting -- a better system

RAZD writes, in part: From List of United States presidential elections in which the winner lost the popular vote - Wikipedia we have: I would hardly ever call that losing badly.

http://www.rangevoting.org/USsplitLine.png

One has to be careful of course, that you don't end up with

http://www.rangevoting.org/SSHR/co_final.png

Because this may be arguably better to the current boundaries in Colorado, it's still a bit mad. See details here

Make the number of Representatives in Congress from each state be determined by the voter turnout. State X has population P. Of the P, only T turnout to vote. They would get T/310,000 (or something like that) = N Representatives, proportioned out by the top N vote tallies for all the candidates. N would be something like the smallest integer greater than T/310,000.0Something to encourage voting turnout.The number, 310,000, would be selected to give 436 Representatives usually as of today, unless this actually significantly increases turnout soon....With a turnout in 2016 of some 134.5 million, the 436 Representatives, on average, represent 308,486 voters that actually turned out.

Couple this with ranked IRV and my At Large Representatives, and, viola'!, no gerrymandering, no spoilers, just the will of the people who actually want to get out and vote.

Curiously I also see this number put into perspective with this graphic Which says to me that she lost badly because she failed to energize or inspire democrats and independents to bot for her.And I find it sadly ironic to see the Hillary people grasping at straws to say it wasn't her fault that she actually won ... when they were all over the Bernie people that were saying Bernie actually won.The election was hers to lose, she did.Enjoy

Your bargraph is already way out of date for 2016. The current tally is: from 2016 United States presidential election - Wikipedia-------------------------------------------------------------------------------------------------
...and they're still being counted.this means that Clinton got more votes than any White Male candidate for President in the history of the country. She probably wont catch Obama's 2012 count and certainly not his 2008 count.so redraw your bargraph.

Now Popular vote 62,235,228 Trump ==> but 64,235,413 Clinton Edited by xongsmith, : No reason given.Edited by xongsmith, : No reason given.

Are they also recounting senate races? Personally I think that is much more important ... supreme court nominations.And how is it going for the swing states (electoral college delegates).If the recount overturns the election we can expect uprising war.Enjoy

That gives us 70% 'like' or 'can live with' Hillary ... if your assumed values are correct.But I think 50% of Trump voters for Clinton is high for this election.
and I think 75% of Bernie voters for Clinton is high for this election.In this year of breaking establishment politics and Hillary hate baggage, what I see is more people unhappy with the choices and staying home, that more people voted against rather than for the candidates,If 50% of Trump voters can 'live with' Bernie
If 75% of Hillary voters can 'live with' Bernie
That gives us 75% 'like' or 'can live with' Bernie.Anecdotal tid-bit: I overheard 3 professional nurses discussing the election, all three were democrats, but 2 voted for Trump and one did not vote. Solid middle-class women with educations.EnjoyEdited by RAZD, : .

RAZD writes: So sad that 30 years of hateful propaganda and lies from the GOP worked.

RAZD ponders: Looks like your gal Jill is riding in to the rescue, demanding recounts in Pennsylvania, Wisconsin and Michigan.I think they ought to recount Florida and North Carolina as well.If the evidence proves credible enough that the machine votes were rigged, then we may get the necessary 38 Faithless Electors we need for an historic event.We should be whispering into their ears that their names will go down in history like no one else's. We have 25 days left to do it.ObSenate: My friend in Wisconsin told me that the Koch brothers inundated the TV with 100's of negative ads about Russ Feingold, but that Russ's compaign was hardly visible on the airwaves at all.
----------------------------------------
62,327,523 64,424,986 lead for Hillary as of yesterday.

That doesn't matter, though, does it? The numbers were for illustration only. You could just easily say 25% and Clinton still has a clear majority of 'tolerants'. Also, given how many people voted for Clinton it strongly suggests that way more than 75% of Bernie supporters actually voted for her suggesting that in a choice between her and Trump, most of them 'could live with' Hillary. As I said 'on the typical simplified two dimensional plane of left / right'. Obviously a real example from an atypical election may not have party 'wing' politics as its predominant force. Hahaha. Your anecdote suggests you are literally referring to 'people who voted for Trump in this specific election' rather than in the fictional one. If Sanders was running as a second Democrat nominee these people would likely not have been Trump voters...they aren't Trump supporters.But long story short, if it turns out that Sanders really is the middle ground appeal in this particular election (Rather than a left/right, there is a status quo/shake it up divide dominating) then just change the fictional results toTrump: 39
Sanders: 20
Hillary: 41And Sanders wins, even though he'd lose in IRV. These numbers aren't realistic, you may protest, never fear, I reply. Just imagine Sanders runs as a Third Party and continues to suffer from dirty tricks from establishment Party politics and the usual FUD.Clinton wins in IRV and plurality, and majority (assuming a run-off occurs and Clinton wins the popular vote and that's what matters as per the actual numbers)Yes, the plurality method, the first past the post with run-off method and the Condorcet method have problems. But many of those same problems or others exist in IRV too, and that's my point. You seem to want to focus on the other systems, which is fine to an extent, but you are excluding dealing with the problems with IRV,Another problem with the Instant-Runoff method: It's possible for a voter to harm a candidates chances by ranking them higher. Given ballots are not publicly released for scrutiny, we'd likely never know if this has ever been the case, and estimates at probability of it happening in a given election that some voters actually did harm a candidate like this (or help a candidate by ranking them lower).Again, imagine the three candidates were running and every voter only voted for two candidates (because they 'can't stand' the third candidate). There are eight possible combos so let's run a fictional election, contrived to create this scenario for illustration, and using your alternative perspective on where second preferences may actually lie {as much as is possible}:Trump- Bernie 28 million
Trump -Clinton 5
Clinton -Bernie 30
Clinton -Trump 5
Bernie -Clinton 16
Bernie -Trump 16Clinton wins plurality with 35 million votes. Trump gets 33 and comes second. Bernie is eliminated. Since Bernie's voters are split evenly Clinton wins the majority.But imagine if two million of those Trump -> Bernie folk had decided to vote for Bernie -> Trump instead we'd getTrump- Bernie 26 million
Trump -Clinton 5
Clinton -Bernie 30
Clinton -Trump 5
Bernie -Clinton 16
Bernie -Trump 18Now Trump gets eliminated and Bernie wins. So in some sense, some people's votes actually acted against their preferences completely. It is not monotonic.So what else? Well an important point, given the electoral college, is that it does not return consistent results. Imagine after everyone has voted we tally up their preferences and Bernie Sanders is the winner. That's nice. But let's say there is some kind of {simplified} electoral college. We divide the nation into 538 pieces on some arbitrary measure like, geography or something, each piece getting one electoral vote based on the local IRV winner. Each piece gets an electoral vote and it's the electoral college that matters ultimately. Now Trump is the winner. Is this a good thing? Not really. The Borda count system manages to avoid this inconsistency:Borda count - WikipediaAnother way to look it: Imagine if Bernie wins as before, but then we discover a cache of uncounted votes. We count them separately and Bernie is the IRV winner of this subset. If we add these ballots to the total, is Bernie guaranteed to be the winner if a second IRV using both the original and the cache ballots? The answer is no.Furthermore, it's a real risk that after the votes have been counted, one group of voters would have been better off not bothering to vote at all. Not voting would have got them their preferences.Let's collapse our table down and widen it and simplify aggressively.Bernie > Clinton > Trump: 50 million
Clinton > Trump > Bernie: 40 million
Trump > Bernie > Clinton: 60 millionIn IRV Clinton is eliminated and then Trump wins. I appreciate this is not realistic given the specifics of this election I'm just using names over letters for human readability.Let's say 20 million Bernie supporters didn't bother turning up at all. That *should* hurt the Democrats, right? They all prefer Democrat candidates, after all. But noBernie > Clinton > Trump: 30 million
Clinton > Trump > Bernie: 40 million
Trump > Bernie > Clinton: 60 millionBernie is eliminated, Clinton wins and the Bernie supporters at least don't have Trump which is their preference.That's really strange.http://www.sciencedirect.com/...article/pii/0165489686900247

quote:

---

In a recent paper Brams and Fishburn (1983) have demonstrated that the ‘single transferable vote’ system for a single member electoral district admits a ‘no show paradox’. The present work indicates that under the single transferable vote the probability of occurence of a no show paradox is very high in situations where the outcome under the single transferable vote differs from that under plurality.

---

The paper estimates that in the very edge cases that make IRV so appealing or interesting to some, where the 1st preference plurality and final IRV results differ...there is as much as a 50% chance that some voters would have done better by not turning up.How about this:50 million voters prefer Bernie then Trump then Clinton
40 million voters prefer Trump then Clinton then Bernie
20 million voters prefer Clinton then Bernie then TrumpIn IRV Bernie wins right?But what if the same electorate had instead completely reversed their preferences? We should get a completely different result? The last 20 million's first preference is irrelevant and Bernie gets their votes regardless. Since Bernie gets first preference of the 40 million he still wins even though everybody switched their first and last preferences around. Odd. This means the system is not monotonic, does not elect the condorcet winner, not naturally separable into consistent 'electoral votes', means non-participation can sometimes be advantageous, has the possibility of encountering the reversal symmetry paradox, as well as potentially other issuesHere is handy table comparing preferential voting methods by these kinds of criteria:Schulze method - WikipediaOf course, other methods are a little more difficult to understand. I don't expect they'd be more popular than IRV even if they are technically superior unless enough people got annoyed at the results of one if its quirks after it has presumably been in play for a while and people have grown up with it.{examples and criteria courtesy of wiki:
Electoral system - Wikipedia}Edited by Modulous, : No reason given.Edited by Modulous, : No reason given.

What about kicking candidates off the island?No parties, start with the whole country's population of POTUS eligibles, and vote week1. Throw out everyone in the lower 80-90%* Repeat Week2. Repeat until there are close to 32 survivors left, and then make it the top the 32 survivors. Call this the Line, then.Throw out half each Week (5 more times) until we have only one survivor.Should be the one most can live with?
??* or what ever it takes to get the field reduced to 32 in the old Primary Season time.Edited by xongsmith, : No reason given.

So sad that the DNC was blind to this being a significant factor in her unfavorability rating. In politics perception is 9/10s of reality ...And that they were blind to what the base wanted to see in a candidate.Enjoy

They need $ to cover the costs for Michigan, Pennsylvania and Wisconsin costs are now covered ...Indeed. Donations to the cause should be made at:
https://jillstein.nationbuilder.com/RecountFor general volunteering sign up at:
http://www.jill2016.com/postelectionVolunteers to help recount in Wisconsin can sign up at:
http://www.jill2016.com/RecountWIVolunteers to help recount in Michigan can sign up at:
http://www.jill2016.com/RecountMIVolunteers to help recount in Pennsylvania can sign up at:
http://www.jill2016.com/recountpaPersonally I think recounts should be mandatory whenever the vote is close (within 10%?) and certainly when the difference between the top two candidates is less than the remaining numbers of of voters.And this is also where a ranked voting system is really needed, whether IRV or Condorcet.Participation in democracy is more than just going to the polls, it is also actively demanding, protesting and volunteering to ensure that the vote counting is accurate and faithfully reflects the votes of the people.Enjoy