Political Polling, Nov 7, 2016

[BoiseBYU]

Interestingly, HRC has more total votes than Trump. It is close, but still she has more overall votes than he.

[SpiffCoug]

Unfortunately for Hillary and those still struggling to cope w/ Trump losing the popular vote but winning the electoral college, they don't understand how our system works.

Here's an analogy. Elections are like a football game. The popular vote is passing efficiency rating. The electoral college is points scored.

You can have a PER as high as you want, and while it is important, it's still not as important as points scored. Yes, there is a high correlation between a high PER (popular vote) and points scored (electoral college), but you can still win games with a low PER and you can lose games with a high PER.

Whining about winning the PER battle and somehow claiming that was the more important metric than points scored, makes a bitter SpiffCoug on Cougar Corner.

[BoiseBYU]

Unfortunately for Hillary and those still struggling to cope w/ Trump losing the popular vote but winning the electoral college, they don't understand how our system works.

Here's an analogy. Elections are like a football game. The popular vote is passing efficiency rating. The electoral college is points scored.

You can have a PER as high as you want, and while it is important, it's still not as important as points scored. Yes, there is a high correlation between a high PER (popular vote) and points scored (electoral college), but you can still win games with a low PER and you can lose games with a high PER.

Whining about winning the PER battle and somehow claiming that was the more important metric than points scored, makes a bitter SpiffCoug on Cougar Corner.

Agreed. To quote the famous philosopher, dem da berries. I hope you aren't thinking I'm whining. I wasn't and I'm not. It is not the norm though for the President elect not to earn the most votes. It happened here though.

[StatsCougar]

Post Mortem:
The failure of the polls this cycle is going to be a hot topic in the industry over the next few months. This was a real black eye and trust will be in short supply. There are several reasons why this may have occurred, some of which were have been suggested by people on this board. Here are my thoughts as to why the results of the polling were so far off.

Bradley Effect: This was a term coined in the 80's when polling for the mayoral race in LA showed one result and the actual results were very different. It was theorized that Mr. Bradley, an African-American, was polling higher because voters wanted to do the PC thing and say they were voting for the minority when they actually were going to vote for his opponent.

Much the same has happened here, with some people not wanting to say they were voting for Trump, as the media had insulted the "Trump Voters". It wasn't PC. There was some surveys that sought to look at this issue and find a way to look at the bias. They did it by asking "Who is your neighbor voting for?" As most people will project their beliefs on others, you could use this to detect for bias.

When creating the simulations, I noticed these polls in the final run and thought it was odd. I thought it was biased methodology towards Trump. It was correcting for the bias against Trump.

Women were one of the most common people to have this difference.

Likely Voters: Every polling company has there own secret sauce that they use to define what a likely voter would be. Some ask how likely you are to vote, combined with how often you have voted in the past. This is done knowing that people who have voted frequently in the past are more likely to vote again. It appears that people who have been infrequent voters in past came out in higher proportions then in previous years.

Voter Turn Out: It appear that rural counties had a higher turn out than normal, and overall was not as good as in previous years. This could be for several reasons. The first is people wanted a change. The second is enthusiasm. Republicans had it with many people really wanting to vote for Trump and the Democrats did not. The announcement of the FBI looking at Hillary's email again would help to default this.

It is interesting to note, that if you combine the 2012 and 2016 votes, Trump would have finished last with the number of total votes received. A low voter turnout favors the republicans. This piece is also a that modeled professional that is often hard to get right.

Third Party Switch: The percent of voters that will go third party is often hard to get right. Polling before the election showed about 7% of the vote going to third parties, when in reality, it was closer to 4%. I don't know how those voter actually voted, but I would guess many ended up voting for Trump.

It is also interesting to note that if voters were disliked both Hillary and Trump, they were more likely to vote for Trump.

[BroncoBot]

Unfortunately for Hillary and those still struggling to cope w/ Trump losing the popular vote but winning the electoral college, they don't understand how our system works.

Here's an analogy. Elections are like a football game. The popular vote is passing efficiency rating. The electoral college is points scored.

You can have a PER as high as you want, and while it is important, it's still not as important as points scored. Yes, there is a high correlation between a high PER (popular vote) and points scored (electoral college), but you can still win games with a low PER and you can lose games with a high PER.

Whining about winning the PER battle and somehow claiming that was the more important metric than points scored, makes a bitter SpiffCoug on Cougar Corner.

Agreed. To quote the famous philosopher, dem da berries. I hope you aren't thinking I'm whining. I wasn't and I'm not. It is not the norm though for the President elect not to earn the most votes. It happened here though.

I've really come to appreciate the electoral college. California has nearly 50M people. If they had all voted Hillary, California still only gets 55 votes. It provides each individual and state to have its say in the election.

interesting that I've heard a lot of complaining from those who wanted Hillary regarding the popular vote. In 2008 Hillary also won the popular vote against one BHO, but I can't remember hearing a word about it then...

[BoiseBYU]

Unfortunately for Hillary and those still struggling to cope w/ Trump losing the popular vote but winning the electoral college, they don't understand how our system works.

Here's an analogy. Elections are like a football game. The popular vote is passing efficiency rating. The electoral college is points scored.

You can have a PER as high as you want, and while it is important, it's still not as important as points scored. Yes, there is a high correlation between a high PER (popular vote) and points scored (electoral college), but you can still win games with a low PER and you can lose games with a high PER.

Whining about winning the PER battle and somehow claiming that was the more important metric than points scored, makes a bitter SpiffCoug on Cougar Corner.

Agreed. To quote the famous philosopher, dem da berries. I hope you aren't thinking I'm whining. I wasn't and I'm not. It is not the norm though for the President elect not to earn the most votes. It happened here though.

I've really come to appreciate the electoral college. California has nearly 50M people. If they had all voted Hillary, California still only gets 55 votes. It provides each individual and state to have its say in the election.

interesting that I've heard a lot of complaining from those who wanted Hillary regarding the popular vote. In 2008 Hillary also won the popular vote against one BHO, but I can't remember hearing a word about it then...

I think it is understandable when someone gets the most votes but still does not win to be upset about it. Didn't Gore win more votes than Bush in 2000? But it is the rules of the game and everyone knows it going in. I do not like the Electoral College myself. It means states like CA or ID or UT are irrelevant in Presidential campaigns because the outcome in those States is so clear there is no need to campaign and engage the voters there. That is why you see all the campaigning in the Battleground States and the raising of issues that appeal to voters in those states. IF there were no electoral college, your vote would be worth as much as a Floridian vote, campaign speaking-wise (is that a word?), and there would be more incentive to win your vote because now your vote counts as much as any other vote. But it isn't going to be changed and it is the system we have. It is why it ws only when UT looked like the race was close did they send Pence out for a day. Every ten years when they reconfigure each States' congressional delegations census wise we get closer to having the EC equal popular votes, but it is never perfect. But like I said, no complaining from me! C'est la vie!

[BroncoBot]

Yeah, we definitely don't agree.

[BOID]

300K to 350K absentee votes were not processed/counted in the 2000 election. An estimated 7000 absentee applications were thrown away by dem operatives in Florida. They used the excuse that the applicants had forgotten to record their zip codes. Zip codes had historically been provided and added as a courtesy at the polling registration stations. Dem election workers in some Florida offices, sought for, and obtained, permission to throw them out. Lawsuits were threatened over this, but GWB ended up winning anyway.

In many other states, huge numbers of absentee ballots were not processed because their home states had their respective elections called before the out-standing ballots' legally determined mail arrival times had elapsed. State elections were called earlier, one way or the other, because the number of remaining absentee ballots, known to be due, but not yet delivered, were insufficient in number to change the leads established in their home states.

However, the number of yet undelivered absentee ballots Nation-wide were, in summation, sufficient in number to potentially overcome Al Gore's small lead in the total popular vote. Had the Nation-wide popular vote mattered for the outcome/win, then of course, all of the out-standing ballots would have been eagerly awaited and tallied.

My point? No one knows if Gore really won the popular vote. Democrats have trumpeted that popular/electoral disparity with great zeal, in an attempt to keep a moral victory, and undermine the legitimacy of George W's win. The absentee voter block typically is dominated by registered Republicans, and it therefore remains a possibility that Gore lost, or would have lost, on both counts.

[hawkwing]

Interesting point BOID. I have heard as well that Donald Trump could be the popular vote winner of this election as well. I, in fact, heard that the AP reported that fact, but I just heard that and haven't actually seen it with my own eyes, so I can't confirm it.

[SpiffCoug]

Since we're talking popular vote, I thought it was interesting to see how many states had a majority vote for one candidate. It turned out 38 states saw a majority and 24 were won by Trump. Why should the 14 states were Clinton won a majority have more say than the other 37 states. (Yes I know that adds up to 51, due to D.C.).


Trump (24 states)
Alabama - 62.9%
Alaska - 52.9%
Arkansas - 60.4%
Georgia - 51.3%
Idaho - 59.2%
Indiana - 57.2%
Iowa - 51.8%
Kansas - 57.2%
Kentucky - 57.2%
Louisiana - 58.1%
Mississippi - 58.3%
Missouri - 57.1%
Montana - 56.5%
Nebraska - 60.3%
North Carolina - 50.5%
North Dakota - 64.1%
Ohio - 52.1%
Oklahoma - 65.3%
South Carolina - 54.9%
South Dakota - 61.5%
Tennessee - 61.1%
Texas - 52.6%
West Virgina - 68.7%
Wyoming - 70.1%


Clinton (14 states)
California - 61.5%
Connecticut - 54.5%
Delaware - 53.4%
Hawaii - 62.2%
Illinois - 55.4%
Maryland - 60.5%
Massachusetts - 60.8%
New Jersey - 55.0%
New York - 58.8%
Oregon - 51.9%
Rhode Island - 55.4%
Vermont - 61.1%
Washington - 54.9%
Washington, D.C. - 92.8%


Neither (13 states)
Arizona - Trump 49.5%
Colorado - Clinton 47.3%
Florida - Trump 49.1%
Maine - Clinton 47.9%
Michigan - Trump 47.6%
Minnesota - Clinton 46.9%
Nevada - Clinton 47.9%
New Hampshire - Clinton 47.6%
New Mexico - Clinton 48.3%
Pennsylvania - Trump 48.8%
Utah - Trump 46.6%
Virginia - Clinton 49.9%
Wisconsin - Trump 47.9%

Trump won 60% or more in 9 of his 24 majority states (.375). Clinton won 60% or more in 6 of her 14 states (.429). Clinton won seven of the "battleground states". Those arguing against the electoral college simply 1) don't understand how our system works and 2) are upset they lost.