Seth Rich story resurfacing

[Paratus]

Awesome source, they totally called the election spot on.

Who will win the presidency?
Chance of winning

Hillary Clinton

71.4%

Donald Trump

28.6%

538 was spot on. If you think otherwise you don't understand scientific polling.

Right before the election they had it 33% for The Donald and 66% for Clinton.

Hillary did win the popular vote by several %.

They warned that the polling was close and Trump had a legitimate chance - which he obviously did.

Let me put it this way. You load two rounds into a revolver and spin the cylinder. 538 says you have a 33% chance of blowing your head off.

You pull the trigger and blow your head off. Would you have said 538 wrong for only giving 33% chance of that happening?

 

[fskimospy]

What do you mean by "it is true in the general as well"? I'm not referring to polls of Sanders during the general. I'm pointing out according to Sam Wang's article, the polls don't even become as accurate as February until August (i.e. general election polls of the one who was nominated!). And I don't see any reason to assume that if a candidate loses in the primary, that it somehow means their February polls weren't accurate.

I think there is a very, very good reason why. The presumptive nominees are generally known in February and people who aren't the nominee generally get far more favorable treatment from those who lean towards the other political party. This is a well known phenomenon. People are always far more willing to vote for an opposite party candidate that they don't actually have to vote for.

Silver's model sucked. It changed on the dime, and he was being conservative due to how humiliated he got after being so wrong on Trump and the Cubs. lmao

Silver did incorrectly underestimate Trump's chances in the primary but that came from him not trusting his own model that said Trump had a very good chance. I agree that his model was too sensitive to changes in the middle of the election season but overall it performed quite well. Far better than Sam Wang's, who you apparently think highly of.

Sam Wang's issue was simple. He was modeling after a margin between polls and results that was considerably smaller than what recent elections suggested (0.5 vs. 1.2 or so). If he put the correct input, it was around 88-90% chance she would win. She still essentially won, though. It was 2% and a loss of only about 80K votes divided among three states.

No, from what I remember his failure was assuming polling errors weren't correlated.

 

[Maxima1]

538 was spot on. If you think otherwise you don't understand scientific polling.

Right before the election they had it 33% for The Donald and 66% for Clinton.

Hillary did win the popular vote by several %.

They warned that the polling was close and Trump had a legitimate chance - which he obviously did.

Let me put it this way. You load two rounds into a revolver and spin the cylinder. 538 says you have a 33% chance of blowing your head off.

You pull the trigger and blow your head off. Would you have said 538 wrong for only giving 33% chance of that happening?

They weren't "spot on". The model was crap. What did they give a popular vote upset anyway? 1-2% chance?

 

[Maxima1]

I think there is a very, very good reason why. The presumptive nominees are generally known in February and people who aren't the nominee generally get far more favorable treatment from those who lean towards the other political party. This is a well known phenomenon. People are always far more willing to vote for an opposite party candidate that they don't actually have to vote for.

Again, I'm referring to the fact that Sam Wang is making the claim that polls don't even become as accurate as February until August (which is some point during the general).

Silver did incorrectly underestimate Trump's chances in the primary but that came from him not trusting his own model that said Trump had a very good chance. I agree that his model was too sensitive to changes in the middle of the election season but overall it performed quite well. Far better than Adam Wang's, who you apparently think highly of.

BS His model assigned high probabilities of Trump winning by 3% or more. He probably assigned a very low probability of a popular vote upset of 1-2% chance.

No, from what I remember his failure was assuming polling errors weren't correlated.

I'll try looking for it.

 

[fskimospy]

They weren't "spot on". The model was crap. What did they give a popular vote upset anyway? 1-2%?

Nah, they actually gave it around a 10% chance, which when coupled with Trump's ~28% chance shows that they thought if Trump did win a popular vote split with the electoral vote was highly likely.

Their model has performed really well in every election so far. I agree it's volatility in the medium term is a problem but overall it's excellent.

 

[fskimospy]

Again, I'm referring to the fact that Sam Wang is making the claim that polls don't even become as accurate as February until August (which is some point during the general).

Again, he is only making claims as to polls between eventual nominees. His analysis is obviously inapplicable to those who are not nominated. Why matchups between non-nominees and nominees are not comparable is covered by what you quoted.

 

[brycejones]

duhversion thread is duhversion thread shame on the OP.

 

[Paratus]

They weren't "spot on". The model was crap. What did they give a popular vote upset anyway? 1-2%?

538's analysis told me:

• The vote would be close
• The polls were volatile
• A Trump win was a significant possibility

What did we see:

• A popular vote - EC split
• States were very close - with Clinton losses of a few 10's of K in certain states
• A Trump win

It's possible that "changing on a dime" was actually due to the rapidly changing negative media stories about each candidate being reflected in the polling.

If I was running a campaign and saw those numbers I would have been worried.

 

[Maxima1]

Nah, they actually gave it around a 10% chance, which when coupled with Trump's ~28% chance shows that they thought if Trump did win a popular vote split with the electoral vote was highly likely.

Their model has performed really well in every election so far. I agree it's volatility in the medium term is a problem but overall it's excellent.

No, I should have clarified Hillary winning by 2% and still losing. That is likely a 1-2% chance according to his model. The electoral college loss of 10% is including very narrow pop vote losses.

Again, he is only making claims as to polls between eventual nominees. His analysis is obviously inapplicable to those who are not nominated. Why matchups between non-nominees and nominees are not comparable is covered by what you quoted.

'
What factors would make it so that the eventual nominee has numbers that reflect accurately back to February but the loser doesn't?

• A popular vote - EC split
• States were very close - with Clinton losses of a few 10's of K in certain states
• A Trump win

Trump won barely. You wouldn't be saying this had Hillary picked up just 81K votes.

 

[Maxima1]

Nah, they actually gave it around a 10% chance, which when coupled with Trump's ~28% chance shows that they thought if Trump did win a popular vote split with the electoral vote was highly likely.

Their model has performed really well in every election so far. I agree it's volatility in the medium term is a problem but overall it's excellent.

Here is what I was referring to. Sam Wang messed up, and people were baffled by his choice, since it was clearly wrong. He posted it on Nov. 6th.

http://election.princeton.edu/2016/11/06/is-99-a-reasonable-probability/

 

[Paratus]

No, I should have clarified Hillary winning by 2% and still losing. That is likely a 1-2% chance according to his model. The electoral college loss of 10% is including very narrow pop vote losses.


'
What factors would make it so that the eventual nominee has numbers that reflect accurately back to February but the loser doesn't?


Trump won barely. You wouldn't be saying this had Hillary picked up just 81K votes.

An 81K swing in Hillarys favor and I still would be saying 538 was basically correct.

So I'm not sure I get your point.

 

[Maxima1]

An 81K swing in Hillarys favor and I still would be saying 538 was basically correct.

So I'm not sure I get your point.

Uh, well your standards are low.

 

[Paratus]

Uh, well your standards are low.

Ok then

 

[Balt]

Apparently Sean Hannity made this Seth Rich thing his lead story tonight.

Fox News has gone full Infowars and /r/the_donald. Pathetic.

 

[alcoholbob]

http://www.foxnews.com/politics/201...contact-with-wikileaks-investigator-says.html

If the Seth Rich story turns out to be true and he's the leaker who spilled the story about the DNC conspiring against Bernie Sanders, doesn't that disprove the entire "Russians colluded to undermine our election"

Even if Seth Rich was the one who provided the document dump to Wikileaks, that's doesn't mean the Russians weren't trying to influence the election, since they were obviously meeting with Republican and Trump officials and involving themselves in cyber espionage, fake news and mass online trolls. The only thing this lead, if true, would simply clear the Russians of being involved in the specific charge of giving Wikileaks the DNC documents.

 

[woolfe9998]

Even if Seth Rich was the one who provided the document dump to Wikileaks, that's doesn't mean the Russians weren't trying to influence the election, since they were obviously meeting with Republican and Trump officials and involving themselves in cyber espionage, fake news and mass online trolls. The only thing this lead, if true, would simply clear the Russians of being involved in the specific charge of giving Wikileaks the DNC documents.

Don't say "even if." The hypothetical is pointless. Not only did those docs have Russian metadata which any member of the general public could see, but our entire intel community says it was Russia. There have been reports that they have wiretapped conversations between Russians who are discussing the fact that they did it. It's open and shut.

 

[akenbennu]

http://money.cnn.com/2017/05/16/med...sponse-claims-of-wikileaks-contact/index.html

From the article: "Earlier Tuesday, Brad Bauman, a spokesman for the Rich family, released a statement in which he said the family had "seen no evidence" to suggest Seth Rich had been in contact with Wikileaks."

Really looks like a conspiracy theory here.

 

[Vic]

Apparently Sean Hannity made this Seth Rich thing his lead story tonight.

Fox News has gone full Infowars and /r/the_donald. Pathetic.

The Trump camp is desperate.

 

[fskimospy]

http://money.cnn.com/2017/05/16/med...sponse-claims-of-wikileaks-contact/index.html

From the article: "Earlier Tuesday, Brad Bauman, a spokesman for the Rich family, released a statement in which he said the family had "seen no evidence" to suggest Seth Rich had been in contact with Wikileaks."

Really looks like a conspiracy theory here.

And what sucks is that dishonest people have turned this into an issue that needs to be rebutted almost certainly to distract from criminal behavior from elected officials.

What is wrong with them.

 

[piasabird]

It is sounding like the FBI and the DC Police knew about this and someone destroyed all the evidence.

 

[Vic]

It is sounding like the FBI and the DC Police knew about this and someone destroyed all the evidence.

What it sounds like is a baseless conspiracy theory to distract Trump's loyal base from his own misdeeds.

 

[Victorian Gray]

It is sounding like the FBI and the DC Police knew about this and someone destroyed all the evidence.

Been to any good pizza joints lately?

 

[Vic]

Was just reading comments over at Fox News and some other right wing sites.. c'mon people, this story is almost a year old and yet it was suddenly thrown to the top of the right wing news cycle with no new information just as Trump is having a major meltdown. I'd have to say it's more than obvious which story is the distraction.

 

[agent00f]

538 was spot on. If you think otherwise you don't understand scientific polling.

Right before the election they had it 33% for The Donald and 66% for Clinton.

Hillary did win the popular vote by several %.

They warned that the polling was close and Trump had a legitimate chance - which he obviously did.

Let me put it this way. You load two rounds into a revolver and spin the cylinder. 538 says you have a 33% chance of blowing your head off.

You pull the trigger and blow your head off. Would you have said 538 wrong for only giving 33% chance of that happening?

This is an interest contest between educated people's ability to explain moderately simple math and morons' ability to never understand anything.

 

[mnewsham]

WaPo article about trump release, May 15th 7:45PM EST.

FIRST twitter blogger or news reporter to touch Seth Rich in WEEKS, May 15th 9:48PM EST.
Surprised it only took them 2 hours to come up with a distraction

Also funny how people went a half dozen+ months without any mention of Seth Rich and suddenly within 24 hours its as if Seth Rich is some heroic American that EVERYONE should know about. Just sad how obvious of a distraction this is, and more sad how many are falling for it.