Total Posts:53|Showing Posts:1-30|Last Page
Jump to topic:

Stony Brook Professor's Electoral Model

YYW
Posts: 36,286
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 3:57:01 PM
Posted: 9 months ago
http://www.syracuse.com...

President Trump? A SUNY professor says he's "almost certain" it will happen.

Helmut Norpoth, a political science professor at Stony Brook University, says that his statistical model predicted Donald Trump will win the election if he's the Republican nominee. The billionaire businessman has a 97 percent chance of beating Hillary Clinton in the general election and a 99 percent chance of beating Bernie Sanders, according to Norpoth.

The Statesman reports Norpoth uses statistics from candidates' performances in their party's primary and patterns in the electoral cycle to forecast national election results. The equation correctly predicted Bill Clinton's victory in 1996, George W. Bush in 2004, and Barack Obama in 2012.

In fact, Norpoth's model has correctly predicted the outcome of every presidential election since 1912 except one, according to the Daily Mail. In 1960, John F. Kennedy beat Richard Nixon in one of the closest races in U.S. history.

"When I started out with this kind of display a few months ago, I thought it was sort of a joke," Norpoth said at a gathering Monday at the SUNY Global Center in Manhattan. "Well, I'll tell you right now, it ain't a joke anymore."

He added that the Republican party overall has a 61 percent chance of putting a GOP candidate in the White House, but Clinton has a 55 percent chance of winning in a hypothetical race with Cruz or Marco Rubio.

But Trump's success in the primaries, winning three out of four GOP contests to date, has made him the favorite by far in Norpoth's model -- especially if he gets the Republican nomination.

"The probability of that [outcome] is almost complete certainty, 97 percent. It's almost 'Take it to the bank,'" Norpoth said.
Tsar of DDO
imabench
Posts: 21,215
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 4:18:49 PM
Posted: 9 months ago
I heard about this, and im skeptical.... The Professor's model is almost infallible according to the professor himself, meaning he's the only one claiming that his model has predicted almost every race successfully.

I also am skeptical about models in general since this election more than demonstrates how elections cant be forecasted very well.

Lastly: Trump has the absolute worst head-to-head ratings against Clinton compared to everyone else in the race, same with Sanders http://www.realclearpolitics.com...
Kevin24018 : "He's just so mean it makes me want to ball up my fists and stamp on the ground"

7/14/16 = The Presidency Dies

DDO: THE MOVIE = http://www.debate.org...
http://www.debate.org...

VP of DDO from Dec 14th 2014 to Jan 1st 2015
1harderthanyouthink
Posts: 13,102
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 4:23:37 PM
Posted: 9 months ago
He logistically cannot pull in the vote needed to win, unless a large group of Democrats stay at home.
"It's awfully considerate of you to think of me here,
And I'm much obliged to you for making it clear - that I'm not here."

-Syd Barrett

DDO Risk King
imabench
Posts: 21,215
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 4:31:03 PM
Posted: 9 months ago
At 2/26/2016 4:23:37 PM, 1harderthanyouthink wrote:
He logistically cannot pull in the vote needed to win, unless a large group of Democrats stay at home.

which they wont, since a Trump presidency will scare the living hell out of people and encourage them to come out and vote to stop him
Kevin24018 : "He's just so mean it makes me want to ball up my fists and stamp on the ground"

7/14/16 = The Presidency Dies

DDO: THE MOVIE = http://www.debate.org...
http://www.debate.org...

VP of DDO from Dec 14th 2014 to Jan 1st 2015
1harderthanyouthink
Posts: 13,102
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 4:32:29 PM
Posted: 9 months ago
At 2/26/2016 4:31:03 PM, imabench wrote:
At 2/26/2016 4:23:37 PM, 1harderthanyouthink wrote:
He logistically cannot pull in the vote needed to win, unless a large group of Democrats stay at home.

which they wont, since a Trump presidency will scare the living hell out of people and encourage them to come out and vote to stop him

However, Debbie Wasserman Schultz ought not come off as rigging the election for Hillary and treating Sanders as "obstruction". That's turned a lot of people off from the DNC - and it could be dangerous for them.
"It's awfully considerate of you to think of me here,
And I'm much obliged to you for making it clear - that I'm not here."

-Syd Barrett

DDO Risk King
xus00HAY
Posts: 1,382
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 4:43:18 PM
Posted: 9 months ago
I don't think you have ever been to SUNY Stony Brook Ms. Bench, believe me if you were able to get a job as a professor there you might consider yourself smarter than any person you don't know, but are conversing with. You get to live and work in a beautiful place and work with teen aged girls. The hours you work are short and you get plenty of vacation time. You don't really need much vacation time because if you live in Stony Brook you can just drive for an hour or 2 and be someplace where you would like to visit if you were on a trip.
I suppose you could find another job where you can make more money, if money could buy you happiness.
This is a computer model. This happens to be a kind of program that really works well, It may choose a candidate that you don't like, but we can assume that Mr. Trump will be our next president, unless you can think of a compelling reason why this could not happen.
Look, his opponent will be Hillary, that is not much of a challenge.
imabench
Posts: 21,215
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 4:45:02 PM
Posted: 9 months ago
At 2/26/2016 4:43:18 PM, xus00HAY wrote:

This is a computer model. This happens to be a kind of program that really works well, It may choose a candidate that you don't like, but we can assume that Mr. Trump will be our next president, unless you can think of a compelling reason why this could not happen.

I can think of several
Kevin24018 : "He's just so mean it makes me want to ball up my fists and stamp on the ground"

7/14/16 = The Presidency Dies

DDO: THE MOVIE = http://www.debate.org...
http://www.debate.org...

VP of DDO from Dec 14th 2014 to Jan 1st 2015
xus00HAY
Posts: 1,382
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 4:51:04 PM
Posted: 9 months ago
Again , I am talking about a good reason why Trump couldn't possibly win. not reasons why you think we should not vote for him.
imabench
Posts: 21,215
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 4:52:36 PM
Posted: 9 months ago
At 2/26/2016 4:51:04 PM, xus00HAY wrote:
Again , I am talking about a good reason why Trump couldn't possibly win.

And again, I can think of several
Kevin24018 : "He's just so mean it makes me want to ball up my fists and stamp on the ground"

7/14/16 = The Presidency Dies

DDO: THE MOVIE = http://www.debate.org...
http://www.debate.org...

VP of DDO from Dec 14th 2014 to Jan 1st 2015
YYW
Posts: 36,286
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 5:05:41 PM
Posted: 9 months ago
At 2/26/2016 4:23:37 PM, 1harderthanyouthink wrote:
He logistically cannot pull in the vote needed to win, unless a large group of Democrats stay at home.

Which they might.
Tsar of DDO
1harderthanyouthink
Posts: 13,102
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 5:10:51 PM
Posted: 9 months ago
At 2/26/2016 5:05:41 PM, YYW wrote:
At 2/26/2016 4:23:37 PM, 1harderthanyouthink wrote:
He logistically cannot pull in the vote needed to win, unless a large group of Democrats stay at home.

Which they might.

A model can't really account for that.

A large group of falsely righteous Republicans can stay home, too.
"It's awfully considerate of you to think of me here,
And I'm much obliged to you for making it clear - that I'm not here."

-Syd Barrett

DDO Risk King
YYW
Posts: 36,286
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 5:13:26 PM
Posted: 9 months ago
At 2/26/2016 5:10:51 PM, 1harderthanyouthink wrote:
At 2/26/2016 5:05:41 PM, YYW wrote:
At 2/26/2016 4:23:37 PM, 1harderthanyouthink wrote:
He logistically cannot pull in the vote needed to win, unless a large group of Democrats stay at home.

Which they might.

A model can't really account for that.

Actually it can (which I know because I have worked with people who do this for a living), which is the entire point of these models.

A large group of falsely righteous Republicans can stay home, too.
Tsar of DDO
EndarkenedRationalist
Posts: 14,201
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 5:18:16 PM
Posted: 9 months ago
At 2/26/2016 4:32:29 PM, 1harderthanyouthink wrote:
At 2/26/2016 4:31:03 PM, imabench wrote:
At 2/26/2016 4:23:37 PM, 1harderthanyouthink wrote:
He logistically cannot pull in the vote needed to win, unless a large group of Democrats stay at home.

which they wont, since a Trump presidency will scare the living hell out of people and encourage them to come out and vote to stop him

However, Debbie Wasserman Schultz ought not come off as rigging the election for Hillary and treating Sanders as "obstruction". That's turned a lot of people off from the DNC - and it could be dangerous for them.

If Sanders had never gained traction, and Clinton remained a shoo-in for the nominee, she would easily defeat Trump. But Sanders has sparked the whole 'Bern or bust' movement, and the DNC's response has alienated the populist base. A Trump nomination may scare them back into supporting Clinton, but if it doesn't, she could well lose to Trump.
imabench
Posts: 21,215
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 5:31:36 PM
Posted: 9 months ago
At 2/26/2016 4:32:29 PM, 1harderthanyouthink wrote:
At 2/26/2016 4:31:03 PM, imabench wrote:
At 2/26/2016 4:23:37 PM, 1harderthanyouthink wrote:
He logistically cannot pull in the vote needed to win, unless a large group of Democrats stay at home.

which they wont, since a Trump presidency will scare the living hell out of people and encourage them to come out and vote to stop him

However, Debbie Wasserman Schultz ought not come off as rigging the election for Hillary and treating Sanders as "obstruction". That's turned a lot of people off from the DNC - and it could be dangerous for them.

5 out of 6 Americans cant find Ukraine on a map so I doubt that any aspect of the American electorate will even know who Debbie Schultz is, let alone know about the f*ckery she has pulled with these debates being organized the way they are
Kevin24018 : "He's just so mean it makes me want to ball up my fists and stamp on the ground"

7/14/16 = The Presidency Dies

DDO: THE MOVIE = http://www.debate.org...
http://www.debate.org...

VP of DDO from Dec 14th 2014 to Jan 1st 2015
TBR
Posts: 9,991
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 5:44:55 PM
Posted: 9 months ago
At 2/26/2016 5:13:26 PM, YYW wrote:
At 2/26/2016 5:10:51 PM, 1harderthanyouthink wrote:
At 2/26/2016 5:05:41 PM, YYW wrote:
At 2/26/2016 4:23:37 PM, 1harderthanyouthink wrote:
He logistically cannot pull in the vote needed to win, unless a large group of Democrats stay at home.

Which they might.

A model can't really account for that.

Actually it can (which I know because I have worked with people who do this for a living), which is the entire point of these models.

A large group of falsely righteous Republicans can stay home, too.

I don't disagree that the model cal work, or that it is accurate for the last 100+ years he says. He has gone back in time, and run the algorithm. How else would he have any basis for the research?

It could fail for this cycle, then he would go back and adjust.
YYW
Posts: 36,286
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 5:46:09 PM
Posted: 9 months ago
At 2/26/2016 5:44:55 PM, TBR wrote:
At 2/26/2016 5:13:26 PM, YYW wrote:
At 2/26/2016 5:10:51 PM, 1harderthanyouthink wrote:
At 2/26/2016 5:05:41 PM, YYW wrote:
At 2/26/2016 4:23:37 PM, 1harderthanyouthink wrote:
He logistically cannot pull in the vote needed to win, unless a large group of Democrats stay at home.

Which they might.

A model can't really account for that.

Actually it can (which I know because I have worked with people who do this for a living), which is the entire point of these models.

A large group of falsely righteous Republicans can stay home, too.

I don't disagree that the model cal work, or that it is accurate for the last 100+ years he says. He has gone back in time, and run the algorithm. How else would he have any basis for the research?

It could fail for this cycle, then he would go back and adjust.

Yeah. The point is that it's the best model out there.
Tsar of DDO
dylancatlow
Posts: 12,245
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 5:52:11 PM
Posted: 9 months ago
There are many reasons to be skeptical of his model. First, the near flawless track record is misleading. A model can't receive credit for agreeing with past election results, since all you have to do is find a pattern in the data and construct your model accordingly. Some patterns might reflect underlying forces that make the pattern likely to continue, while others might just be random noise (and in the case of his model, the pattern doesn't even run the whole way through, as it fails to predict the 1960 election). So really, the model only gets points for successfully predicting the last three elections: Clinton, Bush, and Obama. The only one that wasn't obvious was Bush, and the model can't even receive credit for that one. The model bases its prediction on the estimated general election numbers, and of course Bush lost in the 2000 general election yet managed to snatch the presidency anyway, so the model just got lucky.

Second, the model is actually quite unsophisticated. The model's prediction is primarily a function of how well candidates do in the primaries - the relevance to the presidential election is not entirely clear. Trump is doing better than Clinton in the primaries, but so what? Trump could go uncontested in the primaries (which, according to the model, would make him even more likely to win) and still lose to Clinton if he can't win over enough independents. Trump is not a typical candidate. His base of support "runs narrow but deep," as Nate Silver puts it. That means he's likely to run out of steam far sooner than most candidates. Moreover, Trump, as well as the rest of the Republican field, get a major boost because of an assumption built into the model: political parties are less likely to win when they've held office for the previous two terms, which is true historically speaking, but not necessarily true going forward.
TBR
Posts: 9,991
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 5:58:25 PM
Posted: 9 months ago
At 2/26/2016 5:46:09 PM, YYW wrote:
At 2/26/2016 5:44:55 PM, TBR wrote:
At 2/26/2016 5:13:26 PM, YYW wrote:
At 2/26/2016 5:10:51 PM, 1harderthanyouthink wrote:
At 2/26/2016 5:05:41 PM, YYW wrote:
At 2/26/2016 4:23:37 PM, 1harderthanyouthink wrote:
He logistically cannot pull in the vote needed to win, unless a large group of Democrats stay at home.

Which they might.

A model can't really account for that.

Actually it can (which I know because I have worked with people who do this for a living), which is the entire point of these models.

A large group of falsely righteous Republicans can stay home, too.

I don't disagree that the model cal work, or that it is accurate for the last 100+ years he says. He has gone back in time, and run the algorithm. How else would he have any basis for the research?

It could fail for this cycle, then he would go back and adjust.

Yeah. The point is that it's the best model out there.

Yea, again, no argument.

I should also point out, when he says "99% chance" etc. that is not exactly how you or I would discuss such high "chances". What I am saying is the model is also predicting an overall 60% chance of a GOP victory. So, when he says "99%" he is talking about setting conditions in the model that are not yet in place in reality.

Point is, these models work. They are useful. They take a little more understanding than just put a quarter in and get your fortune out the other side.
xus00HAY
Posts: 1,382
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 6:00:02 PM
Posted: 9 months ago
People watch Trump on TV saying "your'e fired".
They believe that he is the kind of guy we need to be the chief executive of the federal government.
So, While Mr. Trump is far from perfect, voting for him sends a message to all federal employees, that is like I think you should have a boss who will fire you.
This is what the voters will be doing election day.
YYW
Posts: 36,286
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 6:00:32 PM
Posted: 9 months ago
At 2/26/2016 5:52:11 PM, dylancatlow wrote:
There are many reasons to be skeptical of his model. First, the near flawless track record is misleading. A model can't receive credit for agreeing with past election results, since all you have to do is find a pattern in the data and construct your model accordingly. Some patterns might reflect underlying forces that make the pattern likely to continue, while others might just be random noise (and in the case of his model, the pattern doesn't even run the whole way through, as it fails to predict the 1960 election). So really, the model only gets points for successfully predicting the last three elections: Clinton, Bush, and Obama. The only one that wasn't obvious was Bush, and the model can't even receive credit for that one. The model bases its prediction on the estimated general election numbers, and of course Bush lost in the 2000 general election yet managed to snatch the presidency anyway, so the model just got lucky.

Second, the model is actually quite unsophisticated. The model's prediction is primarily a function of how well candidates do in the primaries - the relevance to the presidential election is not entirely clear. Trump is doing better than Clinton in the primaries, but so what? Trump could go uncontested in the primaries (which, according to the model, would make him even more likely to win) and still lose to Clinton if he can't win over enough independents. Trump is not a typical candidate. His base of support "runs narrow but deep," as Nate Silver puts it. That means he's likely to run out of steam far sooner than most candidates. Moreover, Trump, as well as the rest of the Republican field, get a major boost because of an assumption built into the model: political parties are less likely to win when they've held office for the previous two terms, which is true historically speaking, but not necessarily true going forward.

You are using anecdotal evidence to rebut an empirical model, which is pretty much the same thing as punching a brick wall with your bare hands. It doesn't work, and it hurts your hand.
Tsar of DDO
YYW
Posts: 36,286
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 6:02:09 PM
Posted: 9 months ago
At 2/26/2016 5:58:25 PM, TBR wrote:
At 2/26/2016 5:46:09 PM, YYW wrote:
At 2/26/2016 5:44:55 PM, TBR wrote:
At 2/26/2016 5:13:26 PM, YYW wrote:
At 2/26/2016 5:10:51 PM, 1harderthanyouthink wrote:
At 2/26/2016 5:05:41 PM, YYW wrote:
At 2/26/2016 4:23:37 PM, 1harderthanyouthink wrote:
He logistically cannot pull in the vote needed to win, unless a large group of Democrats stay at home.

Which they might.

A model can't really account for that.

Actually it can (which I know because I have worked with people who do this for a living), which is the entire point of these models.

A large group of falsely righteous Republicans can stay home, too.

I don't disagree that the model cal work, or that it is accurate for the last 100+ years he says. He has gone back in time, and run the algorithm. How else would he have any basis for the research?

It could fail for this cycle, then he would go back and adjust.

Yeah. The point is that it's the best model out there.

Yea, again, no argument.

I should also point out, when he says "99% chance" etc. that is not exactly how you or I would discuss such high "chances". What I am saying is the model is also predicting an overall 60% chance of a GOP victory. So, when he says "99%" he is talking about setting conditions in the model that are not yet in place in reality.

Point is, these models work. They are useful. They take a little more understanding than just put a quarter in and get your fortune out the other side.

That's exactly right, and the whole idea is to forecast. Sophisticated statistical modeling isn't something that your run of the mill armchair pundit (i.e. DylanCatlow) or even most real pundits (e.g. S.E. Cupp, who is the stupidest person I have ever seen on TV, including competition from such networks as MTV's noted show, Jack@ss) have the intellectual capacity to understand.
Tsar of DDO
YYW
Posts: 36,286
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 6:02:09 PM
Posted: 9 months ago
At 2/26/2016 5:58:25 PM, TBR wrote:
At 2/26/2016 5:46:09 PM, YYW wrote:
At 2/26/2016 5:44:55 PM, TBR wrote:
At 2/26/2016 5:13:26 PM, YYW wrote:
At 2/26/2016 5:10:51 PM, 1harderthanyouthink wrote:
At 2/26/2016 5:05:41 PM, YYW wrote:
At 2/26/2016 4:23:37 PM, 1harderthanyouthink wrote:
He logistically cannot pull in the vote needed to win, unless a large group of Democrats stay at home.

Which they might.

A model can't really account for that.

Actually it can (which I know because I have worked with people who do this for a living), which is the entire point of these models.

A large group of falsely righteous Republicans can stay home, too.

I don't disagree that the model cal work, or that it is accurate for the last 100+ years he says. He has gone back in time, and run the algorithm. How else would he have any basis for the research?

It could fail for this cycle, then he would go back and adjust.

Yeah. The point is that it's the best model out there.

Yea, again, no argument.

I should also point out, when he says "99% chance" etc. that is not exactly how you or I would discuss such high "chances". What I am saying is the model is also predicting an overall 60% chance of a GOP victory. So, when he says "99%" he is talking about setting conditions in the model that are not yet in place in reality.

Point is, these models work. They are useful. They take a little more understanding than just put a quarter in and get your fortune out the other side.

That's exactly right, and the whole idea is to forecast. Sophisticated statistical modeling isn't something that your run of the mill armchair pundit (i.e. DylanCatlow) or even most real pundits (e.g. S.E. Cupp, who is the stupidest person I have ever seen on TV, including competition from such networks as MTV's noted show, Jack@ss) have the intellectual capacity to understand.
Tsar of DDO
TBR
Posts: 9,991
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 6:04:11 PM
Posted: 9 months ago
At 2/26/2016 5:52:11 PM, dylancatlow wrote:
There are many reasons to be skeptical of his model. First, the near flawless track record is misleading. A model can't receive credit for agreeing with past election results, since all you have to do is find a pattern in the data and construct your model accordingly. Some patterns might reflect underlying forces that make the pattern likely to continue, while others might just be random noise (and in the case of his model, the pattern doesn't even run the whole way through, as it fails to predict the 1960 election). So really, the model only gets points for successfully predicting the last three elections: Clinton, Bush, and Obama. The only one that wasn't obvious was Bush, and the model can't even receive credit for that one. The model bases its prediction on the estimated general election numbers, and of course Bush lost in the 2000 general election yet managed to snatch the presidency anyway, so the model just got lucky.

Second, the model is actually quite unsophisticated. The model's prediction is primarily a function of how well candidates do in the primaries - the relevance to the presidential election is not entirely clear. Trump is doing better than Clinton in the primaries, but so what? Trump could go uncontested in the primaries (which, according to the model, would make him even more likely to win) and still lose to Clinton if he can't win over enough independents. Trump is not a typical candidate. His base of support "runs narrow but deep," as Nate Silver puts it. That means he's likely to run out of steam far sooner than most candidates. Moreover, Trump, as well as the rest of the Republican field, get a major boost because of an assumption built into the model: political parties are less likely to win when they've held office for the previous two terms, which is true historically speaking, but not necessarily true going forward.

Not disagreeing with the assessment, but the model has predicted all but the Kennedy election, so...

Failed:
Kennedy

Success:
Nixon
Nixon
Carter
Reagan
Reagan
Bush
Clinton
Clinton
Bush
Bush
Obama
Obama
Cody_Franklin
Posts: 9,483
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 6:04:19 PM
Posted: 9 months ago
What is the model, exactly? I may have glossed over it in the article.

I'm generally pretty suspicious of predictions made in social science circles, so I would be interested to see how this professor seems to have [seemingly spontaneously, from our point of view] come up with this rigorous little machine.
xus00HAY
Posts: 1,382
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 6:10:52 PM
Posted: 9 months ago
C'mon Benchy,
If a professor at the university of Miami said that trump will win, and this guy who works in an elementary school in the everglades bringing the kids to school in an air boat. Which one are you gonna believe?
YYW
Posts: 36,286
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 6:12:56 PM
Posted: 9 months ago
At 2/26/2016 6:04:19 PM, Cody_Franklin wrote:
What is the model, exactly? I may have glossed over it in the article.

I'm generally pretty suspicious of predictions made in social science circles, so I would be interested to see how this professor seems to have [seemingly spontaneously, from our point of view] come up with this rigorous little machine.

Most of social science is bullsh!t, but political modeling isn't. It's one of the few legit areas of political science.

I'm trying to find a PFD of it, but I'll prob. have to go on JSTOR for it. Will do that later.
Tsar of DDO
TBR
Posts: 9,991
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 6:13:38 PM
Posted: 9 months ago
At 2/26/2016 3:57:01 PM, YYW wrote:
http://www.syracuse.com...

President Trump? A SUNY professor says he's "almost certain" it will happen.

Helmut Norpoth, a political science professor at Stony Brook University, says that his statistical model predicted Donald Trump will win the election if he's the Republican nominee. The billionaire businessman has a 97 percent chance of beating Hillary Clinton in the general election and a 99 percent chance of beating Bernie Sanders, according to Norpoth.

The Statesman reports Norpoth uses statistics from candidates' performances in their party's primary and patterns in the electoral cycle to forecast national election results. The equation correctly predicted Bill Clinton's victory in 1996, George W. Bush in 2004, and Barack Obama in 2012.

In fact, Norpoth's model has correctly predicted the outcome of every presidential election since 1912 except one, according to the Daily Mail. In 1960, John F. Kennedy beat Richard Nixon in one of the closest races in U.S. history.

"When I started out with this kind of display a few months ago, I thought it was sort of a joke," Norpoth said at a gathering Monday at the SUNY Global Center in Manhattan. "Well, I'll tell you right now, it ain't a joke anymore."

He added that the Republican party overall has a 61 percent chance of putting a GOP candidate in the White House, but Clinton has a 55 percent chance of winning in a hypothetical race with Cruz or Marco Rubio.

But Trump's success in the primaries, winning three out of four GOP contests to date, has made him the favorite by far in Norpoth's model -- especially if he gets the Republican nomination.

"The probability of that [outcome] is almost complete certainty, 97 percent. It's almost 'Take it to the bank,'" Norpoth said.

Just for kicks, I looked him up on rate my professor
http://www.ratemyprofessors.com...

He has a A grade with 3.7 for quality. I mention this because my brother on the same site gets an A and 4.8 with a flaming red pepper for "hotness"
YYW
Posts: 36,286
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 6:14:24 PM
Posted: 9 months ago
At 2/26/2016 6:13:38 PM, TBR wrote:
At 2/26/2016 3:57:01 PM, YYW wrote:
http://www.syracuse.com...

President Trump? A SUNY professor says he's "almost certain" it will happen.

Helmut Norpoth, a political science professor at Stony Brook University, says that his statistical model predicted Donald Trump will win the election if he's the Republican nominee. The billionaire businessman has a 97 percent chance of beating Hillary Clinton in the general election and a 99 percent chance of beating Bernie Sanders, according to Norpoth.

The Statesman reports Norpoth uses statistics from candidates' performances in their party's primary and patterns in the electoral cycle to forecast national election results. The equation correctly predicted Bill Clinton's victory in 1996, George W. Bush in 2004, and Barack Obama in 2012.

In fact, Norpoth's model has correctly predicted the outcome of every presidential election since 1912 except one, according to the Daily Mail. In 1960, John F. Kennedy beat Richard Nixon in one of the closest races in U.S. history.

"When I started out with this kind of display a few months ago, I thought it was sort of a joke," Norpoth said at a gathering Monday at the SUNY Global Center in Manhattan. "Well, I'll tell you right now, it ain't a joke anymore."

He added that the Republican party overall has a 61 percent chance of putting a GOP candidate in the White House, but Clinton has a 55 percent chance of winning in a hypothetical race with Cruz or Marco Rubio.

But Trump's success in the primaries, winning three out of four GOP contests to date, has made him the favorite by far in Norpoth's model -- especially if he gets the Republican nomination.

"The probability of that [outcome] is almost complete certainty, 97 percent. It's almost 'Take it to the bank,'" Norpoth said.

Just for kicks, I looked him up on rate my professor
http://www.ratemyprofessors.com...

He has a A grade with 3.7 for quality. I mention this because my brother on the same site gets an A and 4.8 with a flaming red pepper for "hotness"

lol right on
Tsar of DDO
Cody_Franklin
Posts: 9,483
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 6:16:50 PM
Posted: 9 months ago
At 2/26/2016 6:12:56 PM, YYW wrote:
At 2/26/2016 6:04:19 PM, Cody_Franklin wrote:
What is the model, exactly? I may have glossed over it in the article.

I'm generally pretty suspicious of predictions made in social science circles, so I would be interested to see how this professor seems to have [seemingly spontaneously, from our point of view] come up with this rigorous little machine.

Most of social science is bullsh!t, but political modeling isn't. It's one of the few legit areas of political science.

I'm trying to find a PFD of it, but I'll prob. have to go on JSTOR for it. Will do that later.

Please do. I thank the Lord Jehovah Yahweh God that I know enough about the mechanics of statistical analysis to squint my eyes knowingly at complicated-looking models. I've got an open offer on a bet of not-insignificant consequence concerning the results of the coming election, and, if this model is as good as competent people in this thread claim it to be, it will be a great asset.
xus00HAY
Posts: 1,382
Add as Friend
Challenge to a Debate
Send a Message
2/26/2016 6:18:47 PM
Posted: 9 months ago
"but the model has predicted all but the Kennedy election, so..."

The model did not predict Johnson vs Goldwater in '1964