538 is sensationalist af, no longer trust it

S

Saiyajin

Superstar
Joined
Nov 3, 2015
Messages
10,081
Reputation
3,365
Daps
54,030
I used to go on that site and assumed it was trustworthy, but now I can see it has purposely skewed the data to make the race appear much closer, and seeing it's election progress indicate when it has last been updated I am 100% they want to increase revenue by making people come back and see the probabilities of a Trump win over and over as the election draws closer.

I found a much better source that is non profit, a run by a man who seems to know what he's talking about based on his past predictions. Princeton Election Consortium — A first draft of electoral history. Since 2004

In 2004, Wang was one of the first to aggregate US Presidential polls using probabilistic methods.[10] The method's applications included correct Election-Eve predictions, high-resolution tracking of the race during the campaign, and identification of targets for resource allocation. Wang's calculation, based on polls only, ended up precisely at the actual electoral outcome, Bush 286, Kerry 252 EV. In 2008, Wang and Andrew Ferguson founded the Princeton Election Consortium blog, in which he analyzes U.S. national election polling.[11][12] His statistical analysis in 2012 correctly predicted the presidential vote outcome in 49 of 50 states and even the popular vote outcome of Barack Obama's 51.1% to Mitt Romney's 48.9%.[13] That year, the Princeton Election Consortium also correctly called 10 out of 10 close Senate races and came within a few seats of the final House outcome
Click to expand...

The source has given Hillary a 99%+ chance of winning while 538 has had Hillary's chance go from 91% to 66 within 1-2 weeks.

I've also noticed that he has every swing state basically 50/50 that never changes. He has almost every swing state shaded in to indicate the state as slightly in favor of Trump which seems odd when it should be left blank it is pointless to differentiate.

Nevada, Florida, NC, for example are extremely close I wanna see how the actual results line up to the predictions he's given.

I no longer trust 538 for election related info :francis:
 
Last edited:
S

Saiyajin

Superstar
Joined
Nov 3, 2015
Messages
10,081
Reputation
3,365
Daps
54,030
Sam Wang also correctly predicted Trump's nomination

but 538 ignored the data and wrote shyt like this

Donald Trump’s Six Stages Of Doom

:francis:

Afterwards Sam correctly predicted that Trump would easily get the delegates needed while the rest of the world was pushing the narrative that he would have the highest # of delegates but not enough Does Trump’s ceiling matter?
 
Last edited:
N

No1

Retired.
Supporter
Joined
Apr 30, 2012
Messages
32,198
Reputation
5,472
Daps
73,212
Professor Emeritus

Professor Emeritus

Veteran
Poster of the Year
Supporter
Joined
Jan 5, 2015
Messages
51,331
Reputation
19,940
Daps
204,141
Reppin
the ether
All of you propping up Sam Wang need to reconcile the fact that he was wildly off in 2010 and 2014.

In 2008 and 2012, Wang and Silver were about equally accurate.

So by pure history, you have to say that Silver has the better track record.


Silver's methodology believes in more variance than Wang's does, and thus a greater chance for underdogs. Wang believes in a narrower chance for underdogs and thus is more "certain" in his predictions, which has led him to errors so great that in 2010, the actual Republican results were something that Wang had pegged at something like trillions to one, or basically impossible.


Yes, Silver has had some inaccurate punditry regarding Trump. But that had to do with his own personal opinions, not his polling model.
 
hashmander

hashmander

Hale End
Supporter
Joined
Jan 17, 2013
Messages
21,252
Reputation
5,577
Daps
91,227
Reppin
The Arsenal
The Dankster said:
All of you propping up Sam Wang need to reconcile the fact that he was wildly off in 2010 and 2014.

In 2008 and 2012, Wang and Silver were about equally accurate.

So by pure history, you have to say that Silver has the better track record.


Silver's methodology believes in more variance than Wang's does, and thus a greater chance for underdogs. Wang believes in a narrower chance for underdogs and thus is more "certain" in his predictions, which has led him to errors so great that in 2010, the actual Republican results were something that Wang had pegged at something like trillions to one, or basically impossible.


Yes, Silver has had some inaccurate punditry regarding Trump. But that had to do with his own personal opinions, not his polling model.
Click to expand...
then sam will get kicked to the bushes in 2018. fortunately this is a presidential election year. maybe his methods don't work as well in poor quality polling years like you get in off-year elections.

anyway nate got a taste of that corporate money. gotta keep it flowing. the good thing about all this is that everyone gets a light shined on them after the election.
 
Professor Emeritus

Professor Emeritus

Veteran
Poster of the Year
Supporter
Joined
Jan 5, 2015
Messages
51,331
Reputation
19,940
Daps
204,141
Reppin
the ether
The Dankster said:
All of you propping up Sam Wang need to reconcile the fact that he was wildly off in 2010 and 2014.

In 2008 and 2012, Wang and Silver were about equally accurate.

So by pure history, you have to say that Silver has the better track record.


Silver's methodology believes in more variance than Wang's does, and thus a greater chance for underdogs. Wang believes in a narrower chance for underdogs and thus is more "certain" in his predictions, which has led him to errors so great that in 2010, the actual Republican results were something that Wang had pegged at something like trillions to one, or basically impossible.
Click to expand...


I think I can fairly give an "I told you so" here. Silver was right, and the Princeton Election Consortium ain't nothing.


Not much condolence on this ugly night, of course. :huhldup:
 
You must log in or register to reply here.
Top