GDT: 11/3/20, Election Day, Donald J. Trump vs. Joseph R. Biden

Status
Not open for further replies.
striker42 said:
The live election blog he's been running has been desperately trying to pull out ways his site was right. The model sucked and it should have been clear it was going to be unreliable. This election was like nothing we've seen before with the uniqueness of a Trump re-election campaign and the massive absentee voting.
Click to expand...

Except for Florida, what did he get super wrong? A few house seats?
 
msstate7 said:
Lol, the polls were horrible. Did you factor in house and senate too?
Click to expand...

Senate on the no tossups was 50/50

Which probably is wrong but isn't that far off.

House they don't do a no tossup map. But they had 43 seats as tossups. They probably screwed some up because I don't want to verify everyone, I don't think they were that wrong.
 
zitothebrave said:
What's interesting is everyone is jumping on the pollsters. If you look at RCP's average polls and see their no tossups, they have 319 to 219 if the results hold as they are it will be 306 to 232. Not that far off the mark. The off the mark was Florida, but after the chst pumping of how wrong the pollsters were, the vote closed a lot. I think Ohio will be the same.
Click to expand...

The polls were wrong in that they had Biden with huge leads in a lot of these states. That's simply not the case.
 
zitothebrave said:
Except for Florida, what did he get super wrong? A few house seats?
Click to expand...

A lot of house seats and the Senate. He called a 75% chance Democrats take the Senate and seemed more sure than that.

He also had Biden way farther ahead in a lot of states. The electoral vote count might not end up being the surprise, the margin of victory in a lot of states is a huge surprise.
 
I get the battle ground states were close percentage wise, but it looks like Biden is gonna end up with 306 electoral votes and Trump with 216... when you add the fact that Biden is winning the popular vote by almost 4 million votes (which will only get bigger)... he has won the election pretty convincingly
 
thewupk said:
The polls were wrong in that they had Biden with huge leads in a lot of these states. That's simply not the case.
Click to expand...

But that's why polls have Margin of Error. Average poll has a 2-3 point margin of error.

Let's look big picture. RCP average for the race was Biden 51.2 Trump 44. As of right now Biden is at 50.5 and Trump at 47.7. Of course the average doesn't account for people who are unsure who're they're voting for or for 3rd party, so if we split the undecided minus the votes actually cast for 3rd party (which tend to be overrepresented in polls) you have Biden at 52.7 and Trump at 45.5 which is well within the margin of error.
 
thewupk said:
The polls were wrong in that they had Biden with huge leads in a lot of these states. That's simply not the case.
Click to expand...

FL, OH, and WI seem like the worst culprits, though votes are still being tallied in OH. Maybe PA too as Biden seems to have had at least 4 point advantage in most polls. Most of the other states look to have ended up pretty good, at least on POTUS.
 
Last edited:
For states that are going to Biden...

Wisconsin ended up being a 0.3 percentage point difference. Nate Silver had it at over 8.3.

Michigan is a 2.7 percentage point difference. Nate Silver had it at 8.0.

Pennsylvania is currently at 0.1 percentage point difference and will certainly end less than percentage point difference. Nate Silver had it at 4.7.

Nevada will probably end up around 1 percentage point difference. Nate Silver had it at 6.1.


For states that are going to Trump...

Ohio is 8.2 percentage points in Trump's favor. Nate Silver had it at 0.6

Iowa is also 8.2 percentage points in Trump's favor. Nate Silver had it at 1.5

Texas is 5.8 percentage points in Trump's favor. Nate Silver had it at 1.5


So the model definitely predicted Biden performing much better in the majority of battleground states than he did. The only places he was really close to accurate that were battlegrounds were Georgia and Arizona. In none of the states did he predict Trump doing better than he did. So the model was definitely wrong.
 
striker42 said:
A lot of house seats and the Senate. He called a 75% chance Democrats take the Senate and seemed more sure than that.

He also had Biden way farther ahead in a lot of states. The electoral vote count might not end up being the surprise, the margin of victory in a lot of states is a huge surprise.
Click to expand...

You do realize that 75% isn't 100% right? Also realize that part of that was the high probability of Biden winning. Which if the Votes come in right and the dems win 2 of the 3 non-called seats, they flipped the senate.

The logic was R seats were more vulnerable than D seats, which is certainly true as they were many risky seats.
 
Well Pelosi threw out the word "mandate" in reference to Biden's election. I don't think even Biden would go that far. Honestly that's the last word that needs to be thrown around. After a tumultuous Trump presidency and an bitter election I think the word "reconciliation" needs to be used. I think that's what Biden is doing.
 
striker42 said:
For states that are going to Biden...

Wisconsin ended up being a 0.3 percentage point difference. Nate Silver had it at over 8.3.

Michigan is a 2.7 percentage point difference. Nate Silver had it at 8.0.

Pennsylvania is currently at 0.1 percentage point difference and will certainly end less than percentage point difference. Nate Silver had it at 4.7.

Nevada will probably end up around 1 percentage point difference. Nate Silver had it at 6.1.


For states that are going to Trump...

Ohio is 8.2 percentage points in Trump's favor. Nate Silver had it at 0.6

Iowa is also 8.2 percentage points in Trump's favor. Nate Silver had it at 1.5

Texas is 5.8 percentage points in Trump's favor. Nate Silver had it at 1.5


So the model definitely predicted Biden performing much better in the majority of battleground states than he did. The only places he was really close to accurate that were battlegrounds were Georgia and Arizona. In none of the states did he predict Trump doing better than he did. So the model was definitely wrong.
Click to expand...

But this is an example of again, of margin of error being not taken into effect.

Don't get me wrong. I don't think Silver did a great job. I think he didn't properly calculate the 2016 error. But I don't think the polls were as far off as most people think, except for Florida.
 
zitothebrave said:
But this is an example of again, of margin of error being not taken into effect.

Don't get me wrong. I don't think Silver did a great job. I think he didn't properly calculate the 2016 error. But I don't think the polls were as far off as most people think, except for Florida.
Click to expand...

Simply falling within the margin of error doesn't automatically eliminate problems with the polls. If the polls were off simply because of margin of error then you would expect them to be off both ways. Some would have Trump doing better, others would have Biden doing better. It would be random as they would be flukes of random sampling.

However, what we saw is the polls consistently were off in favor of Biden. And not off by small amounts. Off in amounts that stretch the margin of error to its limit. When you see polls consistently off in one direction there's a good chance something was wrong with the polls.
 
Status
Not open for further replies.
Back
Top