Trump jumps in Rasmussen Poll
-
@George-K said in Trump jumps in Rasmussen Poll:
@jon-nyc said in Trump jumps in Rasmussen Poll:
Nate Silver tweeted yesterday “Guess which polling outfit has Trump leading nationally”?
<past performance is no guarantee of future results>
How did Nate's predictions work out, oh, about 4 years ago?
https://projects.fivethirtyeight.com/2016-election-forecast/
In all fairness, Nate was saying a week or so before the election that something was off with the numbers and Democrats should be somewhat careful. He then got lambasted...
@LuFins-Dad said in Trump jumps in Rasmussen Poll:
Nate was saying a week or so before the election
That link is from November 8 - Election day.
"Now look here, you Baltic gas passer... " - Mik, 6/14/08
The saying, "Lite is just one damn thing after another," is a gross understatement. The damn things overlap.
-
@jon-nyc said in Trump jumps in Rasmussen Poll:
Nate Silver tweeted yesterday “Guess which polling outfit has Trump leading nationally”?
Rasmussen is consistent. So an uptick like at 53% vs mid to high 40’s for a long time has meaning especially in the brutal week Trump seems to have suffered.
Also the political scientist from Stonybrook who’s model predicted Trump at 80% chance in 2016 has it at over 90% this time. It is based on primary data from 1918 and has been right I think 90% of the time.
-
@jon-nyc said in Trump jumps in Rasmussen Poll:
Nate Silver tweeted yesterday “Guess which polling outfit has Trump leading nationally”?
Rasmussen is consistent. So an uptick like at 53% vs mid to high 40’s for a long time has meaning especially in the brutal week Trump seems to have suffered.
Also the political scientist from Stonybrook who’s model predicted Trump at 80% chance in 2016 has it at over 90% this time. It is based on primary data from 1918 and has been right I think 90% of the time.
@Loki said in Trump jumps in Rasmussen Poll:
the political scientist from Stonybrook who’s model predicted Trump at 80% chance in 2016 has it at over 90% this time
I remember seeing this guy a while ago, but I thought he had backtracked his prediction saying that Trump will lose based on economy, etc.
Got a link for me?
-
-
There’s a bunch of guys that every cycle. They invent an overfitted model and brag about its post-dictive abilities and assume it must be predictive.
Only non-witches get due process.
- Cotton Mather, Salem Massachusetts, 1692
-
@Loki said in Trump jumps in Rasmussen Poll:
the political scientist from Stonybrook who’s model predicted Trump at 80% chance in 2016 has it at over 90% this time
I remember seeing this guy a while ago, but I thought he had backtracked his prediction saying that Trump will lose based on economy, etc.
Got a link for me?
@George-K said in Trump jumps in Rasmussen Poll:
@Loki said in Trump jumps in Rasmussen Poll:
the political scientist from Stonybrook who’s model predicted Trump at 80% chance in 2016 has it at over 90% this time
I remember seeing this guy a while ago, but I thought he had backtracked his prediction saying that Trump will lose based on economy, etc.
Got a link for me?
https://patch.com/new-york/threevillage/stony-brook-professor-predicts-president-trump-landslide-win
-
There’s a bunch of guys that every cycle. They invent an overfitted model and brag about its post-dictive abilities and assume it must be predictive.
@jon-nyc said in Trump jumps in Rasmussen Poll:
There’s a bunch of guys that every cycle. They invent an overfitted model and brag about its post-dictive abilities and assume it must be predictive.
OK, Jon. Here's the deal.
If Biden wins, you can pummel some of us daily for 1 year, and we will type "you were right, Jon!" every day in response.If Trump wins, you buy that Vette and wear a MAGA hat for one week per month, with the Vette top down.
And, you have to drive to wherever Phibes lives, and listen to him on any subject for an hour. -
There’s a bunch of guys that every cycle. They invent an overfitted model and brag about its post-dictive abilities and assume it must be predictive.
@jon-nyc said in Trump jumps in Rasmussen Poll:
There’s a bunch of guys that every cycle. They invent an overfitted model and brag about its post-dictive abilities and assume it must be predictive.
On that, I would agree.
OTOH, something must be going on in some of the internal polls. I see some changes being made in both campaigns...
-
@jon-nyc said in Trump jumps in Rasmussen Poll:
There’s a bunch of guys that every cycle. They invent an overfitted model and brag about its post-dictive abilities and assume it must be predictive.
On that, I would agree.
OTOH, something must be going on in some of the internal polls. I see some changes being made in both campaigns...
-
@jon-nyc said in Trump jumps in Rasmussen Poll:
There’s a bunch of guys that every cycle. They invent an overfitted model and brag about its post-dictive abilities and assume it must be predictive.
OK, Jon. Here's the deal.
If Biden wins, you can pummel some of us daily for 1 year, and we will type "you were right, Jon!" every day in response.If Trump wins, you buy that Vette and wear a MAGA hat for one week per month, with the Vette top down.
And, you have to drive to wherever Phibes lives, and listen to him on any subject for an hour.@Rainman said in Trump jumps in Rasmussen Poll:
@jon-nyc said in Trump jumps in Rasmussen Poll:
There’s a bunch of guys that every cycle. They invent an overfitted model and brag about its post-dictive abilities and assume it must be predictive.
OK, Jon. Here's the deal.
And, you have to drive to wherever Phibes lives, and listen to him on any subject for an hour.
What did I do to have to talk to Jon for a freaking hour?
-
@Rainman said in Trump jumps in Rasmussen Poll:
@jon-nyc said in Trump jumps in Rasmussen Poll:
There’s a bunch of guys that every cycle. They invent an overfitted model and brag about its post-dictive abilities and assume it must be predictive.
OK, Jon. Here's the deal.
And, you have to drive to wherever Phibes lives, and listen to him on any subject for an hour.
What did I do to have to talk to Jon for a freaking hour?
@Doctor-Phibes said in Trump jumps in Rasmussen Poll:
@Rainman said in Trump jumps in Rasmussen Poll:
@jon-nyc said in Trump jumps in Rasmussen Poll:
There’s a bunch of guys that every cycle. They invent an overfitted model and brag about its post-dictive abilities and assume it must be predictive.
OK, Jon. Here's the deal.
And, you have to drive to wherever Phibes lives, and listen to him on any subject for an hour.
What did I do to have to talk to Jon for a freaking hour?
Punishment.
-
@jon-nyc said in Trump jumps in Rasmussen Poll:
There’s a bunch of guys that every cycle. They invent an overfitted model and brag about its post-dictive abilities and assume it must be predictive.
OK, Jon. Here's the deal.
If Biden wins, you can pummel some of us daily for 1 year, and we will type "you were right, Jon!" every day in response.If Trump wins, you buy that Vette and wear a MAGA hat for one week per month, with the Vette top down.
And, you have to drive to wherever Phibes lives, and listen to him on any subject for an hour.@Rainman said in Trump jumps in Rasmussen Poll:
@jon-nyc said in Trump jumps in Rasmussen Poll:
There’s a bunch of guys that every cycle. They invent an overfitted model and brag about its post-dictive abilities and assume it must be predictive.
OK, Jon. Here's the deal.
If Biden wins, you can pummel some of us daily for 1 year, and we will type "you were right, Jon!" every day in response.If Trump wins, you buy that Vette and wear a MAGA hat for one week per month, with the Vette top down.
And, you have to drive to wherever Phibes lives, and listen to him on any subject for an hour.NFWLolOnly non-witches get due process.
- Cotton Mather, Salem Massachusetts, 1692
-
@LuFins-Dad said in Trump jumps in Rasmussen Poll:
Nate was saying a week or so before the election
That link is from November 8 - Election day.
@George-K said in Trump jumps in Rasmussen Poll:
@LuFins-Dad said in Trump jumps in Rasmussen Poll:
Nate was saying a week or so before the election
That link is from November 8 - Election day.
https://www.vox.com/2016/11/3/13147678/nate-silver-fivethirtyeight-trump-forecast
-
There’s a bunch of guys that every cycle. They invent an overfitted model and brag about its post-dictive abilities and assume it must be predictive.
@jon-nyc said in Trump jumps in Rasmussen Poll:
There’s a bunch of guys that every cycle. They invent an overfitted model and brag about its post-dictive abilities and assume it must be predictive.
Yep. LOL
A friend once quit work to start a precious metals trading company. He had a model that was going to work and give great returns. I invested a small amount with him, knowing (and figuring) that I could lose it, but that was okay. After a couple of years, I got my exact amount of money back, but did not make any. He admitted that the model was constantly being updates to fit past history, but that was not enough to predict the future. LOL
-
There’s two candidates, the elections only happen once every 4 years.
There’s a shitload of prognosticators. If you have enough of them, a portion of them will have a perfect prediction record through sheer chance.
We see the patterns we want to see.
-
@LuFins-Dad said in Trump jumps in Rasmussen Poll:
Nate was saying a week or so before the election
That link is from November 8 - Election day.
@George-K said in Trump jumps in Rasmussen Poll:
@LuFins-Dad said in Trump jumps in Rasmussen Poll:
Nate was saying a week or so before the election
That link is from November 8 - Election day.
Trump had a 30% chance and won. A 30% chance is a very good chance.
-
There’s two candidates, the elections only happen once every 4 years.
There’s a shitload of prognosticators. If you have enough of them, a portion of them will have a perfect prediction record through sheer chance.
We see the patterns we want to see.
@xenon said in Trump jumps in Rasmussen Poll:
There’s two candidates, the elections only happen once every 4 years.
There’s a shitload of prognosticators. If you have enough of them, a portion of them will have a perfect prediction record through sheer chance.
We see the patterns we want to see.
Not too difficult when it's 50/50.
-
There’s two candidates, the elections only happen once every 4 years.
There’s a shitload of prognosticators. If you have enough of them, a portion of them will have a perfect prediction record through sheer chance.
We see the patterns we want to see.
@xenon said in Trump jumps in Rasmussen Poll:
There’s two candidates, the elections only happen once every 4 years.
There’s a shitload of prognosticators. If you have enough of them, a portion of them will have a perfect prediction record through sheer chance.
We see the patterns we want to see.
It’s not even that. It’s post-diction not pre-diction. They curve fit with full hindsight and choose coefficients and variables until all previous results are matched then pat themselves on the back for a model that “predicted” all previous elections.
