Nearly 60 percent Want Trump Prosecuted

[classhandicapper]

The FBI is politically corrupt. Almost everything they say or do must be considered biased, spun, an outright lie, or partial truth.

[PaceAdvantage]

Originally Posted by Actor

War is "an intense armed conflict between states, governments, societies, or paramilitary groups such as mercenaries, insurgents, and militias." Those participating in the Jan 6 attack on the Capital were, at the very least, insurgents. We all heard Trump tell them to "fight like hell." Ergo, Trump was making war upon the United States and thereby committing treason according to the Constitutional definition. No matter how many you post that fact will not change.

And you have made it quite clear that you approve of the attack on the Capital.

I'm happy your delusions are on full display for all to see and enjoy for years to come.

[Elkchester Road]

Originally Posted by PaceAdvantage

I'm happy your delusions are on full display for all to see and enjoy for years to come.

Exactly.

You'd think this Election Cycle is going bad for Democrats...or something. Straight up empty-headed Lunacy.

[stuball]

These guys should really have their heads split open to let Trump out to get some air --he has been stuck in there for a very long time

[Inner Dirt]

Originally Posted by stuball

These guys should really have their heads split open to let Trump out to get some air --he has been stuck in there for a very long time

[cratman]

Originally Posted by Jeff P

Link to the ABC News/Ipsos poll:
https://www.ipsos.com/en-us/news-pol...t-following-it



According to the Statista website there were approximately 168.31 million registered voters in the US during 2020.

Doing some quick math:

545/168,310,000 = 3.238072604e-6

Or, maybe this way is easier to visualize:

545/168,310,000 = 0.000003238072604

There are approximately 35,800 thoroughbred races in my 2021 database.

(35,800) x (3.238072604e-6) = 0.11592

or:

(35,800 races) x (3.238072604e-6) = 0.11592 races

Fyi, 0.11592 races is roughly the same as a single horse from a 9 horse field.

Imo, the statement by the authors of the USA Today article "Almost six out of 10 Americans want former President Donald Trump criminally charged in connection with the Jan. 6, 2021, Capitol riot" has about the same statistical accuracy I'd have if I created a model for betting thoroughbreds after randomly selecting a single horse from a 9 horse field and using that as my development sample.

It might be an interesting exercise.

But the sample size is way too small to make accurate unbiased predictions about future events.



-jp

.

It would be useful for those who seek success betting on races to have some understanding of statistics.
http://www.wral.com/they-asked-only-...ple-/13507173/

[PaceAdvantage]

Originally Posted by cratman

It would be useful for those who seek success betting on races to have some understanding of statistics.
http://www.wral.com/they-asked-only-...ple-/13507173/

[cratman]

Originally Posted by Jeff P

Link to the ABC News/Ipsos poll:
https://www.ipsos.com/en-us/news-pol...t-following-it



According to the Statista website there were approximately 168.31 million registered voters in the US during 2020.

Doing some quick math:

545/168,310,000 = 3.238072604e-6

Or, maybe this way is easier to visualize:

545/168,310,000 = 0.000003238072604

There are approximately 35,800 thoroughbred races in my 2021 database.

(35,800) x (3.238072604e-6) = 0.11592

or:

(35,800 races) x (3.238072604e-6) = 0.11592 races

Fyi, 0.11592 races is roughly the same as a single horse from a 9 horse field.

Imo, the statement by the authors of the USA Today article "Almost six out of 10 Americans want former President Donald Trump criminally charged in connection with the Jan. 6, 2021, Capitol riot" has about the same statistical accuracy I'd have if I created a model for betting thoroughbreds after randomly selecting a single horse from a 9 horse field and using that as my development sample.

It might be an interesting exercise.

But the sample size is way too small to make accurate unbiased predictions about future events.



-jp

.

It would be useful for those who seek success betting on races to have some understanding of statistics.
http://www.wral.com/they-asked-only-...ple-/13507173/

[PaceAdvantage]

why you post the same thing twice bro?

[Jeff P]

Actually, he has a point.

Creating a poll that measures public opinion and modeling the likelihood of a horse winning a race are two completely different things.

Just as the authors contend in the article he cited:

Poll accuracy does see diminishing returns as sample size increases.

So I admit I used a bad argument in my post.

That said:

I think the vast majority of Americans have moved on from January 6th, and I don't believe for one second that 60% of Americans actually want to see Trump criminally charged.

I also think a sample size of 545 is way too small for the ABC New/Ipsos poll USA Today used as the basis for their article.

Asking likely voters who they plan on voting for: CandidateA or CandidateB is one thing.

But asking about criminal charges for a polarizing figure like Trump has a high likelihood of eliciting a politically charged emotional response.

Getting an accurate read on something like that requires a much larger sample size.

That's just my opinion.


-jp

.

[cratman]

Sample bias may be more of an issue than sample size. Supporters of Trump may distrust polls and therefore be reluctant to participate. This could be the reason so many of the 2020 polls were off. It is probably very difficult for pollsters to control for an unknown group who decline to participate.
A classic example of sample bias were the Dewey Truman polls only of those who had telephones (the rich of their time).

[king kong]

Putting America first with 2.oo gas and very little inflation is OK with me!!!

[tucker6]

Originally Posted by cratman

Sample bias may be more of an issue than sample size. Supporters of Trump may distrust polls and therefore be reluctant to participate. This could be the reason so many of the 2020 polls were off. It is probably very difficult for pollsters to control for an unknown group who decline to participate.
A classic example of sample bias were the Dewey Truman polls only of those who had telephones (the rich of their time).

The greatest issue relating to poll accuracy is the formulation of the question itself. Something as mundane as the changing of sentence structure of a question (no change in actual words) can have a meaningful change in results.