The Off Topic Topic

[Valmy]

I am just curious why this is considered some kind of important new trend.

Seriously? You jumped into defending the polls which is a strange way to express that curiousity

Asking a question is a weird way to express curiosity? Really?

They predicted Lula would get the most votes and he did. Was there ever a point in history where polls were consistently nailing every election down exactly? Because I don't recall this.

It is right there. I asked the question. Generally all I think the polls have ever been good for is just kind of knowing generally how this election is trending. Is it going to be close? And if not who is favored? So it said Lula should get the most votes and by a decent margin. Well five points is not super close. Yeah they predicted a bigger margin but as you pointed out it wasn't completely outside every polls result. Bottom line is that is seemed to give us a general picture of how the election was supposed to go. But maybe I was wrong. Maybe in the past we did get extremely predictive results. Hence why I asked the question.

I pointed out that the largest sample size poll had Lula winning by 14% and if you have a 9% margin of error that makes the poll completely meaningless. Anyone knowledgeable about Brazil should have been able to guess the results to within 18% without the effort of getting responses from over 12,000 people (and considering response rates they probably asked a multiple of 12,000 people).

There has been a lot of literature on recent polling challenges, but if you want something right at hand...that link I gave above from 538 has the polling errors for 4 types of elections (president, governor, senate, and house) in the most 12 recent cycles - that is 48 total data points. Since 2010, in 15 of the 18 data points the results of the election have exceeded the average for the type of election. 2020 the biggest miss since 2000 in every category. In the US context, it does look like polling quality is deteriorating.

The second biggest miss was 1998 and the 2018 cycle was one of the closer ones. It looks like the error is inconsistent and all over the damn place. And without data from before 1998 how would this be data that indicates it deteriorating? It looks likes it is consistently inconsistent and we could never count on it doing much more than giving us a general idea.

[alfred russel]

I am just curious why this is considered some kind of important new trend.

Seriously? You jumped into defending the polls which is a strange way to express that curiousity

Asking a question is a weird way to express curiosity? Really?

They predicted Lula would get the most votes and he did. Was there ever a point in history where polls were consistently nailing every election down exactly? Because I don't recall this.

It is right there. I asked the question. Generally all I think the polls have ever been good for is just kind of knowing generally how this election is trending. Is it going to be close? And if not who is favored? So it said Lula should get the most votes and by a decent margin. Well five points is not super close. Yeah they predicted a bigger margin but as you pointed out it wasn't completely outside every polls result. Bottom line is that is seemed to give us a general picture of how the election was supposed to go. But maybe I was wrong. Maybe in the past we did get extremely predictive results. Hence why I asked the question.

You are defensive of the polling for reasons I don't get. The way that you are asking the question presupposes that the results somehow vindicate the polling...I mean the consensus predictions across all polss as reflected by the aggregations by CNN and the economist was 8-9% off. The fact that there was a poll somewhere off just 3% doesn't mean the polls were good. There is going to be randomness in individual polls conducted at points in time--which is why aggregation gets used.

If you are surveying a population and coming up with an expected percentage, the measure of success is the percentage. Imagine if before the 2020 election, the aggregation of polls said that Biden was going to win the popular vote by 8-9%, and in the event Trump won the election in an electoral landslide and narrowly carried the popular vote. That is the magnitude of the miss that we just saw.

I don't really have time to dig through papers right now that go into the problems that plummeting response rates and call screenings are having on polling, but maybe later I'll post them.

In the meantime you can read about the 2021 Peruvian presidential election, in which the current president was totally off the radar of all the pollsters but came out first in round 1 (polling did much better in round 2).

The consensus of the polls

[grumbler]

What is fairly interesting from Brazil is it is another case of a polling miss in a major election. Sort of like in the last US election, the expected blowout didn't happen but the favorite still won (or likely will win), but it is something else to shake confidence in polling.

Lula got pretty much exactly the number of votes predicted.  The polls undercounted the Bolsanaro voters because, like Trump voters, they think that pollsters are all tools of the devil and so they lie or refuse to respond to polls.  The pollsters even noted that the Bolsanaro poll count was short of the real count, but by an unknown margin.

[Eddie Teach]

TIL there are owls that live underground and mimic rattlesnakes. 

[alfred russel]

What is fairly interesting from Brazil is it is another case of a polling miss in a major election. Sort of like in the last US election, the expected blowout didn't happen but the favorite still won (or likely will win), but it is something else to shake confidence in polling.

Lula got pretty much exactly the number of votes predicted.  The polls undercounted the Bolsanaro voters because, like Trump voters, they think that pollsters are all tools of the devil and so they lie or refuse to respond to polls.  The pollsters even noted that the Bolsanaro poll count was short of the real count, but by an unknown margin.

Are you basing that on analysis out of brazil or just talking out of your ass? It doesn't seem logically consistent to me: if Bolsanaro voters refused to answer the pollsters, why would Lula's votes be correct? If 48% of people want to vote for Lula and told pollsters that, and Bolsanaro voters to an extent refused to answer the polls and thus removed themselves from the sample, Lula would poll higher than 48%. Similarly if Bolsanaro voters decided to lie, unless they universally decided not to say they were voting for Lula and instead chose random people, wouldn't Lula also be overstated?

What people actually knowledgeable about Brazil have speculated is that in working class neighborhoods, which are Lula strongholds and neighbors can hear phone conversations because of the close proximity of housing, people were reluctant to announce their support for Bolsanaro because they didn't want neighbors to know.

[Zanza]

The Asiatic Winter Games in 2029 will be held in Saudi Arabia. 

[The Brain]

The Asiatic Winter Games in 2029 will be held in Saudi Arabia. 

Asia is a different country; they do things differently there.

[Duque de Bragança]

The Asiatic Winter Games in 2029 will be held in Saudi Arabia. 

Must be Allah's will. 

[Jacob]

Scandalous!

Professional Walleye Anglers Caught Stuffing Fish with Weights During a Tournament on Lake Erie
In front of a crowd of fans and other competitors, Chase Cominsky and Jake Runyon were busted for stuffing walleyes with lead weights

A professional walleye fishing tournament on Lake Erie ended in pandemonium last Friday when the leading team was caught cheating during the weigh-in. Professional anglers Chase Cominksy, of Hermitage, Pennsylvania, and Jake Runyon, of Cleveland, Ohio, were declared the winners of the Walleye "Fall Brawl," until tournament director Jason Fischer became suspicious of their bulging walleye, which seemed unusually heavy for their size. Cutting open the fish in front of the crowd, Fischer discovered that each of them had been stuffed with lead weights.

"We got weights in fish!" Fischer yelled when he pulled the first 12-ounce egg sinker out of the team's winning batch.

...

The recorded footage shows that in addition to lead sinkers, the walleye had also been stuffed with extra fillets. All this additional heft helped push the team's total weight close to 34 pounds, which, not surprisingly, blew the rest of the field out of the water. (The two only needed 19 pounds of fish to win.)

...

Had the two not been disqualified on Friday, Cominksy and Runyon would have been declared the winners of the one-day tournament—a title that carried with it nearly $30 thousand in winnings. The two also would have secured the team-of-the-year award, as they had won several other tournaments (along with tens of thousands in additional prize money) on the Lake Erie Walleye Trail in recent months.

https://www.outdoorlife.com/fishing/professional-anglers-caught-cheating-in-tournament/

[DGuller]

The Asiatic Winter Games in 2029 will be held in Saudi Arabia. 

Do they know something about Putin's war plans that we don't?

[Zanza]

It is part of that stupid line city by Prince Bonesaw.

[Barrister]

Okay, in order to solve a 2019 cold case, Edmonton Police sent a DNA sample of the rapist to a DNA phenotype company and released the results.

https://www.edmontonpolice.ca/News/MediaReleases/DNAPhenotypeOct4



I'm... I'm not sure how useful this is.  I don't think the tech is anywhere at the point where it can help predict facial features.  The article points out DNA can't tell you anything about age, hairstyle or body weight.

So really all the police now know is the suspect is black.  Umm, okay.  That's something I guess in a city with less than 5% black population.  But not much.


Ayone ever seen DNA tech used like this?

[Josquius]

Not seen it. Kind of interesting though I hope they don't put too much stock in the image

Narrowing down to 5% of the population, presumably a small % of crooks operating in a particular field is useful no?

And it strikes me they can get a bit more specific than just white if that's what they get.

[grumbler]

Lula got pretty much exactly the number of votes predicted.  The polls undercounted the Bolsanaro voters because, like Trump voters, they think that pollsters are all tools of the devil and so they lie or refuse to respond to polls.  The pollsters even noted that the Bolsanaro poll count was short of the real count, but by an unknown margin.

Are you basing that on analysis out of brazil or just talking out of your ass? It doesn't seem logically consistent to me: if Bolsanaro voters refused to answer the pollsters, why would Lula's votes be correct? If 48% of people want to vote for Lula and told pollsters that, and Bolsanaro voters to an extent refused to answer the polls and thus removed themselves from the sample, Lula would poll higher than 48%. Similarly if Bolsanaro voters decided to lie, unless they universally decided not to say they were voting for Lula and instead chose random people, wouldn't Lula also be overstated?

You seem to be confusing the words "number" and "percentage."  Look it up, and you will see that the idea that the polls could predict quite accurately the number of Lula voters but not get the percentages correct.

What people actually knowledgeable about Brazil have speculated is that in working class neighborhoods, which are Lula strongholds and neighbors can hear phone conversations because of the close proximity of housing, people were reluctant to announce their support for Bolsanaro because they didn't want neighbors to know.

Are you basing that on analysis out of brazil or just talking out of your ass?

[Jacob]

Philodendron plant controlling a machete wielding robot arm.



Video here: https://www.dwbowen.com/plant-machete