Hacker News new | ask | show | jobs
by TomDavey 1635 days ago
> anticipates 538 and all the other media on elections that ultimately disenfranchise voters.

Serious question: how does "media on elections" disenfranchise voters? I would have thought that they usefully inform voters, or at least remind voters of the importance of voting.

3 comments

Go read the sources I point to.

In the case of Franchise a computer does an interview of one voter and then calculates who the president should be.

In real life, if 538 was perfect at simulating the election there would be no need to have the election.

The point of Insider Baseball is that ‘horse-race’ coverage and coverage that pretends to give you an insider view of the campaign as the candidates and their staff see it completely avoid any real discussion of who the voters are, what they really want, what really motivates them, what alternatives they really have, etc.

538 in perfect irony applies the techniques and terminology of sports betting to politics.

The funny thing is, they're just as often wrong with their sports predictions. No serious sports wagerer I know would ever rely on 538 to make their picks. Gamblers understand that touts serve two masters. Apparently, punters are less gullible than political pundits when it comes to believing that a set of chosen stats represents a neutral, scientific prediction.
Everyone likes to hear what they want to hear - until the bill comes anyway.
Predictions are hard, particularly about the future.
They by no means have total control. People can still think. The point is media also provides interpretation. So, instead of going into a voting booth reflecting over your own needs or principles, you go into the booth thinking about polling and statistics. You vote for somebody because their number is higher on 538. It gets worse when you consider that the people being polled are themselves thinking about who is most viable or likable.
It could just as easily work the other way. e.g. all the polling says Trump can't beat Hillary, so no one bothers going out to vote for Hillary.
In fact, Hillary won the popular vote by quite a lot, 2.8M or more than 2% of the vote.
There are some people that are suspicious of polling averages and models have under-performed Trump vs. the official vote totals. They think it’s a form of “suppression polling” where the opponents supporters are demoralized by a conspiracy of weak polling numbers that suppress turnout.

I believe there are simpler and more convincing explanations for polling errors that seemed consistently biased against Trump, but some people are happy to jump to far fetched conspiracy theories.

One counter-vailing fact is Republican internal polls showed Trump losing up until election night in 2016. Polls they did for themselves, which they did not share at the time, showed him losing.

Aside from this, to influence opinions with sharing polling information, one just needs to change the question, or the audience, or both. "Should the US seek peaceful resolutions with Russia" will get a different answer than "Should Biden oppose Russia's military buildup on Ukraine's border". You can tailor the question to the answer, and have Americans either supporting or opposing abortion, or whatever.

Also, a poll of everyone will yield different results from a poll of, say, likely voters. You have to look to who is polled along with what is polled.

With these things done, there is little need to fudge the numbers, other polling organizations can ask the same question to the same demographic and get similar answers.

Why associate the comment you replied to with Trump, and why is associating that comment to Trump enough to dismiss criticism of publishing continuous polls as "far-fetched conspiracy theories"?

There have certainly been massive polling failures (and obvious push polling) in elections over the past decade that can be discussed rather than being dismissed for the sake of partisanship. The organizations who actually do the polling discuss these issues constantly without accusing each other of being deluded.

I think the simplest explanation is that pollsters are from the media/elite; when Trumpers get a phone call from a pollster, it's either their chance to "own the libs" by lying, or they're embarrassed to admit their reactionary beliefs so they say what they think a centrist or liberal would say. Then they vote for the furthest right wing loonies they can find.

You can see this kind of thing in practice if you've ever been a non-white person in a redneck bar.