Indiana GOP/Democratic Primaries

[UpaLoompa]

I didn't mean to come off as a dick but there's nothing wrong with loving the cult of personality around a person come unraveled yet still believing that person is pretty good (though not perfect) at what he does.

There's no true comparable level of Trump to other candidates in other elections and it also requires not taking into context the voting issues of the day. Instead of admitting he was wrong, he wants to blame everyone else. Not only does being consistently wrong make Silver look bad, so does bitching about it and externalizing blame. It shows he cares more about his reputation than it does about being truthful.

His rationale about underestimating "tribalism" isn't wrong though. Trump's nativist populism isn't something that their models could take into account because there was insufficient data so--instead--he was treated as a more typical candidate. That's a data failure really.

 

[UpaLoompa]

As for coming off as a dick, really you're just coming off as weird. Did Silver do something personal to you?

 

[theBLADE1]

It's ridiculous how many people think Nate Silver is the Golden God of predicting this shit. If anything it just goes to show that you really never can tell even if all indicators point to one thing over the other. Then people would get pissed off at me for bringing up how Nate completely fucked up his predictions on Sanders chances at winning Michigan but then go silent when pretty much the same thing happens in Indiana.

The guy can be good but let's not pretend it doesn't have his own personal biases and always has the best possible information.

 

[UpaLoompa]

It's ridiculous how many people think Nate Silver is the Golden God of predicting this shit.

He's demonstrably the best in the business. If your preferred candidate's chances of winning are predicated on Silver being wrong, good luck.

 

[in the]

Silver was so bad at wrongly predicting the inevitable Trump nomination that a satirical statistician account (who predicted many races Silver got wrong) was challenging him to a forecasting duel.

But Carl Diggler is especially satisfying because pundits ostensibly trade in credibility, which is tied to making these concrete, measurable predictions, and Carl is just objectively beating all of them at it. In Iowa, he predicted Clinton over Bernie because (from his column the day before the primary):

Sure, the Bernie Bros make effective attack dogs, but my gut tells me that when it’s time to vote, these basement-dwelling slobs won’t be able to get their moms to drive them to the caucus. These Millennial misogynists have probably squandered their car privileges on too many trips to Best Buy to stock up on Mountain Dew Code Reds and gaming headsets they use to scream slurs at women of color on Xbox. Furthermore, Hillary’s cadre of web-savvy Gen Xers (the much-bandied “Hillary Men”) have been destroying Sanders supporters with logic in full view of everyone thanks to Twitter’s .@ reply function. This is the first election that will mostly be determined by online interactions, and Hillary’s veritable professional militia-style online wing has ruthlessly massacred these terrorists.

Hillary did beat Bernie in Iowa. Probably just because Iowa democrats turned out to be a little too conservative to take a chance on Bernie, but indulging the insane BernieBro narrative, it turns out, is no less reliable than the most data-driven, PhD-analyzed political sabermetrics that money can buy.

 

[theBLADE1]

He's demonstrably the best in the business.

That's fine and all but everyone pretending he's right hundred percent of the time and you should just rely on everything he says is fooling themselves. He may in fact be the best but he damn sure isn't infallible.

 

[theBLADE1]

Silver was so bad at wrongly predicting the inevitable Trump nomination that a satirical statistician account (who predicted many races Silver got wrong) was challenging him to a forecasting duel.

That's Golden LOL

 

[in the]

His rationale about underestimating "tribalism" isn't wrong though. Trump's nativist populism isn't something that their models could take into account because there was insufficient data so--instead--he was treated as a more typical candidate. That's a data failure really.

Exactly. He's supposed to be "data driven," but had a data-driven failure because he let his personal biases get in the way. Like Work Play said, there should have been radio silence when he lacked that data but instead he got cocky and started letting his personal feelings dictate his own data which affected his terrible predictions during this primary cycle. I'm repeating myself by now but at least you've seen the light.

As for coming off as a dick, really you're just coming off as weird. Did Silver do something personal to you?

I always find it fascinating when cult of personality bubbles eventually burst, and it's fun seeing it in real time and seeing people still defending him on here despite his track record of failure during this election cycle. "Well the tee vee there said I could trust this man!"

 

[in the]

It's ridiculous how many people think Nate Silver is the Golden God of predicting this shit. If anything it just goes to show that you really never can tell even if all indicators point to one thing over the other. Then people would get pissed off at me for bringing up how Nate completely fucked up his predictions on Sanders chances at winning Michigan but then go silent when pretty much the same thing happens in Indiana.

The guy can be good but let's not pretend it doesn't have his own personal biases and always has the best possible information.

I think this shows that when track records are relevant and most variables can be accounted for, statistical probabilities are way more reliable. But when we designate someone as "king" of predictions who can do no wrong, we're setting ourselves up for an inevitable letdown because not all track records are solid (or even interpretable) and many variables simply cannot be controlled for. It's the reason why so many people were fucked over by Long-Term Capital Management.


I mean, if Ryan Reilly can admit when he looked foolish, certainly Nate Silver could, too.

 

[UpaLoompa]

That's fine and all but everyone pretending he's right hundred percent of the time and you should just rely on everything he says is fooling themselves. He may in fact be the best but he damn sure isn't infallible.

No one is pretending he's right 100% of the time. My comment was "If your preferred candidate's chances of winning are predicated on Silver being wrong, good luck."

That stands.

 

[theBLADE1]

No one is pretending he's right 100% of the time. My comment was "If your preferred candidate's chances of winning are predicated on Silver being wrong, good luck."

That stands.

That has zero to do with what I was talking about. And yes some of the time Nate Silver was wrong to the benefit of my preferred candidate. My gripe and the gripes of others is how everyone pretends like he is some mastermind when he has clearly been wrong a number of times.

See trumps Republican nomination, and Bernie's victories in Michigan and Indiana. I bet he'll be wrong on some of the states going forward as well.

 

[UpaLoompa]

That has zero to do with what I was talking about. And yes some of the time Nate Silver was wrong to the benefit of my preferred candidate. My gripe and the gripes of others is how everyone pretends like he is some mastermind when he has clearly been wrong a number of times.

See trumps Republican nomination, and Bernie's victories in Michigan and Indiana. I bet he'll be wrong on some of the states going forward as well.

He is a mastermind, he's got the record to demonstrate it. That doesn't make him infallible and he doesn't claim to be. Nor have those supportive of him. What's been argued is he's very, very good. You're arguing against a strawman (see soda, that's an actual proper use of it).

 

[Fawlty]

I don't totally disagree here but would argue that rather than necessarily being bias the subjective inputs are historically based and don't properly apply to Trump.

Nativist populists aren't a new thing in the US but they're not common enough to where there's going to be enough data so Silver uses information that has been useful in the past.

I hadn't considered that. I'll keep that in mind.

It's ridiculous how many people think Nate Silver is the Golden God of predicting this shit. If anything it just goes to show that you really never can tell even if all indicators point to one thing over the other. Then people would get pissed off at me for bringing up how Nate completely fucked up his predictions on Sanders chances at winning Michigan but then go silent when pretty much the same thing happens in Indiana.

The guy can be good but let's not pretend it doesn't have his own personal biases and always has the best possible information.

Michigan re: Sanders wasn't a fuckup. All good analysis was turned on its head. The only thing people could have picked up on, if I remember right, was the debate analysis which was clearly biased toward Hillary after Bernie clearly did better.