Forecasting the race for the House

[Grenz]

No my argument is it’s the law, you conviently didn’t mention that part

'It's the law' is circular again. The legal code is fluid. That's how you get rid of things like slavery. There may be good reasons for this particular law, but you're not laying any of them out. You're deferring to the law as the highest power like it's an absolute, but it's not, it's constantly changing.

 

[JamesRussler]

If they crafted a strong plan to funnel resources into races at all levels, even no-name local shit, then control of State legislatures and governorships might not have gotten blown out of their hands. And then maybe the GOP wouldn't have been hogs at the trough during the last census cycle.

Just saying, there might be something about that message that really drives people away – even at the local level.

 

[DIDM]

'It's the law' is circular again. The legal code is fluid. That's how you get rid of things like slavery. There may be good reasons for this particular law, but you're not laying any of them out. You're deferring to the law as the highest power like it's an absolute, but it's not, it's constantly changing.

I said use your voice to try and change it

 

[DIDM]

Except for the fact that a lot of the people sitting on their asses last time not voting were reassured by most polls that Clinton would win. And then that didn't happen. How many of them do you think would have voted against Trump if they knew that's what it would take to sway the vote? Perhaps the majority? Current approval ratings would suggest that would be the case now, at least.

They sit on their ass every time

2020 will be no different

 

[Limbo Pete]

Just saying, there might be something about that message that really drives people away – even at the local level.

That's your prerogative. Imo the organizational administration is bedrock.

 

[ocean size]

IMO it is platform and organisation. I don't think that country is ready for a fully democratic socialist platform but elements that appeal to swing states independents and moderates would go a long way.

Even if everyone turns out there aren't enough self described liberals in the country to get a far left Dem elected without messaging that appeals to the middle. If Trump has taught us anything it is that people want their heads fucked with regardless of realities.

 

[Grenz]

I said use your voice to try and change it

'Use your voice to try and change it' is not an answer to the question 'why should the state supersede the individual's vote?'.

 

[DIDM]

'Use your voice to try and change it' is not an answer to the question 'why should the state supersede the individual's vote?'.

Dude, do you just enjoy circles?

It’s the law, if you don’t like it, use your voice to change it, for fucks sake

 

[Grenz]

Dude, do you just enjoy circles?

It’s the law, if you don’t like it, use your voice to change it, for fucks sake

....isn't a valid argument for or against the electoral college. But it may take you some time to get that, certainly longer than the life of this thread.

 

[Wingslaught]

You were implying credibility is lost if someone predicts 30% for something to happen and it occurs, right?

I'm saying that 538 didn't turn out to be so accurate after all. If they really had super-accurate models that predicted results better than everyone else they would have shown Trump had a greater chance of winning than most thought.

 

[DIDM]

....isn't a valid argument for or against the electoral college. But it may take you some time to get that, certainly longer than the life of this thread.

Is the electoral college what is law right now.

Well then what in the fuck are you confused about

 

[Lead]

I'm saying that 538 didn't turn out to be so accurate after all. If they really had super-accurate models that predicted results better than everyone else they would have shown Trump had a greater chance of winning than most thought.

Again, you're proving my point. Trump could've actually only had a 30% chance. You're saying if someone wins, it means the chance should've favored them but chance is running something a ton of times and seeing what are the results for each to happen. We only have the one event. It doesn't mean the 30% was incorrect.

If someone rolls a die and says the odds of it landing on 1 though 5 is 83% and we roll the die and it lands on six, was the 83% chance wrong? No.

 

[JamesRussler]

I'm saying that 538 didn't turn out to be so accurate after all. If they really had super-accurate models that predicted results better than everyone else they would have shown Trump had a greater chance of winning than most thought.

In spite of his reputation, Nate Silver has always been an overblown political meteorologist. It's amazing anyone can get paid for being wrong as often as he is.

 

[waiguoren]

Do you understand when someone is given a 70% chance of winning, that means they are saying 3 times out of 10, they lose with the information presented?

Again, you're proving my point. Trump could've actually only had a 30% chance. You're saying if someone wins, it means the chance should've favored them but chance is running something a ton of times and seeing what are the results for each to happen. We only have the one event. It doesn't mean the 30% was incorrect.

If someone rolls a die and says the odds of it landing on 1 though 5 is 83% and we roll the die and it lands on six, was the 83% chance wrong? No.

At the moment you pronounce the probability of the die landing on 4 as 1/6, the outcome of the die roll is indeterminate.

On election day morning, when Nate Silver decided Clinton had a 71.4% chance of winning, the cake was baked. The election result was better modeled as a deterministic and not a probabilistic process at that point. How do we know this? The electoral victory for Trump was large enough that random variation (e.g., unpredictable weather patterns) could not have tipped the scales to Clinton. People like to talk about "77,000 votes in three states", but those three states were correlated for reasons totally unrelated to random variation. So random variation that would give Clinton a boost in PA would (on average) be balanced by random variation that gave Trump an equal boost in WI.

Silver's election-day forecast is properly interpreted as an estimate of his confidence in his Clinton pick. In this regard, he was just wrong. His state-level predictions were also wrong. I beat Silver's 2016 state-level predictions in my bet with @m52nickerson.

I might go head-to-head vs. Silver again in November 2018. I think I can call the House more accurately than he can.

 

[aitkenmike]

I'm saying that 538 didn't turn out to be so accurate after all. If they really had super-accurate models that predicted results better than everyone else they would have shown Trump had a greater chance of winning than most thought.

That's exactly what they did. They were roundly attacked by a number of those on the left for scaremongering with Trumps odds being a lot better than other publications were showing.
https://www.vox.com/2016/11/3/13147678/nate-silver-fivethirtyeight-trump-forecast
https://www.huffingtonpost.com/entry/whats-wrong-with-538_us_581ffe18e4b0334571e09e74

 

[kpt018]

71.4%

https://projects.fivethirtyeight.com/2016-election-forecast/

Also, I went head-to-head against them in four swing states and beat them easily.

I could have sworn I saw 65%, but maybe that was after the Comey letter and maybe the odds changed? Could be my memory is fucking with me lol.

 

[sickc0d3r]

Mid-terms always swing away from the president's party. Plus turnout favors the democrats, and I expect Trump-hate-fueled record turnouts. Should be a fairly decisive ass-whooping, but maybe not quite the blue wave the dems are hoping for just because of the seats that are in play.

 

[VivaRevolution]

In spite of his reputation, Nate Silver has always been an overblown political meteorologist. It's amazing anyone can get paid for being wrong as often as he is.

I remember when he did fantasy football.......he should stick to politics.

 

[waiguoren]

I could have sworn I saw 65%, but maybe that was after the Comey letter and maybe the odds changed? Could be my memory is fucking with me lol.

You did see 65%, but that wasn't the final estimate (election day morning). I think it was 65% the day before.

 

[Dirt Road Soldier]

Democratic voters love to not show up at midterms. The economy is doing quite well. Unemployment is on the down swing. And idiots like Antifa are making the Democrats look bad.

I'd say the Republicans have a decent shot at holding the line and breaking even on both sides.