That high of 88.1% was October 17. It decayed pretty steadily until November 4. That's 18 days, almost 3 weeks. That is not just days before.
Then it recovered from the low of 64.5% up to 71.4% over the next 4 days. That is just days before.
And I never called you crazy. Please don't put words in my mouth.
Either way, you are ridiculously splitting hairs that were already split once already. /shrug
The bottom line is, this outlet was discredited, and it continues to be wrong in elections since. They are just not using sound science, or they have a shortage of valid data, or both.
Almost 3/4 of a month is splitting hairs?
And how is it discredited? Their model reacted to the latest polling data, and based on that they assigned a chance of victory for Hillary at 71.4%. I wouldn't bet anything important if my odds of winning were that low, because that's more than 1 in 4 chance of losing. That is a significant probability.
And what is your idea of sound science, in this regard? What do you expect out of a statistical model? If we could know for certain, we wouldn't turn to statistics in the first place.
The data is as much at fault as 538 is. The figures being used right now clearly under represent Republicans, when they sample as low as 20% in some of these. The pollsters, on both sides, are clearly trying to use polls to impact public perception, rather than gauge it.
If you read the methodology that comes with the poll results, any pollster worth their salt weights the results to account for under/over representation. Not only for party breakdown, but for other characteristics that could affect the results like landline/cell.
And whether or not this is relevant, a completely representative sample would still have less Republicans than Democrats or Independents, because more people identify as the latter two.
That is why polls mean nothing at this point and even the public realizes it. The internet has saturated the average voter with info that they are bombarded daily with conflicting numbers. That is why the average voter only looks at things more critically as an election cycle actually nears.
That is actually smart use of time to only examine things when the time is necessary (eg vote). Trying to dissect tiny nuances, policy decisions, etc doesn't change the House, Senate, state legislators, etc.
The problem with the Democratic party is that they used the "centrist" strategy to come back in the 90's politically but have failed to move beyond that paradigm. At least Obama attempted to move the party to a new path but he too bogged himself down with 90's centrist DLC ideology.
Howard dean said wooooo at a fucking rally and it cost him the primary.
Do i need to list the barrage of shit trump said on his way to the presedency?
Its evident that on some level the same rules do not in fact apply. Can you imagine if jeb grabbed em by the pussy?
I agree with some of this. But, I'm not sure I agree that the Democrat party was further left prior to the 1990's. I think both parties were pretty centrist until Bill Clinton took office. (GOP moved right to counter Clinton, sparking a war on the center, imo.)
- - - Updated - - -
He didn't just say wooo. He about fucking lost it in an embarrassing display of over hype. That video was viral before viral was even a thing....
If they are so accurate, how come they are so inaccurate? I'm just asking questions here.
So, in your mind, merely 20% of America identifies as Republican, despite Republicans holding the overwhelming majority of the political offices? How does that math out, in your mind, if the polls are sampling as correctly, as you insist?
- - - Updated - - -
If the site is predicting victory of a presidential victory, based on the popular vote, then it really is useless and of no merit. /shrug
Because there are multiple sources of error. There are systematic and random errors. You're describing systematic errors, and even if you reduce those you are still subject to random error. Because we're still doing statistics, and not a census. There was a relative lack of state polling compared to national polling, and since it is state results that matter for winning the election, the models have to take those into account. Not having enough state polling translates into higher random error.
Again, please stop putting words in my mouth. You combined two different points. I addressed the 20% thing by noting that deviation from known data is accounted for by pollsters to increase accuracy.So, in your mind, merely 20% of America identifies as Republican, despite Republicans holding the overwhelming majority of the political offices? How does that math out, in your mind, if the polls are sampling as correctly, as you insist?
And Republicans win because independents don't all break for the Democrats. And there are a lot of independents. This should have been obvious, really.
- - - Updated - - -
They didn't predict anything other than a chance of winning, and it wasn't based on the popular vote. It was based primarily on state polling, because states decide the presidency.
I think we can all agree, with our hind sight knowledge of Dean, that we made a good choice to not nominate him. He is as fucking crazy as he looked in that video. He shows TERRIBLE judgement, regarding what he says, on a regular basis. He often says things his own side won't back him up on.
But, the best point here, that I can think of, is what does Trump's victory tell you about Democrats, given the Pussygate comments? What does it say about the Democrat party, that they could BADLY lose to a ridiculous clown of a man, whose own party doesn't even support him? Has this resulted in ANY change to their platform, tactics, or tact? Nope. Nope. Nope.
Expect more of the same in 2018, when you will lose to another ridiculous clown of a man, who isn't supported by his own party, Kid Freaking Rock in the Michigan Senate race.
I actually dont think the overall point is wrong. Their arent that many republican voters or ersatz republican voters ("independents" who break for R). The republicans just rig the game so well.
If you poll on policy and issues its not even close.
- - - Updated - - -
It says republicans are craven and will do anything for power. And furthermore their voters largely vote for the R regardless of whos behind it.
I can't believe how people continue to show how objectively stupid they are by saying the models were "wrong" because they put Trump at a 30% chance to win and he won.
How is a model wrong if something has a 30% chance of happening and it happens?
Are weather men wrong when they say there's a 30% chance of rain and it rains?
Is WoWhead wrong if raptor flesh has a 30% chance of dropping and the first raptor you kill gives you raptor flesh?
I can't believe people so willingly put their ignorance on display when it comes to statistics and probability.
Well, I can. All of the current models are showing Trump and the Republicans being 15-25 points down across the country, meaning 2018 is going to be a TERRIBLE year for them if this trend keeps up. They have to comfort their fears of a Democrat take-over by completely setting aside all logic and intelligence and acting like they don't understand probabilities to say "POLLS ARE FAKE!"
2014 Gamergate: "If you want games without hyper sexualized female characters and representation, then learn to code!"
2023: "What's with all these massively successful games with ugly (realistic) women? How could this have happened?!"
So, in your mind, an election that is sitting in the aggregate polls as within the margin of error for victory, makes sense to be put in terms of 88% in favor of one side? To you, that seems like some pretty sound science? Like...really?
If deviation from the 20% sample is represented in the figures, how come the results were different on election day, and in every special election? If they are doing such a good job at 538, why are they always wrong?
The definition of an Independent is a Conservative who thinks Republicans are assholes. I know, because I am one. Hillary never had a chance at them, given her extreme left views.
If their methodology is so accurate, why did they predict an 88% chance of victory, in an election that went wildly the other direction? An 88% chance of victory would at least presume the election will still be in doubt when the last two time zones had polls still open.
Classic @Tijuana, making totally baseless claims without anything to back them up, demanding sources from others, then not reading them and making more baseless claims.
Since I'm calling you out on your bullshit I guess the next step is to ask why I'm attacking you, right?
Proud far-left Democratic Socialist
Conservatism is a plague on society.
This honestly makes little sense. It sounds like technical word salad.
I just told you that there are different types of errors, and not having enough state polling means larger random errors. Which has nothing to do with whether or not Republicans are over represented.If deviation from the 20% sample is represented in the figures, how come the results were different on election day, and in every special election? If they are doing such a good job at 538, why are they always wrong?
And what do you mean by 'always' wrong? I've been reading the site pretty regularly, and I don't recall ever reading an article predicting special election winners. They did predict the two elections before 2016 really well, however, because there was a lot less variance there.
They didn't predict an 88% chance. They predicted around ~70% chance. You have to go back nearly 3 weeks to get to 88%, which is arbitrary and unfounded.The definition of an Independent is a Conservative who thinks Republicans are assholes. I know, because I am one. Hillary never had a chance at them, given her extreme left views.
If their methodology is so accurate, why did they predict an 88% chance of victory, in an election that went wildly the other direction? An 88% chance of victory would at least presume the election will still be in doubt when the last two time zones had polls still open.
And your last point about it still being in doubt with two time zones is based on what exactly? Elections are generally decided by results in battleground states, not by results in time zones.