Nate Silver averages polls. The problem is that any asshole can average polls so he has to do punditry to stay relevant."Chance of winning" lol get fucked
11/5/2016 10:27:19 PM
simply averaging polls is a gross oversimplification of what nate silver does.
11/5/2016 11:10:05 PM
I don't understand "chance of winning." They do the same thing for sports...the Panthers have a 55% chance of winning tomorrow. But either they win or they lose, so there's really a 50% chance that the chance of winning was correct, and there are a million factors that can effect the end result.Hillary has a 65.8% chance of winning. But what about people lying in polls? Undecideds becoming decideds? A large number of supporters of either candidate don't bother to vote? A sudden wave of people not planning to vote decide they should? There's no way to factor all of that in. And in the end, either they get to say "WE WERE RIGHT!" even if it isn't close to that same margin or they get to say "here are the reasons we got it wrong beyond our control, so our number was still right based on what we knew"
11/6/2016 3:19:42 AM
Look less at the % and more at the margin state by state. At base level all the % is saying is the polls say X and there is Y% chance they are accurate.538, which I trust, is why despite some people on here think this is a blowout I think it's very close. I can't tout him in 2012 when I wanted Obama to win and 538 was getting a lot of heat saying it wasn't too close and then throw it out in 2016 because it's closer than I like.3 days and this madness all stops. Hopefully my obsessive TWW use does too!
11/6/2016 6:18:30 AM
Going forward, for all future elections, I'm going to keep a tally of all of the races that I get mailers for, and whichever candidate wastes the least amount of paper will get my vote.
11/6/2016 2:44:36 PM
11/6/2016 3:33:10 PM
What? It clearly shows Clinton leading 292 to 246 in the EC.And no, Florida definitely matters for Trump. It's the only real way he could win, along with flipping solid blue states red. I said it before, but 323 Clinton, 191 Trump.
11/6/2016 3:48:43 PM
the tally at the bottom doesn't match the map for some reason. It doesn't match the map graphic at the bottom of the page either, which shows NH as being the most critical state for swinging the election currently.
11/6/2016 3:57:41 PM
NH might be close, but VA/PA/WI/CO/NV won't be. He needs to win a big blue and all the swing states. Don't see it happening now, given the crazy Latino turnout.
11/6/2016 4:05:20 PM
In dumbed down terms for myself, 538 assumes that one state being off correlates with other states. That's why when many are close they have a higher percentage chance of Trump winning (but also a higher percentage chance of a Clinton landslide) than other models.
11/6/2016 4:18:40 PM
11/6/2016 4:25:45 PM
^ This was enough to derail a candidate in 2004 https://www.youtube.com/watch?v=RwkNnMrsx7Q
11/6/2016 5:23:17 PM
Trump gets a pass on anything he says. As long as he appears authentic it buys him SO much leeway. That and the volume of factually inaccurate things makes correcting him impractical.That's how much people hate politicians. If you're not one you can almost say anything.
11/6/2016 6:23:07 PM
11/6/2016 6:49:37 PM
A lot of people are better off than they were four years ago. You don't luck your way into the approval rating he has in year eight if that isn't the case. The only group of people you can argue are far worse off is the white working class constituency that is the core of Trump's support. I hope, but don't think, Clinton works very hard at helping this group..
11/6/2016 6:52:39 PM
people are better off right now but everyones future (25+ years from now) has gotten much worse over the course of the obama presidency. Approval ratings only confirm people don't understand/care about that.
11/6/2016 7:11:33 PM
Work together to fix the issues?
11/6/2016 7:21:39 PM
11/6/2016 7:38:58 PM
I don't disagree with much of that but what is ~truth? Trump lies more than any human I've ever seen. Not political candidate. Human.
11/6/2016 7:43:36 PM
He lies a lot, but so does Clinton. No one has ever talked to a politician like he did in, say, the second debate. And many of the things he said there were not untrue. If a regular person had done the things that Hillary Clinton had done (talking about classified emails on an insecure server here), they would have faced serious consequences.So, yes, he lies, but he also points out some uncomfortable truths.
11/6/2016 7:51:29 PM
There isn't an equivalency between his lying and any other person I've ever seen.Clinton lies like every other politician lies. Not good, of course, but run of the mill.Her server and classified information is a different conversation
11/6/2016 7:54:16 PM
Yeah, there's a difference between "telling it like it is" and "telling it like however you see fit, regardless of whether it's true or not"I'm sure that years down the road, his rise to power will be a hot topic of debate among political scientists and historians. Unfortunately for him though, it's going to be more of a cautionary tale than anything else.
11/6/2016 7:59:26 PM
11/6/2016 11:16:34 PM
trumps lies are not as dangerous because they are easy to spot. its the lies democrats make that convince people problems are solved that are our biggest problem.
11/7/2016 9:37:41 AM
That's horseshit. Go to a rally and see if his support thinks he is lying at all.Though I am glad you finally admitted Friday evening that you want Trump to win
11/7/2016 9:42:22 AM
11/7/2016 10:21:34 AM
You should buy two tickets then50% + 50% = 100% BOOM
11/7/2016 10:26:23 AM
it's bulletproof
11/7/2016 10:31:42 AM
Just as I expected, 538 is flickering the toss ups like NV, NC, and FL back to blue. Markets are going nuts given the email update too. I bet Trump really wishes he had access to his Twitter account right now...
11/7/2016 11:39:15 AM
one item to note is that 538 doesn't take in to account EV at all. in one of their casts they gave a pretty good explanation as to why. gist is that it's so small scale and relatively new that they don't have the data to translate EV turnout into a variable for their model that works.[Edited on November 7, 2016 at 12:01 PM. Reason : .]
11/7/2016 12:01:41 PM
Senate is about the only thing that has me worried right now, but I don't see a scenario where Hillary wins by +3 or more and doesn't pick up the necessary seats to give the Dems a majority.
11/7/2016 12:03:46 PM
Say what you want about 538 blowing the predictions for the Republican primary, and possibly the general election, their graphics are just awesome:
11/7/2016 12:13:32 PM
i think romney is going to win
11/7/2016 12:21:32 PM
11/7/2016 1:35:15 PM
but my ticket either wins or loses. as you stated:
11/7/2016 2:28:33 PM
There's a 50% chance a predicted Hillary win is correct and a 50% chance a predicted Hillary win is incorrect. because there are only 2 options.If there was a legit 3rd candidate who had any real chance of winning, then the chances of a predicted Hillary win being correct would be 33%.(And there are no other factors on the powerball...you can't change your numbers after you've submitted them, unlike the polls and predictions about votes that haven't been cast yet)Statistics create an expectancy based on past behavior or occurrences up to this point. A baseball player can bat .350 over the last 5 games, but nothing about that will tell you how many hits he's going to get in tomorrow's game.[Edited on November 7, 2016 at 2:56 PM. Reason : -]
11/7/2016 2:48:05 PM
srsly. stop. i'm dying over here.
11/7/2016 2:53:23 PM
^^ An election is not a coin toss. Not all choices are equally likely.
11/7/2016 2:56:53 PM
do you think sports handicapping is witchcraft?
11/7/2016 2:59:25 PM
I'm not saying Hillary has a 50% chance of winning. I'm saying Nate Silver has a 50% chance of being correct in saying that Hillary will win.Let's assume goalielax is an asshole 65% of the time. So I'll say there's a 65% chance he will be an asshole in his next post. But there's only a 50% chance that I'm right, because either he will or he won't.If the odds are Team A has a 60% chance of beating Team B by a spread of X, I'm not betting that Team A will win or lose. I'm betting that the odds were right or wrong.[Edited on November 7, 2016 at 3:04 PM. Reason : -]
11/7/2016 3:02:01 PM
you'll either win the lottery or you won't
11/7/2016 3:05:26 PM
^^ wat?
11/7/2016 3:06:58 PM
In this election there is one outcome which would make the prediction of "Hillary wins" right, one outcome which make the prediction wrong.In a lottery there is one outcome which would make the prediction of "You will win" right, and millions of outcomes which make the prediction wrong.
11/7/2016 3:19:23 PM
Even in the case of whether or not Silver's predictions are correct, the two outcomes are not equally likely.
11/7/2016 3:20:10 PM
It's "chance of winning", not "winning or losing". There's lots of variables.
11/7/2016 3:21:28 PM
there is a 100% chance bdmazur failed stats
11/7/2016 3:24:39 PM
Didnt the market surge before Brexit when everyone thought they would stay in the EU?
11/7/2016 3:28:50 PM
Deans list as a stats major for my first 2 years at State.I switched out because I wanted a career I could do the work directly instead of analyzing other people's work.Statistics is useful in scientific testing. Not in opinion polling.
11/7/2016 3:30:20 PM
lol ok - good job on getting good grades before you started any major coursework.
11/7/2016 3:38:02 PM
LOL at 538 updating it's site every few hours to show a higher chance of a Clinton victory. It updated only once yesterday.
11/7/2016 3:44:15 PM