No. This is dumb. Accusing your political opponent of being a Russian plant delegitimizes the issue.
The government should push it and the media should push it. Clinton should stay away from it.
Bingo. If Russians did do it, NSA knows and will say as much. If they don't say anything by end of DNC, probably because Russians didn't do it.
Rosario Dawson: "if Trump wins, it's not our fault."
If Clinton manages to lose to Trump, it won't be the tiny minority of Sanders voters not voting for Clinton that will have lost the election. Clinton's on track to keep more of the Sanders voters than Obama did for Clinton voters. (We'll see how the DNC goes)
If polls are wildly variable this year, how is Nate's model at all useful?
Because he's compensating for the polls being wild?
Regarding 538, I'm sorry but I think Nate's model is horseshit. The swings are way too wild to make any sense. Especially since there's been nothing on the state polling level to indicate the big reversal in the NOWCAST.
So I looked into it and he's unskewing polls!? I don't mean putting weights on pollsters but actually adjusting the numbers. I was confused how he has Donald up in Pa and I looked and despite nearly every poll showing Hillary up (9 out of 10 polls), he adjusted them to 6 showing Trump up. A Clinton +8 is now Clinton +3. Clinton + 4 (from PPP) is now Trump +1.
What the fuck is this horseshit modeling? How can 9 out of 10 polls show Clinton with a decent lead (the avg is like +5) be Trump +1? Something he calls "trendline" which I assume he just believes because the trend nationally or something is one way that it means the numbers in the state are wrong? Which would be a stupid way of doing it but i won't comment without knowing exactly wtf that means.
All I know is it makes no sense and would explain why 538 is so bullish on Trump right now and everyone else like Upshot and Sam Wang are not in their modelling.
Honestly, and I don't say this because I support the numbers that are better for Hillary, but I think 538 is trying to create a model so amazing that is sucks. Stay simpler, IMO. Their model isn't good right now, IMO. I would basically ignore it and look at Upshot.
It's the same model used in 2012; the idea is fairly straightforward - polls tend to be precise if not accurate. Calculate a poll's precision and compare to actuals, and adjust accordingly. I'm worried that folks are discounting Silver's model because it isn't telling them what they want to hear. (Also, he is trolling people with the now-cast in order to have some fun, and see who actually reads what the algorithm is and understands what he's doing). Nate's model is dependent on pollsters having a high order of precision, though.
The Now-Cast having Trump only at a 57% chance to win at what is probably the single highest point of the race for him (post RNC convention, before DNC convention, right after DNC scandal and after Clinton gets hit for FBI stuff) is...actually pretty heartening.
The remainder aren't even socialists. It's amazing how almost nobody is moving to Stein, and instead going to Trump or Johnson. It just shows that politics is about personality and not policy.
I wish people would get this. We got spoiled by Obama having both.
The difference in Wang's and Silver's attitudes today is something else.
One is calm and collected, sharing data and reasons why nobody should really panic and explaining why his model isn't really changing, and Silver has Trump winning and going crazy on Twitter.
Did the Primaries really shake 538 that much?
No; the models are modeling two different things. Wang's model is based on what will happen in November, Silver's model that everyone is freaking out about is based on having the election today. You know, when Trump is at his peak and Clinton is at her worst point.
Aren't you guys seeing red here over nothing?
It seems like he's using "friends" facetiously here, no?
confirmation bias is a thing. People want to believe Trump is that dumb.
Looks to me like they unite where it counts.
Democrats fall in love, Republicans fall in line.