Stolen! Mine! Back off!
I’ll agree that “we need to step up our game” is absolutely a valid message to take away from this prediction. Not just that, we need to step up our game by an amount statistically more than we’re expected to, since Silver’s numbers probably already assume some baseline amount of game-up-stepping.
Yes. Right before he says +/- 5. I’m not sure why you are denying this. It’s right there in the transcript. Are you so wedded to the sensational headline that the actual details don’t matter?
Um, Evil. You need to level up. Technically Nate doesn’t make predictions, he makes forecasts. There’s a difference (though I almost always use the colloquial definitions on this board).
You can’t predict the outcome of the election at this time: there’s too much noise. You can only provide probabilistic forecasts.
Now the center of that forecast doesn’t look good for the Dems, based on the historical data. Luckily, the historical data is a little misleading, given the Tea Partier’s recent propensity to nominate jackasses in the Spring. Still those are strong winds the Dems are facing.
Here’s the transcript:
How many are they going to pick up? I’d say exactly six. But it’s probably six plus or minus five.
Reporter: That means they could pick up 11 seats. They could. Yeah.
Reporter: What you’re saying is a 60% chance that republicans win the senate. Something like that. Kind of imagine like a bell curve distribution, sort of. I can see characterizing this as a wake up call for the Dems, but honestly the pros are fully aware of the headwinds they face. Just as the Republican pros understand they have a tea party/grifter problem, of which they typically are a part.
ETA: Apropos nothing, it’s nice to have somebody who isn’t in the tank provide us with a baseline. If the Republicans can’t take the Senate under these circumstances, it will be a humiliation. Or actually not really given the odds. But you can spin it that way just as easily as you can call it a dire warning for the Dems.
I have never, ever before seen any distinction drawn between the two.
I got the distinction from Nate Silver’s book. It’s a technical one, and I’m using it to convey an idea. I’m really not trying to be picayune.
Here’s the application. Seismologists say they can’t predict earthquakes, they can only forecast them. Now that makes a lot of sense, right? Seismologists might state that there’s (say) a 50% chance of a magnitude 6+ earthquake in a given region over the next 30 years, but point estimates (eg it will occur on June 7th, 2014) are really beyond them. At least with today’s knowledge.
ETA: I guess if I want to be precise, I should admit that Nate did make a prediction - 6 seats lost. But it wasn’t especially serious - people should focus on the probabilistic forecast.
Just as it was a humiliation that they couldn’t win the WH in 2012. All the indicators pointed to a Democrat loss. But the GOP showed that they could never miss an opportunity to miss an opportunity.
It was a complete humiliation and it made their supporters look foolish and pathetic.
Or so I would like to say. The indicators gave the Democrats something like 49% of the vote. The margin between the two parties ended up being something over 2.3%, close to the standard error of 2.5%. (Think of a Gaussian distribution: +/- one std error covers the central 2/3 of the probabilities.)
Romney was a weak candidate (or Obama was a strong one) but the mismatch wasn’t overwhelming. I’m not happy about that as there was more than a little reality-oblivious wingnuttery in his campaign. It makes you miss Eisenhower or even Nixon.
http://fairmodel.econ.yale.edu/vote2012/index2.htm
Sure, I get that Silver qualified his prediction with the notation that there was a plus or minus five spread. So as few as one seat, as many as eleven. But what he SAID was that he thought the Republicans would likely win six seats … it’s right there in the text. Six seats is where the center of distribution of the bell curve goes. As it tails off to only winning one seat in one direction, to winning 11 seats in the other, the odds get increasingly lower. But Silver’s stats show the center of greatest probability hovering right over six seats. Which is why Silver said he thought the Republicans would likely win six seats.
If the polling data changes, Silver will undoubtedly change his prediction to reflect that. But for that to happen, the polling data has to change. And it’s not gong to happen if the Democrats keep playing according to their present playbook, unless the Republicans fuck up badly … which they are perfectly capable of doing. But winners don’t plan based on the hope that their opponent will fuck up.
Please let us not rewrite history again.
There was 0% chance of the Democrats losing the White House in 2012. None. I screamed this at the top of my lungs for all of 2011 and nobody sentient could have thought it was higher than 0 by or after the actual primaries in 2012. It could not happen. Not in any possible real world that didn’t have Obama buggering Bo’s puppies on the Capitol Steps. Only the kool-aid drinking deluded could have thought this for even one second. I wasn’t prescient. This was totally obvious then and even more totally obvious now if such a thing is possible.
It will not be a humiliation if the Senate goes either way this year. It’s that close. Later spin won’t make that any different. Sure, one party could yet do something so stupid that the race will tip into humiliation. That’s high probability. But it hasn’t happened yet.
And Silver is embarrassing himself by speculating without any good data. He seems to be desperate for eyeballs for the new 538. I hope he changes before November himself.
BTW, this is bizarre. And why does an elections blog even have a science writer?
It’s a sabermetrics blog, actually, which has applications pretty much wherever you want to try applying it. As long as there are statistics, and patterns in those statistics that allow for high accuracy predictions, you can use sabermetrics on it.
Partly, I suppose, the illusion that Romney had a chance was just because a party always has to go through the motions regardless of that year’s odds – think Mondale in '84 or Dole in '96 – and the motions require at least pretending you think you can win.
But in this case I think it’s also because of the RW bubble/echo-chamber effect, which has gotten worse since the two main parties became ideologically homogenous, and a whole lot worse since Obama took office. Talking only to each other most of the time, the Pubs appear to have found it inconceivable anyone would not hate Obama.
More than that, they appear to have internalized the notion that the Reagan Revolution was once-and-for-all; that a Dem should ever be POTUS again after Reagan seems to them an impossible reversal of history, like a Jacobite Restoration. That’s why so many of them don’t even accept the legitimacy of Obama as POTUS and never, for that matter, accepted Bill Clinton’s. (I remember the opener of Rush Limbaugh’s TV show in the Clinton years: “Day X! America Held Hostage!”)
The Reagan Revolution is once and for all. No party has ever succeeded in dominating the White House for more than 20 years, and rarely for more than 12. But ideologies are a lot more sticking than political parties. Both parties are still playing by the rules the Reagan Revolution set and until liberals are willing to run explicitly undoing the Reagan Revolution, it will prove as enduring as the New Deal did.
On the one hand, it’s true that small differences can make the most rancorous of struggles.
On the other hand, have you looked at the Pubbies lately? the SRIOTD thread alone can belie the idea that they are similar, if not ideologically, then from a sanity perspective alone, since the GOP has to move to the right to differentiate itself from the Dems.
There’s more to it than that. If you want to move the country further and further to the right, and the “left” party is willing to concede that the country should indeed be governed in a moderately right-wing fashion, then why not move even further to the right and see if you can get the Democrats to follow you? It’s worked so far.
We’ve successfully labelled Eisenhower conservatism as “liberalism” and liberalism as “socialism”. Let’s see how far we can take this. Maybe the Democrats will be Tea Partiers in 2030 and the Tea Partiers will be anarchists.
I forget. Against what was Reagan revolting? Balanced budgets?
Oh, nothing in particular, he was just revolting. [rimshot]
He made the public skeptical of government. That’s a reality that Obama has to concede even when he’s trying to defend government.
He made the public hate taxes. When we’re arguing over whether to tax people more making over $250,000 a year vs. whether to tax people making over $1 million, then big government is no longer viable.
These two changes were a direct reversal of what came before. The public from FDR to Nixon had come to like government programs and were willing to pay steep taxes to fund them. That ended in the 70s and Reagan rode that disgust with government to the White House. “Government isn’t the solution, it’s the problem.”
And that’s the political reality that liberals have not yet succeeded in overturning. Most of them don’t even try, which is probably why so many of you love Liz Warren. You won’t find many unabashed lovers of government out there. She’s a rare one.