September 13, 2018

FiveThirtyEight gives the Democrats a 5-in-6 chance of winning the House and a 1-in-3 chance of winning the Senate.

New analysis today — here for the House and here for the Senate.

109 comments:

Joe Biden, America's Putin said...

MO is a solid red state. Went BIG for Trump. How on earth is the corrupt old hag Claire McCatkill(D) in the lead?

Rae said...

538 gave Clinton a 71.4% of winning the election on November 7th.

rhhardin said...

So the dems have a 1 in 3 chance of winning both houses. #math

Virgil Hilts said...

Hard to believe their prediction for AZ senator. Sinema (who I really like - she's one of my favorite Democrats in the country) may beat McSally but to give her 2:1 odds is just stupid. Latest polls are showing R support gravitating to McSally who is now leading Sinema. Really amazed at that prediction.

Mike Sylwester said...

The linked webpage is well done.

Joe Biden, America's Putin said...

North Dakota needs to wake up. too.

Levi Starks said...

I hope they’re as accurate this time as they were in 2016

Mike Sylwester said...

I feel optimistic that the Republicans will keep the House majority.

Keep in mind that Vladmir Putin again will spend tens of thousands of dollars on Facebook advertisements and that Carter Page, an experienced agent of Russian Intelligence, will be providing advice from President Trump about which House races to target with the Facebook advertisements.

tim maguire said...

Probably right for the House, but Real Clear Politics is predicting at least 1 pickup for the Repubs in the Senate, which is probably right for them.

Mr Wibble said...


Blogger Rae said...
538 gave Clinton a 71.4% of winning the election on November 7th.

9/13/18, 8:14 AM


Yes, and Silver also emphasized back then that the polls were untrustworthy and that if Trump won Pennsylvania or Wisconsin he'd likely do well across the rest of the upper midwest and win the election. If you watch some of his interviews prior to the election, he makes it pretty clear that while he sees Hillary as a favorite, he also saw the race as closer than everyone did.

I'm Full of Soup said...

I have been figuring there was probably a Blue Wave coming since the Dems and the media are so damn angry and crazed. My only hope was that the good economy would tamp down the Blue Wave even though the media had been keeping god economic news under wraps.

Then yesterday, the local librul newspaper the Phila Inquirer had a headline that read something like this:

"Will the voters opt to stop Trump or continue with the good economy". I was shocked to see that newspaper acknowledge the great economy could convince people to vote for the Repubs. So let's hope the Blue Wave peters out.

tim maguire said...

Virgil Hilts said...Sinema (who I really like - she's one of my favorite Democrats in the country)...

I wonder how that happens. I can understand a person of integrity rejecting the Republican Party, but I cannot understand a person of integrity embracing the Democratic Party. How does a good person align with that madhouse?

rehajm said...

If they don’t get it right this time- and I don’t mean LLR Chuck right or probabilistic prediction models mean never having to admit you’ wrong right, I mean Ds don’t take the house- can we please bury Nate and his too cute quantitative models only a leftie nerd could love?

Joe Biden, America's Putin said...

It's still illegal to beat Hillary or to look for any negative information about her.

Crime of the century.

Dave Begley said...

Ann:

Why do you cite this Dem shill who likes to put the gloss of "science" on his predictions? He's a modern day Wizard of Oz. He was massively wrong before. Why will he do better this time?

I predict the Cornhuskers win on Saturday! Badgers too. I have the science behind me.

rehajm said...

I mean like some of his quantitative ideas but it’s his partisan rooting and best spin on it crap that kind of defeats the purpose.

Dave Begley said...

In all seriousness, Trump wins 40 states in 2020. And I did predict Trump's 2016 win on the website neoneocon.com. I actually saw ALL the candidates except Jeb.

Birches said...

I keep wondering how all these prediction makers are so confident in their decisions seeing as they were so wrong last time. They always bring up 2012 and unskewed polls to prove their right but they completely skip over 2016.

rehajm said...

You dig into the assumptions behind many of the polls he uses in his models and he gives lots of weight to special election turnouts. It seems like a similar error to giving Hillary all Obama’s voter turnout. Those same motivations to get rid of the incumbent in Alabama aren’t the same motivations to keep Dakota Heidi in office if she votes thumb down.

Freeman Hunt said...

After 2016 I'm done paying attention to polls.

And now with the revelation of the massively wrong school shooting statistics put out by the Dept. of Ed., I'm pretty done with stats and science based self-reporting generally.

rhhardin said...

Skewed polls are okay. It's called stratification. There's more processing required but you get a better result that straight polling, assuming you know what you're doing.

Oso Negro said...

Americans everywhere yearn for higher taxes, open borders, and shitty service industry jobs. Which party is best positioned to deliver that?

The Crack Emcee said...

Damn it, the Democrats are now a cult, and cults can't get it together enough to win elections anymore (Jim Jones as the San Francisco Housing Authority was enough). If I'm wrong about this, mark my words, we're gonna have zombies in the Space Force.

Wait - that would be cool.

Joe Biden, America's Putin said...

The Socialists want to give everyone free stuff.

Nothing is free so who will pay? Certainly not rich progressives like Jared Polis(D).
Jared Polis(D) is worth upwards of 150 million and he has NOT paid income tax in many many years.

Taxes are for the little people.

btw- if Manafort was such a tax cheat - where was the IRS all this time?

Big Mike said...

I’ll take the Republican side of that bet.

readering said...

I guess folks here keep away from sports books.

Michael K said...

Sinema (who I really like - she's one of my favorite Democrats in the country) may beat McSally but to give her 2:1 odds is just stupid.

I expect to see a couple of TV ads of Sinema in her pink tutu on 9/11 before November.

McSally is very impressive in person, aside from her career history.

SteveR said...

It’s the economy stupid!

mccullough said...

538 are the baseball experts who get baseball wrong all the time.

I have no idea who will win and neither do they. Polls were spectacularly wrong last time. Maybe they are better this time but the results will show this.

The Red Sox are already at 100 wins and are going to win their division. Betts is a much better player than Stanton it Hudge. The geeks at 538 didn’t know this. Basic stuff. They are busy crunching fake numbers while Betts was making adjustments.

William said...

There's a reverse Bradley effect with Trump. People like him a lot more than they're willing to let on. That screws up the polls. I don't know if that's true with polls about Congress. There's a good chance the Dems take back the House, and we get to hear a lot more from Nancy and Maxine, The WWF dimension of our government will be greatly magnified.

The Cracker Emcee Refulgent said...

Rationally, the Democrats should take Congress. It's a mid-term election and it would be odd indeed if the Democrats didn't recoup some of the massive losses they suffered in Obama's time.
But I can't think of a single instance in my life where incontinent, batshit, frantic insanity was a harbinger of triumph.

Darrell said...
This comment has been removed by the author.
Darrell said...

And if that doesn't get you to the polls, we'll make some other shit up.

Original Mike said...

The odds are they will get it right someday.

Darrell said...

Didn't the NYT hire 538 to create that real-time flashy opening page that had Hillary with a 97% chance of victory, early in the night?

Michael K said...

But I can't think of a single instance in my life where incontinent, batshit, frantic insanity was a harbinger of triumph.

That gives me hope although, if they should win the House, it would suggest that a majority of the country is crazy.

I still can't believe the crazies are that many.

Darrell said...

All the Democrats had to do was stay sane to win the midterms.
They went full crazy.
They will not win.

Original Mike said...

”I still can't believe the crazies are that many.”

Yeah, but they vote.

wwww said...



The Senate map is favourable to Republicans. North Dakota, Missouri, Montana, West Virginia, Tennessee. Rs should win all of these states. These aren't races in Maine or Massachusetts. Scott is running a good campaign in Florida. He's done a lot of outreach to Puerto Ricans in Orlando. He should win.

That's 6 Senate seats Rs should be predicted to win.

House is a toss up because Trump has activated people who normally don't care & don't vote in midterms. Midterms tend to run under the radar. The voting population is normally older compared to a Presidential. The usual midterm turn-out would re-elect a Republican house. 23 seats are a whole lotta seats. It will come down to each race.

It's possible Trump's election has changed the dynamics of midterm races. The Trump coalition includes more non-college educated people then the normal R candidate. The coalition includes less college educated people then the normal R. candidate. Non-college educated people are less likely to vote in midterms. Will this change turnout in midterms?

Independents: look at their voting tendencies. They'll be a critical swing group in each race.

Suburban women: If white college educated women are too lopsided in favor of Ds over Rs will increase the challenge factor of each race with a significant college educated populations. Some polls are suggesting gender gaps that are eye popping. Rs can't bleed out here. A loss of some, sure, survival is still possible. But numbers like 29/63 will make an impact.

Birkel said...

Oh, look! Another poll that is massaged to shape public opinion but which will move at the end to reflect what the supposed "reality" actually is.

This game is boring.

Mark O said...

Let's see. I predict an October Surprise.

Now that's a safe bet.

Ficta said...

"Didn't the NYT hire 538 to create that real-time flashy opening page that had Hillary with a 97% chance of victory, early in the night?"

No, that was not 538. 538 had Trump at ~30%.

Birkel said...

The hurricane causes flooding but the resources the local, state, and federal government bring to bear are sufficient to minimize the natural disaster.

Trump declassifies FBI/DOJ documents about a week after the hurricane mess is out of the news.

The declassification shows massive corruption, Democrat complicity in spying on Trump, Mark Warner's (and Richard Burr's) total corruption of the SSCI.

Indictments are delivered against some FBI/DOJ officials who are pressured to flip - and who have actual knowledge of illegal behavior.

Let's see... All those things would radically alter the terrain for all the midterms.

Qwinn said...

The notion that Dems have ANY chance of taking the Senate is utterly delusional.

I still predict a minimum of 58 Republican Senators after November. Been predicting at least 57 for a year and events since then only drove me to up it a bit. Even 62 is not totally improbable.

Tommy Duncan said...

The Democrats have a 5 in 6 chance of taking the house?

5/6 = 83.3%

Why am I skeptical? I'm expecting a Republican pickup in my Minnesota District 1.

The Cracker Emcee Refulgent said...


"House is a toss up because Trump has activated people who normally don't care & don't vote in midterms."

Maybe, but I question that. Partisan fervor usually comes from...fervent partisans. Those who "normally don't care" are seldom motivated to care unless something arouses them very close to home.

Tommy Duncan said...

Consider this:

Dow 26,100
Nasdaq 8,000
S&P 2,900

Achilles said...

Ficta said...

No, that was not 538. 538 had Trump at ~30%.

Bullshit.

2 months before the election they have had teens and single digits.

Just like every other democrat push poller they slink back towards the truth when the election gets closer.

Achilles said...

If Democrats were going to win Facebook wouldn’t be so clearly fascist.

If Democrats were going to win they would be saving the violence and thuggery for AFTER the election.

Rae said...

If you watch some of his interviews prior to the election, he makes it pretty clear that while he sees Hillary as a favorite, he also saw the race as closer than everyone did.

"Clinton will win, unless I'm wrong" is not a "scientific poll".

The Cracker Emcee Refulgent said...

When I became politically aware in my late twenties, I was struck by how difficult, if not impossible, it is to wake the indifferent to what I saw as important issues. Importuning others with queries of "Are you saved?" is a poor way to make converts. And right now, Democrats are like Jehovah's Witnesses on an epic meth jag.

Bruce Hayden said...

At least in this election, I think that polls are silly. For one thing, there are enormous forces working in the background. No surprises se, I think, but I expect record amounts of money to have been spent on this election. At least in MT, the Dems have been spending heavily (almost all coming from the two coasts) since the first of the year. Republicans are just starting to match them. So, the Dems have an 8 month unanswered advantage in advertising. It is now being answered, and Trump has been to the state twice now, both times for standing room only crowds.

But I think maybe more important is what else is going on. I think it highly likely that the Carter Page FISA warrant application will be unredacted within a week or so, showing conclusively that the Deep State, working closely with the Crooked Hillary campaign and the DNC weaponized law enforcement and counterintelligence tools for their political advantage. The CIA attempted to entrap Trump campaign workers and the FBI spied on him, his transition, and his White House, with the tools designed to prevent the next 9/11. It is clear that the #2 in both the DoJ (Yates) and FBI (McCabe) were actively involved, along with the DNI and CIA Dir. But it is looking more and more like the Obama White House was actively involved. That all will very likely be out well before the election. On the flip side, Mueller doesn’t seem to have anything, and it looks like his investigation was based, initially, at least partially on the infamous Steele Dossier, that too many in the DoJ and FBI knew at the time to have been bogus. Latest is that Mueller’s lead investigator, Weismann, was involved with Steele, Simpson, and the Ohrs well before the election.

On the flip side, the left, with the aid of the tech mega corporations, has been deplatforming their opponents as fast as they can. It started almost innocently with big banks refusing to handle credit card transactions with gun related businesses, and Amazon terminating their associate relationship with some conservative sites, such as Legal Insurrection. Then we saw Google and MSFT biasing their search results, and this is becoming more blatant. We had a bunch of shadow banning, but over the last week or two, they have taken to deplatforming with a vengeance. Trump TV on YouTube was deleted, including thousands of videos, a week ago. Then Alex Jones, and yesterday the Great Awakening was shut down on Reddit, with, of course, all archives deleted. Why is that significant? Because it was the main public outlet for QAnons. They are called conspiracy theorists, but by now, it is fairly clear that Q is close to Trump, and has been driving a lot of the revelations damaging the Dems for some time, probably including the two major #MeToo casualties this last week (Monves and the 60 minutes guy). And today we discover that one of the big voting machine companies installed PCAnywhere on its voting machines. Why today? Why the timing. A lot of this looks like move and countermove in the background.

n.n said...

Democrats are preparing to pull one out of their nether hole. A million to one chance, doc. A million to one.

Left Bank of the Charles said...

The 16.5% chance is the "classic" forecast. The "lite" forecast gives Republicans a 26.1% chance of retaining the House. In other words, two coin flips in a row that come up heads. Or flipping heads three times in a row if you accept the classic forecast.

Michael K said...

With the deplatforming going on in the "Social Media" sites, blogs are more and more important.

Even some fo the crazy conspiracy theories we see here are at least still available and not hidden.

Hidden forces grow more powerful, not less.

Ray - SoCal said...

FISA warrants seems to have been highly politicized and approval rubber stamped.

Sheryl Atkinson may have had a FISA warrant on her.

Think what happens if this comes out?

Or Trump pardons Martha Stewart.

And Trump is a master of playing the media...

Ray - SoCal said...

The google video that came out of the first all hands meeting after Trump got elected is bad.

Something major is going to happen to google by the us government due to it.

Matt Sablan said...

"How on earth is the corrupt old hag Claire McCatkill(D) in the lead?"

-- Never underestimate the power of incumbency.

MadTownGuy said...

Matthew Sablan said...
" -- Never underestimate the power of incumbency."

Tell it to Russ Feingold.

Rick said...

How on earth is the corrupt old hag Claire McCatkill(D) in the lead?

It's normal for people to desire split government. Since Trump can't be voted out a Dem Senate is a convenient brake.

This is why SC appointments are so important. It's one of the few instances where moderates desire Republican unity. Dems have played to Trump's advantage by showing they are incapable of reason or moderation even when facing the mildest possible nominee.

Yancey Ward said...

My theory about Silver and 2016 is pretty simple- he got access to Clinton's internal polling results in the middle of October- something other prognosticators didn't get. That is why he gave Trump a better shot than anyone else on election day.

JPS said...

Say, checking back in on 538 reminds me of a cruel filter you can apply to loudly announced findings in climate science, especially in modeling papers:

Do the authors provide error bars?

I don't see them, but gosh, 83.5% is much more authoritative than 5/6. They must really have a precise method. We'll see if it's an accurate one.

The Cracker Emcee Rampant, 10:37:

"Importuning others with queries of 'Are you saved?' is a poor way to make converts."

I love this comment, but - who the hell tries to make converts anymore? Our team is awesome, the other guy sucks, and you and I should feel proud that we're not them.

TRISTRAM said...

Why am I skeptical? I'm expecting a Republican pickup in my Minnesota District 1.

Don't forget the crooked officials: Judges in PA and elsewhere (NC?) throwing out district maps to get in more democrats. NM looking to change voting rules (one button straight ticket) to freeze out libertarians.

Sure, a few of these have been beaten back, but Dems may get 4-6 seats from PA simply because the top court re-gerrymandered the districts to favor democrats. I am sure there are other regularities (it is hardly irregular to have blue state judges mess with things).

Matt Sablan said...

"My theory about Silver and 2016 is pretty simple- he got access to Clinton's internal polling results in the middle of October- something other prognosticators didn't get."

-- It is possible; the Clinton campaign was very good at co-opting the media with promises of access.

Bushman of the Kohlrabi said...

The latest Marquette Law School Poll had Vukmir within 2 points of Baldwin for Wisconsin Senate. They have a pretty good track record. FiveThirtyEight has Baldwin up by 13. Hard to take that seriously.

Ficta said...


"No, that was not 538. 538 had Trump at ~30%."

"Bullshit."

I was only addressing the question of the source of the NYT's Election Day projection. It wasn't 538, whose Election Day projection was around 30%. I also don't recall Silver's projected percentage for Trump(i.e. not "if the election were held today" but, "what if trends continue as observed until election day") ever dropping below about 20%. I was watching that.

I don't get the extreme vituperation of Silver by some here. Maybe you think all polls are useless. Maybe you think probabilistic election forecasting is useless. Those are arguably valid points of view. But if those aren't your positions, Silver is about as honest as it gets. I'm not saying he doesn't let his bias cloud his judgement, but, to my eyes, he tries harder to avoid bias than anyone else except Real Clear Politics, who avoid it by only doing a much simpler analysis than 538.

I'm Full of Soup said...

I think the PA Supreme Court's unconstitutional re-districting of the state's congressional districts will lead to the Dems picking up at least 4 seats in PA alone.

Yancey Ward said...

Silver's predictions tend to only look good when Democrats do well in elections. That is why I have less regard for him than many- his results look a lot like luck to me.

Bay Area Guy said...

I feel good about the Senate, I am worried about the House. However, anyone who has followed politics for 50+ years understands the natural ebb and flow of political majorities. Majorities don't last. Trump and the GOP have been riding high -- just as Dems were riding high in Congress, until 1994.

So, the question for me is, Does a Democratic House in 2019 obsessed with impeaching Trump help or hurt his chances in 2020?

There's an argument that Dem overreach may, counter-intuitively, help Trump in 2020.

Matt Sablan said...

"Does a Democratic House in 2019 obsessed with impeaching Trump help or hurt his chances in 2020?"

-- The real question is: Will they succeed? Or, worse, how long can the Federal government be paralyzed by impeachment? This won't be like with Clinton, where things still went on as the impeachment happened. The impeachment will be an argument for invalidating everything Trump has done as president, that his appointees have done, etc., etc.

FIDO said...

Here is the deal: from the 1933 to 1994, the Democrats essentially owned the House and Senate (with a few flips)


And what did the Dems have to sell? Socialism and Free Stuff, which worked well since they also frequently owned the Presidency and the President, even if Republican, tended to just sign off on this stuff.


What are the Dems selling now? More immigration? We want to get rid of Trump? More Obamacare? No Russian collusion unless it is a Democrat?

What is the coherent sales message?


Trump is selling prosperity, not giving stuff to foreigners who hate our guts, less immigration, no Obamacare (though weak on this point) and 'I really LIKE America'.


Democrats are selling 'you are racist scum who are too rich and need to let others cut in line'.


So if I remember and feel the urge, I'm willing to put $100 on Republicans holding the House. Not enough to hurt me, but if I get 4 to 1 odds (or better), it will be a nice packet of cash. Cause 'Crazy' doesn't sell in politics. And the Dems have been acting CRAZY.

buwaya said...

"The real question is: Will they succeed? Or, worse, how long can the Federal government be paralyzed by impeachment? This won't be like with Clinton, where things still went on as the impeachment happened. "

The real problem is that the Federal agencies will be free to ignore the executive, and to pursue their own policies. A Democratic Congress in cahoots with the bureaucracy is a nearly-impregnable thing. Congressional investigations of Presidential appointees, Mueller-style, will have full and zealous cooperation.

Henry said...

These 538 posts are all the same.

If you're interested in the analysis, 538 is interesting.

If you're interested in the outcome, 538 is either your BFF or the scum of the earth. Tediousness results.

Earnest Prole said...

Trump was elected on a promise to fuck the establishment and build the wall. He's had multiple opportunities to stare down Congress, and each time he's allowed the establishment (McConnell, Ryan, Schumer, Pelosi) to fuck him instead. That is the reason his numbers are so poor despite the economy, and the reason his party will lose the House. Trump blinked.

CJ said...

Manchin, Tester, McKaskill (the largest recipient of MPAA and RIAA lobbying $$$ in the country. WHY?), and Heitkamp are all going to lose.

The Republicans are running popular former House members or other officials with name recognition and a relatively moderate record.

Howard said...

The LA Times poll gave Trump even money to win. Sabermetrics has it's place, the problem with it is that it is typically used by people with no common sense, so they don't know how to do a straight-face test.

On election night, the NY Times real-time predictor was the first to show Trump winning. They wern't the first to know, however. Watching CNN, you could tell by body language that something was fishy in Denmark wrt Hillary's ascendance.

Roy Lofquist said...

"A Republic, if you can keep it." ~ Benjamin Franklin

The idea is that we elect people to make and enforce the laws. Every two years we have elections to decide whether those elected people are doing a good job or should be replaced.

Given the unprecedented successes of the Trump administration if there is not a "Red Wave" then the very basis of the Republic is in doubt.

Bruce Hayden said...

"Does a Democratic House in 2019 obsessed with impeaching Trump help or hurt his chances in 2020?"

I am thinking of ads showing Maxine Waters shrieking “impeach, impeach”, etc, sowing Pelosi with a gavel, and Antifa fascist thugs whaling on people would be a good campaign ad right now for the Republicans.

Rabel said...

If Arizona voters decide to send the atheistic, bi-sexual, former Green Party activist Krysten Sinema to the Senate, it will still be an improvement over their last two selections.

wwww said...

"I don't get the extreme vituperation of Silver by some here. Maybe you think all polls are useless. Maybe you think probabilistic election forecasting is useless. Those are arguably valid points of view. But if those aren't your positions, Silver is about as honest as it gets."

My theory:

It's a human impulse to want certainty. It's uncomfortable for the future to be uncertain. I have concluded the majority of people do not want a probabilistic analysis. They want Nostradamus to predict the future 100% of the time.

Bay Area Guy said...

@Matthew Sablan,

-- The real question is: Will they succeed? Or, worse, how long can the Federal government be paralyzed by impeachment? This won't be like with Clinton, where things still went on as the impeachment happened. The impeachment will be an argument for invalidating everything Trump has done as president, that his appointees have done, etc., etc.

Not necessarily. The Senate is the prime mover on impeachment. That's where the trial is held. The GOP Senate could treat bullshit impeachment claims with a swift bullshit acquittal.

Matt Sablan said...

I think the problem is that Silver is... shifty. His numbers very rarely track perfectly with his analysis, and if you pay attention to both, that causes you to notice this weird clash. Like, yeah, AFTER the election, he pointed to where he had Trump as having a chance, etc., etc., but throughout the entire election, his analysis was pretty much about how big of a hill Trump had to climb, how far back he was, etc.

The ultimate problem is I might trust Silver's numbers, but I don't trust him to tell me an unbiased analysis about what those numbers mean. He's useful, but I would hesitate to use him without some caveats.

Anonymous said...

This may be right or wrong. Given everyone's polling performance in 2016 I have to take this with a large grain of salt. Not being a mathematician I am not going to attempt a sophisticated mathematical analysis of 538's results in 2016. Here are a couple of points: Silver predicted that Hillary would win 302 electoral votes, she won 227-a forecast error of 25%; Silver predicted that Trump would win 235 electoral votes he won 306 a companion error of 25%. Silver projected an 85% probability that Clinton would win MI; he projected an 83.5% probability that Clinton would win WI; he projected a 73% probability that Clinton would win PA> We know what happened in all three of those states. He was closer in FL and NC giving Clinton a 55% edge. If Silver were making forecasts for your business how much faith would you put in them?

Rasmussen, who was closest in 2016, has Trump approval at around 46/47% since January. Where the 40% approval Silver uses comes from makes me very suspicious. There is no question the House race is going to be close, but I suspect that once again there are a bunch of quiet R voters who are being undercounted and who will show up to vote. Only election day will tell us.

If I were forecasting the Senate race I would give the Democrats O% probability of taking the Senate. I suppose you mathematicians will tell me why that can't be so, but until then I am sticking to it.

Matt Sablan said...

"The GOP Senate could treat bullshit impeachment claims with a swift bullshit acquittal."

-- True. But, a Democrat House may still insist on attempting to void the last two years of Trump's presidency with pointless laws saying things like: "Trump's appointees' decisions no longer carry the force of law, because Russia." Honestly, I generally want some brakes on both parties, but giving the Democrats either chamber pretty much is accepting *nothing* gets done for two years at the Congressional level.

wwww said...

I'm expecting a Republican pickup in my Minnesota District 1.


Larry Sabato says Minn-1 is a "likely Republican" seat in his Crystal Ball analysis. There are a lot of seats that are classified as tossups or lean more likely to flip before that one.

n.n said...

Mr. Chairman, we do not have a crisis at Freddie Mac, and in particular at Fannie Mae


Perhaps not Waters, but a reasonable facsimile.

n.n said...

I might trust Silver's numbers, but I don't trust him to tell me an unbiased analysis about what those numbers mean

It's climate... political science.

Anonymous said...

If the Dems win the House there may be a lot of Sturm and Drang ( how could there be more?), but the game for the next couple of years is in the Senate: appointments in the Senate; treaties in the Senate; budgets have to be passed by the Senate. A Dem House might make a lot of noise and really piss off a lot of people but they can't be very effective. Look how relatively ineffective the House has been even with an R majority in both houses of Congress.

wwww said...


Like, yeah, AFTER the election, he pointed to where he had Trump as having a chance


He had a tweet storm about this. Huffington post accused him of giving Trump too big of a chance, and he exploded at them:

"After dropping his initial f-bomb, Silver went on to argue why his model -- which, in its polls-only version, puts the odds of Hillary Clinton winning the presidential race at 64.7 percent -- is superior to those like the Huffington Post, which rates her election a near-certainty, at 98.3 percent."
https://www.politico.com/story/2016/11/nate-silver-huffington-post-polls-twitter-230815

Ann Selzer ran a poll that gave Trump a 2 point advantage in Florida at the end of October. It's why I was not surprised on election night. I trust every poll Selzer runs. She's good.

If you doubt pollsters, watch Selzer.

Anyways: Nate Silver's tweetstorm.

This article is so fucking idiotic and irresponsible.
The reason we adjust polls for the national trend is because **that's what works best emperically**. It's not a subjective assumption.
It's wrong to show Clinton with a 6-point lead (as per HuffPo) when **almost no national poll shows that**. Doesn't reflect the data
Not just an issue in elections models. Failure to understand how risks are correlated is part of what led to the 2007/8 financial crisis.
Every model makes assumptions but we actually test ours based on the evidence. Some of the other models are barley even empirical.
There are also a gajillion ways to make a model overconfident, whereas it's pretty hard to make one overconfident.
There's a reasonable range of disagreement. But a model showing Clinton at 98% or 99% is not defensible based on the empirical evidence.
We constantly write about our assumptions and **provide evidence** for why we think they're the right ones.
That's what makes a model a useful scientific & journalistic tool. It's a way to understand how elections work. Not just about the results.
The problem is that we're doing this in a world where people—like @ryangrim—don't actually give a shit about evidence and proof.

The philosophy behind 538 is: Prove it. Doesn't mean we can't be wrong (we're wrong all the time). But prove it. Don't be lazy.

And especially don't be lazy when your untested assumptions happen to validate your partisan beliefs.









wwww said...


Nate tweeted that on November 5, 2016.

Huffington Post complained Nate Silver gave Trump too many chances to win.

Silver gave HRC 64.7% and Huffington Post idiotically had HRC at a 98% probability of winning.

tim in vermont said...

Silver gave HRC 64.7% and Huffington Post idiotically had HRC at a 98% probability of winning.

The HuffPo was operating on what they KNEW to be true, not what they could prove. Just like Mueller.

tim in vermont said...

There was an interesting email in the DNC leak, i think, that explained how to invisibly bias a poll in Florida towards the Democrats. It had to do with focusing your calling on specific regions, for example around Orlando, IIRC, where Democrats were known to be overperforming.

tim in vermont said...

The Catholic Church thought that the Bible was absolute truth, therefore they kept it secret from the masses. Nothing will ever change.

tim in vermont said...

Florence was going to be the biggest monster Cat 5 hurricane to ever hit north of Florida and we were all instructed to quake before its awesome majesty and genuflect to the high priests of global warming. Now it’s a Cat 2.... Just saying.

Achilles said...

wwww said...

The philosophy behind 538 is: Prove it. Doesn't mean we can't be wrong (we're wrong all the time). But prove it. Don't be lazy.

And especially don't be lazy when your untested assumptions happen to validate your partisan beliefs.


538 was 71.4% Clinton on November 8 2016. People are trying to edge them close to 35% but it is a lie.

538 was wrong. Brutally so.

Hypothesis: 538 skews democrat leading up to an election and magically slides towards the truth as the election gets closer.

Data: 538 always gives democrats a greater chance 3 months out than they do 2 months out, 1 month out, 1 week out. etc.

Obvious conclusion: 538 is completely unreliable after they were bought by ABC and started doing the bidding of their owners.

538 is just another tool that sold out to the globalists. They are a tool that does what it is told.

They got paid. Most democrat voters vote for people that want to keep them poor and stupid and violent for free.

Notice how not a single democrat is denouncing the constant violence their side is responsible for as well.

Just pliable tools.

Achilles said...

wwww said...

Nate tweeted that on November 5, 2016.

Huffington Post complained Nate Silver gave Trump too many chances to win.

Silver gave HRC 64.7% and Huffington Post idiotically had HRC at a 98% probability of winning.


Then his owners told him what was up.

He changed it to 71.4% by Nov 8.

Despite the data.

It must have hurt poor old Nate to sell his soul to ABC. Now he does what Disney says.

Achilles said...

People are increasingly realizing that a vote for democrats is a vote for violence and repression.

Facebook is overt in it's fascism.

Another democrat tried to stab a republican candidate. Nothing from the left as far as denunciation.

Google will not work with the US Army. But they work with the Communists in China to create "Social Responsibility Scores."

The mask is off. You pretty much have to be an open enemy of freedom to vote for a democrat.

That will not be a successful election strategy.

rehajm said...

538 fired staffer Roger Pielke because Think Progress pressued Nate et al to fire him because his climate science didn’t conform exactly to the narrative the left wanted to project as settled science. ‘Fuck the data’ said Nate, we have a narative to maintain

hombre said...

How cool. Will Dame Nancy appoint Albee Hastings chair of the House Judiciary Committee?

If the Ds regain the House, more lunacy would certainly be in order.

Jim at said...

What's really going to be fun to watch if - If - the Dems take the House and/or Senate?

For one, he won't be removed from office.
And two, Trump will start cutting deals with them, as he is wont to do.

My feelings would depend upon said deal, whereas the left would go completely apeshit because they were sold out.

And Trump wins 40 states in 2020.

wwww said...



The great thing about analysis, is that you don't have to Give Up. Analysts just keep churning out their reports with the help of R.

There will be a bunch more polls and pollster models submitted to the public before November election day.

We will see in a a few weeks if their predictions are informative & on-the-mark or unhelpful & grossly off-target.

tim in vermont said...

’Fuck the data’ said Nate, we have a narative to maintain

There is a website called “skepticalscience” which is basically pap for propagandists and they left the security settings wrong on an internal discussion and their own climate scientist said that the hockey stick was BS. Their response? They said that they had been “hacked” and nobody had a right to hear that discussion. Basically anybody with an actual browser could have “hacked” it. Kind of like when your password is “password.”

Qwinn said...

"whereas the left would go completely apeshit because they were sold out."

How could you possibly tell?

Jim at said...

Oh, they'd still be apeshit.

But then the Ds would have to face the monster they created, and the 2020 D convention would make '68 look like a picnic.

Matt Sablan said...

Trump offered them nearly everything for considering to fund the wall. They refused. They will never compromise with him.

Michael K said...

Google will not work with the US Army. But they work with the Communists in China to create "Social Responsibility Scores."

China and their allies in the US government are behind a lot of this, if not all. Russia is the squirrel.

China bought the Clintons in 1996.

narciso said...

I related suspicions that Equifax employees had re a possible china hack, up thread,

tim in vermont said...

China and their allies in the US government are behind a lot of this, if not all. Russia is the squirrel.

China bought the Clintons in 1996.


Once they found out how cheap it was to buy a Democrat, there has been no looking back. What do you think Putin was thinking when presumptive president’s spouse shows up at his house for a meeting and a half a million dollars? When he have them 150 million bucks or so?

And here we have the Democrats screaming treason over with not the tinies fraction of the evidence that is right there plain as day against the Clintons.

gilbar said...

breaking news from 1862:
The mask is off. You pretty much have to be an open enemy of freedom to vote for a democrat.
Things don't change, point me to a time the democrats supported freedom?
literacy tests?
jim crow?
russia?

That will not be a successful election strategy.
it's been a pretty successful strategy for most of my life