2018 Midterm Elections Coverage

Seen something interesting in the news or on the intertubes? Discuss it here.

Moderators: Zamfir, Hawknc, Moderators General, Prelates

Mutex
Posts: 1342
Joined: Wed Jan 09, 2008 10:32 pm UTC

Re: 2018 Midterm Elections Coverage

Postby Mutex » Wed Aug 08, 2018 3:08 pm UTC

CorruptUser wrote:And it looks like Balderson won by 1750 votes, or just under 1%, but O'Conner is contesting.

According to this there's still 8483 outstanding "absentee and provisional" votes to count before they know who won. I don't know when that will happen.

User avatar
Plasma_Wolf
Posts: 113
Joined: Mon Aug 22, 2011 8:11 pm UTC

Re: 2018 Midterm Elections Coverage

Postby Plasma_Wolf » Wed Aug 08, 2018 3:42 pm UTC

Thesh wrote:Well, pretty sure Collins is out. Not sure how competitive his district was before.

Chris Collins arrested for insider trading.

He won his seat in 2016 with a 34.6% margin.

270towin gives the margins for both the house seat and the presidency's vote in that district now. So it was a very safe seat.

User avatar
sardia
Posts: 6347
Joined: Sat Apr 03, 2010 3:39 am UTC

Re: 2018 Midterm Elections Coverage

Postby sardia » Wed Aug 08, 2018 4:39 pm UTC

Plasma_Wolf wrote:
Thesh wrote:Well, pretty sure Collins is out. Not sure how competitive his district was before.

Chris Collins arrested for insider trading.

He won his seat in 2016 with a 34.6% margin.

270towin gives the margins for both the house seat and the presidency's vote in that district now. So it was a very safe seat.

A scandal like that is worth a 20-30 point loss in the polling. I'll have to dig up the 538 link, but they can quantify how much a crime hurts a politician. Of course there's a survivor bias because the really bad ones resign or are arrested.

Tyndmyr
Posts: 11213
Joined: Wed Jul 25, 2012 8:38 pm UTC

Re: 2018 Midterm Elections Coverage

Postby Tyndmyr » Wed Aug 08, 2018 4:48 pm UTC

Agreed. It's interesting, but it's easy to over-interpret from special elections. On the flip side, we still do have democrats doing fairly well overall in special elections. Even if there's a demographic shift in there, a vote's a vote. It's at least somewhat positive.

Trump campaigning for Balderson didn't sink Balderson. Not sure if there's more to draw there. I haven't seen Trump's involvement/approval as significantly positive for Republican candidates, but I might be wrong on that.

I'd be interested in the crime/vote impact assessment. Couldn't find it offhand, but a lot of the 538 articles are tagged with topical things, so they can be difficult to find later.

Edit: Also, agreed with their assessment that Ocasio-Cortez is a worse indicator of the party than Pelosi. None of the candidates endorsed by the former won. The democrat party seems to be winning largely on traditional grounds, with the youth revolution thing being comparatively fringe.

User avatar
sardia
Posts: 6347
Joined: Sat Apr 03, 2010 3:39 am UTC

Re: 2018 Midterm Elections Coverage

Postby sardia » Wed Aug 08, 2018 6:14 pm UTC

Tyndmyr wrote:Agreed. It's interesting, but it's easy to over-interpret from special elections. On the flip side, we still do have democrats doing fairly well overall in special elections. Even if there's a demographic shift in there, a vote's a vote. It's at least somewhat positive.

Trump campaigning for Balderson didn't sink Balderson. Not sure if there's more to draw there. I haven't seen Trump's involvement/approval as significantly positive for Republican candidates, but I might be wrong on that.

I'd be interested in the crime/vote impact assessment. Couldn't find it offhand, but a lot of the 538 articles are tagged with topical things, so they can be difficult to find later.

Edit: Also, agreed with their assessment that Ocasio-Cortez is a worse indicator of the party than Pelosi. None of the candidates endorsed by the former won. The democrat party seems to be winning largely on traditional grounds, with the youth revolution thing being comparatively fringe.

https://fivethirtyeight.com/features/wh ... -race/amp/
Found it, it's worth about 10-12 points. I.e. senator menendez was a thirty point favorite, and the scandal is expected to bring him down to a 20 point favorite, roughly.
I'm not a fan of the fringe vs mainstream narrative because the party itself has moved left. Sure it looks less than authentic when you change positions to match your voters, but it's a stretch to call them the establishment (implying they are as conservative as before.)

Trump has a good track record on his endorsements overall. Most people he endorses in primaries get a bump. I forget how that affects the general midterm election.

Apparently when Nate silver finishes his election model, we should disregard the special election average. He rates his unreleased model as the best predictor of the house election result. I was not amused that he hasn't released his model.

User avatar
CorruptUser
Posts: 10003
Joined: Fri Nov 06, 2009 10:12 pm UTC

Re: 2018 Midterm Elections Coverage

Postby CorruptUser » Wed Aug 08, 2018 6:38 pm UTC

I think most of this amounts to, well, I'm not sure I have the proper word for it. Living in a fantasy, obsessed with some hopeful outcome in a borderline sexual manner, ignoring all alternative interpretations or evidence to the contrary? There has to be a good word for that. "Wankery"? Don't think that's the term. "Thoughtsturbation"?

You have this idea that, oh, the Republicans are doomed, and how even close elections that Republicans win (tentatively), or how a Republican that was caught being a creepy pedophile lost are "proof" that the Republican party is in trouble in 2018. These strike me as a cut of the same cloth for the Media's predictions of Hillary's landslide election, and how everyone (including myself) missed the warning signs because no one could believe that even someone as disliked as Hillary could lose to an Sesame Street villain.
Spoiler:
Seriously? "Built a swamp in a day"?! Are we sure Sesame Street wasn't written by psychics?

Tyndmyr
Posts: 11213
Joined: Wed Jul 25, 2012 8:38 pm UTC

Re: 2018 Midterm Elections Coverage

Postby Tyndmyr » Wed Aug 08, 2018 6:44 pm UTC

That seems about right. I would think that insider trading would be viewed more dimly, but it doesn't seem to be hated as much as other crimes by other people. Such is life, I guess. I do concur that the party itself has moved left somewhat. As long as it's gradual enough, I don't think it'll cause them large PR concerns, but I still view folks like Ocasio-Cortez as further to the left than the majority of the party. Flip side, of course, is that Republicans have also gotten more partisan, so if moving to the extremes *is* a downside, it's not one that is necessarily easy to pick out since both are doing it.

If memory serves, traditional theory goes that the main two parties are drawn close together to capture the largest part of the electorate, but if that's still the case, it's hard to square that with modern partisanship. I'm not sure I have a good answer for this, but I'm going to noodle it about some more, see if a testable theory comes to mind.

The new election model is...intriguing. I don't really know what to think about it without seeing it, but I am curious. Silver's done good modeling work before, but nobody's perfect, yknow?

CorruptUser wrote:I think most of this amounts to, well, I'm not sure I have the proper word for it. Living in a fantasy, obsessed with some hopeful outcome in a borderline sexual manner, ignoring all alternative interpretations or evidence to the contrary? There has to be a good word for that. "Wankery"? Don't think that's the term. "Thoughtsturbation"?


Well, it depends where you get your information from. Low quality media on both sides will happily triumph how such and such a thing is fated, but as you look at more analytic sources, there's a lot more acknowledgement of limitations. 538's live blog is generally far more informal than other writings, and the authors generally don't hide their hopes. This is probably good, the bias is there anyways, might as well put it up front. Still, their methodology and articles are better than most in this area.

Also, they tend to look for reasons behind the trends, which is useful. Poll numbers in a vacuum only tell you so much. Looking at the larger trends, what they follow, and how to influence them is more informative and predictive.

You also get less wishful thinking when actually money is involved. There's still some. 3-6%, if I recall correctly. I ran through a bunch of politicized races on predictit and averaged out market irrationalities found in political markets. It may be larger, but there are possibly others like me who hunt such irregularities for profit. Hard to say for sure. In any case, making bets based on feelings alone is a brutally rough strategy. So, long term, I think prediction markets are of significant value. Players who use poor methodologies grow increasingly less able to manipulate odds.

User avatar
sardia
Posts: 6347
Joined: Sat Apr 03, 2010 3:39 am UTC

Re: 2018 Midterm Elections Coverage

Postby sardia » Wed Aug 08, 2018 8:20 pm UTC

CorruptUser wrote:I think most of this amounts to, well, I'm not sure I have the proper word for it. Living in a fantasy, obsessed with some hopeful outcome in a borderline sexual manner, ignoring all alternative interpretations or evidence to the contrary? There has to be a good word for that. "Wankery"? Don't think that's the term. "Thoughtsturbation"?

You have this idea that, oh, the Republicans are doomed, and how even close elections that Republicans win (tentatively), or how a Republican that was caught being a creepy pedophile lost are "proof" that the Republican party is in trouble in 2018. These strike me as a cut of the same cloth for the Media's predictions of Hillary's landslide election, and how everyone (including myself) missed the warning signs because no one could believe that even someone as disliked as Hillary could lose to an Sesame Street villain.
Spoiler:
Seriously? "Built a swamp in a day"?! Are we sure Sesame Street wasn't written by psychics?

Are you referring to me? I'm pro Democrats but I don't think I ignored any warning signs. Can you elaborate who is claiming Republicans are doomed? Also define doomed.

User avatar
CorruptUser
Posts: 10003
Joined: Fri Nov 06, 2009 10:12 pm UTC

Re: 2018 Midterm Elections Coverage

Postby CorruptUser » Wed Aug 08, 2018 8:37 pm UTC

Yes you, but I think I was projecting a lot onto you from what I've seen elsewhere on here. So, a bit of irony and shame.

At this point, actually, I don't know where I stand on the political spectrum. Everything is so screwy these days.

User avatar
Dauric
Posts: 3900
Joined: Wed Aug 05, 2009 6:58 pm UTC
Location: In midair, traversing laterally over a container of sharks. No water, just sharks, with lasers.

Re: 2018 Midterm Elections Coverage

Postby Dauric » Wed Aug 08, 2018 8:49 pm UTC

CorruptUser wrote:At this point, actually, I don't know where I stand on the political spectrum. Everything is so screwy these days.


Quoted for truth.

I think part of this is that the current political parties are more a function of legacy alliances rather than actual coherent political ideology. Group A and Group B had some specific issue at some point in time that brought them together under the same banner, and the bureaucratic inertia of organization of those groups continued.

IE: Fiscal "get government out of private lives" conservatives aligned with Social "God says we must dictate what happens in your bedroom" conservatives, and despite fairly fundamental disagreements on the role of government in the lives of the citizenry, both factions are strongly represented in the Republican party, or even not-quite independent parties that will typically align heavily with the Republican party.
We're in the traffic-chopper over the XKCD boards where there's been a thread-derailment. A Liquified Godwin spill has evacuated threads in a fourty-post radius of the accident, Lolcats and TVTropes have broken free of their containers. It is believed that the Point has perished.

User avatar
sardia
Posts: 6347
Joined: Sat Apr 03, 2010 3:39 am UTC

Re: 2018 Midterm Elections Coverage

Postby sardia » Tue Aug 14, 2018 7:04 pm UTC

https://fivethirtyeight.com/features/th ... es-so-far/
Endorsements by establishment figures is highly correlated with winning Democratic primaries.
candidates backed by the party establishment are clearly outperforming those backed by more #Resistance-flavored groups. More established interest groups have also tended to have better endorsement records than their activist brethren.that doesn’t mean that the progressive agenda is losing in 2018. The candidates who beat the Sanders-endorsed El-Sayed and Welder? Gretchen Whitmer supports a $15 minimum wage and the legalization of marijuana, and Sharice Davids is a gay Native American woman campaigning on a child-care tax credit and treating gun violence as a health care crisis. As the data shows, “establishment vs. progressive” is a bit of a false dichotomy.

I find it odd how maligned the establishment party is for Democrats even though they aren't any less progressive than the average progressive outsider group.

User avatar
ObsessoMom
Nespresso Bomb
Posts: 733
Joined: Sun Sep 29, 2013 5:28 pm UTC

Re: 2018 Midterm Elections Coverage

Postby ObsessoMom » Tue Aug 14, 2018 8:49 pm UTC

Rep. Duncan Hunter's chances of hanging on to his seat may go up in smoke

A bit of an ironic headline, as "seat" can also be a synonym for "butt." And he's been desperately trying to save his own butt for the past year or so, due to a number of well-publicized campaign finance scandals.

[Edited to say: Apologies, this isn't a new article--it's an April article with a recent update. He has since won the June primary, and will appear on the November ballot against a Democratic challenger. Anyway.]

I hope he gets indicted soon. He's a ginormous jerk who has barely even had to campaign in order to get elected, because he's riding on his daddy's name recognition. (He's Duncan Hunter, Jr., and Duncan Hunter, Sr., represented the same congressional district for decades. I bet a lot of older Republican voters haven't even noticed that they've been electing a different person lately when they vote for "Duncan Hunter." Which I think is how Junior won the June primary by such a ridiculously high margin, despite all the bad press. A lot of voters just aren't paying attention.)

User avatar
sardia
Posts: 6347
Joined: Sat Apr 03, 2010 3:39 am UTC

Re: 2018 Midterm Elections Coverage

Postby sardia » Thu Aug 16, 2018 5:14 pm UTC

https://projects.fivethirtyeight.com/20 ... ouse/#maps
Nate's long awaited model for the House election is finally online. ]
The layers in FiveThirtyEight’s House forecast
WHICH VERSIONS USE IT?
LAYER DESCRIPTION LITE CLASSIC DELUXE
1a Polling District-by-district polling, adjusted for house effects and other factors. ✓ ✓ ✓
1b CANTOR A system which infers results for districts with little or no polling from comparable districts that do have polling. ✓ ✓ ✓
2 Fundamentals Non-polling factors such as fundraising and past election results that historically help in predicting congressional races. ✓ ✓
3 Expert forecasts Ratings of each race published by the Cook Political Report, Inside Elections and Sabato’s Crystal Ball ✓
Democrats’ chances of winning the House are between 7 in 10 and 3 in 4 in the various versions of the model upon launch — right about what Hillary Clinton’s chances were on election night two years ago! — so ignore those probabilities at your peril.

His metrics are crazy, partially out of necessity. House polling is limited and often bad. Also, I feel Nate is tempting fate here, or is trolling Democrats.

Tyndmyr
Posts: 11213
Joined: Wed Jul 25, 2012 8:38 pm UTC

Re: 2018 Midterm Elections Coverage

Postby Tyndmyr » Thu Aug 16, 2018 5:59 pm UTC

That is... less advantageous for democrats than I expected it to be. I'm not sure that this is wrong, but it essentially has democrats winning by a fairly narrow margin, and a fair number of soft seats. A fairly modest change can upend the result.

Given the quality of the polling, as Silver mentions, that's...actually quite possible. Mmm. I've been going light on bets for who takes the house, and I might need to rethink that. Have to dig through data for some of those tossup areas to see how I feel before doing that, but it's an area where a slight informational edge might provide a great relative advantage.

I don't put a lot of stock in the expert opinion. It doesn't improve accuracy significantly, and any advantage might just be curve fitting. Classic model seems like the most solid model of the three he presents.

Gonna be interesting to see how the forecast develops as we get more/better information.

User avatar
cyanyoshi
Posts: 386
Joined: Thu Sep 23, 2010 3:30 am UTC

Re: 2018 Midterm Elections Coverage

Postby cyanyoshi » Thu Aug 16, 2018 8:23 pm UTC

sardia wrote:https://projects.fivethirtyeight.com/2018-midterm-election-forecast/house/#maps
Nate's long awaited model for the House election is finally online. ]
The layers in FiveThirtyEight’s House forecast
WHICH VERSIONS USE IT?
LAYER DESCRIPTION LITE CLASSIC DELUXE
1a Polling District-by-district polling, adjusted for house effects and other factors. ✓ ✓ ✓
1b CANTOR A system which infers results for districts with little or no polling from comparable districts that do have polling. ✓ ✓ ✓
2 Fundamentals Non-polling factors such as fundraising and past election results that historically help in predicting congressional races. ✓ ✓
3 Expert forecasts Ratings of each race published by the Cook Political Report, Inside Elections and Sabato’s Crystal Ball ✓
Democrats’ chances of winning the House are between 7 in 10 and 3 in 4 in the various versions of the model upon launch — right about what Hillary Clinton’s chances were on election night two years ago! — so ignore those probabilities at your peril.

His metrics are crazy, partially out of necessity. House polling is limited and often bad. Also, I feel Nate is tempting fate here, or is trolling Democrats.


I think he's just reminding readers that probabilities are, well, probabilities. They may want to pump the brakes with the idea of a near-certain Democratic wave in November. Republicans holding onto the House is also a likely scenario, and it would be foolish to dismiss that possibility. We don't need a repeat of the widespread 2016 reporting that Donald Trump was toast, and that Clinton had the presidency in the bag.

User avatar
sardia
Posts: 6347
Joined: Sat Apr 03, 2010 3:39 am UTC

Re: 2018 Midterm Elections Coverage

Postby sardia » Sat Aug 18, 2018 12:58 pm UTC

Tyndmyr wrote:That is... less advantageous for democrats than I expected it to be. I'm not sure that this is wrong, but it essentially has democrats winning by a fairly narrow margin, and a fair number of soft seats. A fairly modest change can upend the result.

Given the quality of the polling, as Silver mentions, that's...actually quite possible. Mmm. I've been going light on bets for who takes the house, and I might need to rethink that. Have to dig through data for some of those tossup areas to see how I feel before doing that, but it's an area where a slight informational edge might provide a great relative advantage.

I don't put a lot of stock in the expert opinion. It doesn't improve accuracy significantly, and any advantage might just be curve fitting. Classic model seems like the most solid model of the three he presents.

Gonna be interesting to see how the forecast develops as we get more/better information.

Nate was more bearish then you, he was surprised that the Democrats had such a high chance in the model.
Given the lack of data expert opinion is probably the way to go especially combined with his model. But they're testing their model this year, so we will find out just how good it is.
His advice about fundraising really stood out to me.
1. Small dollar Fundraising matters. Super PAC does not. You're just burning money at that point. Especially given the favored rates nonpac spending gets for air time.
2. You'd get more impact on elections donating to local journalism than you would to a candidate. The last one might be bias but he brings up a good point about just how wasteful PACs have been. It's half a scam against rich people now.

Tyndmyr
Posts: 11213
Joined: Wed Jul 25, 2012 8:38 pm UTC

Re: 2018 Midterm Elections Coverage

Postby Tyndmyr » Mon Aug 20, 2018 2:44 pm UTC

sardia wrote:His advice about fundraising really stood out to me.
1. Small dollar Fundraising matters. Super PAC does not. You're just burning money at that point. Especially given the favored rates nonpac spending gets for air time.
2. You'd get more impact on elections donating to local journalism than you would to a candidate. The last one might be bias but he brings up a good point about just how wasteful PACs have been. It's half a scam against rich people now.


That makes sense to me. PACs, in general, have been an exercise in confused causality. Getting a ton of support from small backers is a really good indication of support in general. If someone's willing to throw money your way, they're willing to vote for you, most likely. But someone can only vote once, no matter how much they're willing to donate.

So yeah, all else being equal, the guy raising more money generally has odds to win. But any given money amount raised in small donations ought to be far more predictive than the same amount raised from large donors. Sure, money has power for advertising and what not, but number of donors is a good indicator of your message resonating.

Also, local journalism is probably terribly easy to influence with money. Most newspapers and the like are not super profitable these days. Even a modest donation could be a big deal for them. It makes sense that they're growing more cost-efficient than the traditional big budget options. All of the above might be really good news for a candidate like Sanders. It fits his style more so than it fits someone like Trump, giving him a more hopeful look at 2020.

User avatar
sardia
Posts: 6347
Joined: Sat Apr 03, 2010 3:39 am UTC

Re: 2018 Midterm Elections Coverage

Postby sardia » Mon Aug 20, 2018 3:39 pm UTC

These metrics only apply to House elections. Presidential elections are a different ball game. Unless you're talking about Bernie's Senate campaign.

Harry Enten notes that Trump is cutting into his deficit with black people. Given how much they vote, and how concentrated they are in swing states, this improvement is significant for 2020. https://mobile.twitter.com/ForecasterEn ... 6028650496
know this will sound crazy to some of you, but I really do think Trump has gained a few points with black voters since the election. (He's still in the low to mid 10s

In 2016, Trump managed better numbers than Romney for blacks. (Could be Obama, could be Hillary super predators, Trump propaganda, or something else) since 2016, Trump has continued to improve. Effect size of this change would be enough to give Trump another 1 point in a tight election.

Tyndmyr
Posts: 11213
Joined: Wed Jul 25, 2012 8:38 pm UTC

Re: 2018 Midterm Elections Coverage

Postby Tyndmyr » Mon Aug 20, 2018 4:15 pm UTC

sardia wrote:These metrics only apply to House elections. Presidential elections are a different ball game. Unless you're talking about Bernie's Senate campaign.


Don't think there's potential crossover there? The presidential is a different ballgame in some respects, sure, but if PACs are wasteful, I'd imagine that they'd still be so in presidential campaigns.

Tyndmyr
Posts: 11213
Joined: Wed Jul 25, 2012 8:38 pm UTC

Re: 2018 Midterm Elections Coverage

Postby Tyndmyr » Tue Aug 21, 2018 5:02 pm UTC

https://reason.com/blog/2018/08/20/shock-poll-gary-johnson-at-21-in-senate

Gary Johnson polling at twice the support of the Republican in his senate race. Libertarians call for the Republicans to stop running a spoiler candidate.

The Democrat candidate, Martin Heinrich, is clearly in the lead with 39% polling. I'm not sure what broader conclusions can be drawn from this for the Republicans, but...probably nothing good for them. They seem unlikely to take that seat, at minimum.

User avatar
ObsessoMom
Nespresso Bomb
Posts: 733
Joined: Sun Sep 29, 2013 5:28 pm UTC

Re: 2018 Midterm Elections Coverage

Postby ObsessoMom » Wed Aug 22, 2018 6:49 am UTC

ObsessoMom wrote:I hope he [Duncan Hunter Jr. of California's 50th Congressional District] gets indicted soon.

The Los Angeles Times wrote:“The indictment alleges that Congressman Hunter and his wife repeatedly dipped into campaign coffers as if they were personal bank accounts, and falsified FEC campaign finance reports to cover their tracks,” said U.S. Atty. Adam Braverman. “Elected representatives should jealously guard the public’s trust, not abuse their positions for personal gain. Today’s indictment is a reminder that no one is above the law.”

http://www.latimes.com/politics/la-pol- ... story.html

:)

I'm not the only one feeling some schadenfreude about this today:

https://twitter.com/RepAdamSchiff/statu ... 6587971585

User avatar
sardia
Posts: 6347
Joined: Sat Apr 03, 2010 3:39 am UTC

Re: 2018 Midterm Elections Coverage

Postby sardia » Wed Aug 22, 2018 7:37 pm UTC

Tyndmyr wrote:Don't think there's potential crossover there? The presidential is a different ballgame in some respects, sure, but if PACs are wasteful, I'd imagine that they'd still be so in presidential campaigns.

Sure there's crossovers, but what if fundraising is merely a ok proxy due to lack of polling (gold standard)? Or there's diminishing returns on money, which presidential campaigns blow past. That's why I think we shouldn't confuse metrics.
Lastly, there's no guarantee that the path back to democratic power lies with white working class. Aka Bernie could end up running in a lane/coalition that doesn't exist or is tiny.
It's very possible Democrats come back with suburbans elites+females+blacks. How can Bernie's message compete with that? The map is wide open, and coalitions can form anywhere, (or even tilt further towards Republicans).

Tyndmyr
Posts: 11213
Joined: Wed Jul 25, 2012 8:38 pm UTC

Re: 2018 Midterm Elections Coverage

Postby Tyndmyr » Wed Aug 22, 2018 7:46 pm UTC

That's fair. Diminishing returns do seem likely. With the length of modern presidential campaigns, it seems likely that a saturation point is hit for the big two candidates.

May be less true for primaries, but yeah, I don't have hard data to say for sure one way or the other. Gut feeling is that the more minor candidates are not hitting diminishing returns much, but the larger ones, maybe? Both Bernie and Clinton had pretty strong publicity last go-round.

I think Bernie's main hangup is an appeal to youth. It's a demographic that has turnout issues. Yeah, he's got his core believers, but he definitely needs to broaden his appeal in order to make it. Working class is a really good core demographic, and at least some of his core will be older and more experienced this time around, but so far, yeah, he's played to a very specific group, and has much less pull outside of that.

I wouldn't be surprised if the 2020 nominee is someone else who plays a more mainstream Democratic role.

User avatar
CorruptUser
Posts: 10003
Joined: Fri Nov 06, 2009 10:12 pm UTC

Re: 2018 Midterm Elections Coverage

Postby CorruptUser » Wed Aug 22, 2018 7:53 pm UTC

Elizabeth Warren? Please? Dont always agree with her but I can respect her...

Tyndmyr
Posts: 11213
Joined: Wed Jul 25, 2012 8:38 pm UTC

Re: 2018 Midterm Elections Coverage

Postby Tyndmyr » Wed Aug 22, 2018 8:00 pm UTC

Maybe? She seems to come up occasionally as a possible contender, but this far out, I don't feel super confident in calling anyone in particular.

User avatar
ObsessoMom
Nespresso Bomb
Posts: 733
Joined: Sun Sep 29, 2013 5:28 pm UTC

Re: 2018 Midterm Elections Coverage

Postby ObsessoMom » Thu Aug 23, 2018 3:31 pm UTC

Unsurprisingly, Duncan Hunter is unrepentant, and is following Trump's tried-and-true "the best defense is a good offense" playbook:

Rep. Duncan Hunter statement: Prosecution a witch hunt

Some may find this amusing:

The 30 most astounding moments in Rep. Duncan Hunter's 47-page indictment

But where's dat wascally wabbit? I'm disappointed to see the bunny incident mentioned only as taking the family pet on vacation.

User avatar
sardia
Posts: 6347
Joined: Sat Apr 03, 2010 3:39 am UTC

Re: 2018 Midterm Elections Coverage

Postby sardia » Wed Aug 29, 2018 9:39 pm UTC

https://projects.fivethirtyeight.com/20 ... /1/#deluxe
Nate has posted individual ratings, along with the components that add up to 538's predictions for the House on the individual level. You can enter your state, or use wikipedia to find your district. Mine has a 25% chance of Democratic victory :( Lame. I would have to expect Democrat's generic polling to increase by a couple points before my district is likely to flip. Aka, within the margin of error, but not likely. I had thought people would hate Roskam more, he's a real Trump slave/supporter, but I guess that's not bad around here.

Tyndmyr
Posts: 11213
Joined: Wed Jul 25, 2012 8:38 pm UTC

Re: 2018 Midterm Elections Coverage

Postby Tyndmyr » Thu Aug 30, 2018 4:37 pm UTC

Every single Maryland race is 99-1%.

This, honestly, seems correct. It's not very exciting, and doesn't really give much opportunity for interesting insights, but it is what it is, I guess.

Seems to be house only, but it looks like we're gonna keep our Republican governor, despite being a state that's blue as hell.* Odd. I can't fathom why they ran Jealous. He's the sort of candidate you pick when you know you can't possibly win. Hogan's popular, but a state this blue, you'd think they could find a better candidate.

*http://www.baltimoresun.com/news/maryland/politics/bs-md-hogan-jealous-poll-20180813-story.html

User avatar
sardia
Posts: 6347
Joined: Sat Apr 03, 2010 3:39 am UTC

Re: 2018 Midterm Elections Coverage

Postby sardia » Tue Sep 04, 2018 2:43 pm UTC

Ugh, the three models have started to diverge.
Lite - 42%
Classic - 28%
Deluxe - 38%.
Overall, this is a modest improvement from before. The polling in the district got worse, but the surrounding polling (CANTOR) got better. The fundamentals still mostly favors the Republican, but the experts are bullish.
I'll check back next month for reversion to the mean. I hope roskam loses before I move out of his district.

User avatar
sardia
Posts: 6347
Joined: Sat Apr 03, 2010 3:39 am UTC

Re: 2018 Midterm Elections Coverage

Postby sardia » Thu Sep 13, 2018 3:53 pm UTC

https://projects.fivethirtyeight.com/20 ... rms-header
The Senate forecast is finally here. Democrats are defending the best they can, but have an awful map.
Democrats are at 33% odds to win the Senate which is considered an improvement since 2017. Nate expects generic ballot will need to exceed 10-11 points before Democrats are favored to win the Senate. Right now they are only up 8-9 points.

I'm worried about Bill Nelson in Florida. Right now the model is only saying he's even because the fundamentals say he should have an advantage. The polling shows Nelson losing to Rick Scott. Let's not even talk about Heidi...
In personal news, Illinois 6th got better for Democrats, but still Republicans are favored.
Lite 54% Dem
Classic 33%
Deluxe 42%.

User avatar
Plasma_Wolf
Posts: 113
Joined: Mon Aug 22, 2011 8:11 pm UTC

Re: 2018 Midterm Elections Coverage

Postby Plasma_Wolf » Thu Sep 13, 2018 5:55 pm UTC

I wonder how Rick Scott is so popular... Could anyone explain?

There is one thing about this map: I can't see the incumbent party in the Senate race. It's very nicely done in the house race, but in the Senate race that's missing.

Trebla
Posts: 373
Joined: Fri Apr 02, 2010 1:51 pm UTC

Re: 2018 Midterm Elections Coverage

Postby Trebla » Thu Sep 13, 2018 7:26 pm UTC

Plasma_Wolf wrote:I wonder how Rick Scott is so popular... Could anyone explain?


Same reasons Republicans are popular anywhere... they appeal to a conservative base.

Also, Florida's economy (like most of the nation) has been on an upswing since the recovery really got underway in ~2010 or so.. Scott became governor in early 2011. Want to guess if he takes credit?

We're also the only state (I think) dumb enough to elect a member of congress who was impeached (and convicted) for bribery as a judge, so, don't expect people here to be rational with their voting.

Tyndmyr
Posts: 11213
Joined: Wed Jul 25, 2012 8:38 pm UTC

Re: 2018 Midterm Elections Coverage

Postby Tyndmyr » Fri Sep 14, 2018 2:10 pm UTC

Plasma_Wolf wrote:I wonder how Rick Scott is so popular... Could anyone explain?

There is one thing about this map: I can't see the incumbent party in the Senate race. It's very nicely done in the house race, but in the Senate race that's missing.


Incumbents often manage to accrue a significant popularity advantage. I'm not as familiar with Florida, but despite MD being a hardcore blue state, we've got a republican governor with an unassailable popularity advantage. For him, it's mostly being a decent person, and providing fairly moderate governance, plus a bit of the usual politicking. The incumbent gets a lot of popularity opportunities, so long as they don't make a mess of things, that's generally advantageous. I'd assume that something similar is the case for Rick Scott, albeit to a lesser degree.

I'm having trouble imagining the Dem's taking the senate. The silver lining for them is that if they do pull it off, they almost certainly will have the house too. The lite "2 in 7 odds" seems most accurate to me. Deluxe doesn't seem to change the overall odds much, just twerk specific outcomes. As with congress, I'm not sure I see value in the expert opinions.

User avatar
sardia
Posts: 6347
Joined: Sat Apr 03, 2010 3:39 am UTC

Re: 2018 Midterm Elections Coverage

Postby sardia » Fri Sep 14, 2018 3:03 pm UTC

Tyndmyr wrote:Incumbents often manage to accrue a significant popularity advantage. I'm not as familiar with Florida, but despite MD being a hardcore blue state, we've got a republican governor with an unassailable popularity advantage. For him, it's mostly being a decent person, and providing fairly moderate governance, plus a bit of the usual politicking. The incumbent gets a lot of popularity opportunities, so long as they don't make a mess of things, that's generally advantageous. I'd assume that something similar is the case for Rick Scott, albeit to a lesser degree.

I'm having trouble imagining the Dem's taking the senate. The silver lining for them is that if they do pull it off, they almost certainly will have the house too. The lite "2 in 7 odds" seems most accurate to me. Deluxe doesn't seem to change the overall odds much, just twerk specific outcomes. As with congress, I'm not sure I see value in the expert opinions.

You're explanation is incomplete. For one thing, Nelson is the incumbent, Rick Scott is the Challenger. At the very least, both are incumbents.
According to the Classic breakdown on 538, Nelson's polls are even(high weight), Cantor is +10 Dem (minimal weight), and fundamentals is +7 Dem(moderate weight). That leads to a weighted average of +1.3 Dem, hence slightly favored.
The polls only lite shows+0, so even odds.

The Senate is composed of only a few competitive races. Which one do you disagree with?
Donnelly in Indiana being favored 64% lite vs 76% classic
Bresden in TN 50% lite vs 30% classic
Beto in TX 33% in both models
Manchin WV 80+%
Epsy in MS 25% or 16%
Heidi ND 39% lite vs 59% classic?

Tyndmyr
Posts: 11213
Joined: Wed Jul 25, 2012 8:38 pm UTC

Re: 2018 Midterm Elections Coverage

Postby Tyndmyr » Fri Sep 14, 2018 4:54 pm UTC

Ah, good point. I got a bit overly focused on the governor aspect. 'tis a bit of an incumbent vs incumbent battle in practice.

Heidi's probably the race with the most significant swing between them. I think it's actually a pretty even tossup between them, so...somewhere between the lite and classic in that case. Granted, we're making really fine distinctions at this point.

I also don't think Mississippi is actually going to elect a democrat. The only reason it's close right now is because you've got two republican candidates pulling down significant shares of the vote. The republican challenger will lose, and the republican incumbent will stay as such. I'd put that at 90%+ probability of happening, and frankly, the difference between R +14.8 and R +16.3 is mostly unimportant. Unless something really odd happens here, I wouldn't even consider this race competitive.

Basically, when moving up the rankings, I worry that we're taking some of the same effects into account multiple times. Fundamentals ought to be reflected in polling to some degree, generalized trends shown in polling also ought to show up in CANTOR...and the experts are basically making their decisions based on all of the above. It's possible for an effect to be magnified as a result. The tools are useful, mind, it's just something to consider when looking at the results.

User avatar
sardia
Posts: 6347
Joined: Sat Apr 03, 2010 3:39 am UTC

Re: 2018 Midterm Elections Coverage

Postby sardia » Tue Sep 18, 2018 6:01 pm UTC

https://fivethirtyeight.com/features/wi ... blue-wave/
Here's a scary look under the hood for Democratic Senate chances. The odds for Senate is better than the polling indicates, but the model will discount fundamentals more and more as election night approaches in favor of polling. Unless polling improves, several Democratic incumbents will have worse chance of winning.

It's interesting that the model assumes polling is sightly off, and is giving Democrats time for polling to catch up to the fundamentals. Unless it doesn't, and then Democrats chances will decrease.
Other tidbit, model doesn't account for candidate quality inn a previous election. If you run up the table against Todd legitimate rape Akins, the model assumes you're a godlike incumbent instead of really lucky. *

It feels weird to me that the only reason the model rates so many higher is that fundraising indicates broad support. That's actually the main reason Heidi is a slight favorite. I kinda agree with you, but I'm curious if it's an accurate predictor.

Is there really a big difference between something happening 15% of the time instead of %10? It feels the same to us, aka rarely. Like imagine if Mueller or the economy takes a dump on Republicans next month. You couldn't imagine the generic ballot going from+9ish to+15?

*There are rumors that Democrats gave free media exposure to the worst GOP candidate in order to win. So the model isn't entirely wrong here.

Tyndmyr
Posts: 11213
Joined: Wed Jul 25, 2012 8:38 pm UTC

Re: 2018 Midterm Elections Coverage

Postby Tyndmyr » Tue Sep 18, 2018 7:57 pm UTC

sardia wrote:Other tidbit, model doesn't account for candidate quality inn a previous election. If you run up the table against Todd legitimate rape Akins, the model assumes you're a godlike incumbent instead of really lucky. *


Yeah, that's definitely a factor. There's definitely a sense of easier/harder races based on opponent, and I feel like there ought to be a way to quantify that apart from partisan lean. Perhaps it's less of a factor now, given that partisanship is so strong, but the individual still matters to some degree.

Is there really a big difference between something happening 15% of the time instead of %10? It feels the same to us, aka rarely. Like imagine if Mueller or the economy takes a dump on Republicans next month. You couldn't imagine the generic ballot going from+9ish to+15?


In a sense. It's hard to tease out on a single event, because both feel like fairly rare events. However, they will feel differently depending on your partisan lean. In video game design, specifically Civilization, designers found that players expectations were not always in line with statistics. If a fight was one that they were predicted to win 90% of the time, they expected to win, and felt robbed if they did not. Fairly unlikely fights they didn't feel robbed if they lost, but they did not exhibit a similar reaction to the 10% case arising. Biases seem to affect how people process statistics for their identified side.

I *believe* both 10 and 15% are low enough that similar expectations exist. Folks on the winning side are tempted to over-represent it as a sure thing.

But if we're talking gambling odds, it's a significant difference in chance of winning, and being 5% more accurate than the other guy at repeated bets means you'll likely take his money. This is, in practice, a large part of how I make money off of PredictIt. Wishful thinking doesn't go well with betting.

ijuin
Posts: 819
Joined: Fri Jan 09, 2009 6:02 pm UTC

Re: 2018 Midterm Elections Coverage

Postby ijuin » Tue Sep 18, 2018 9:02 pm UTC

The difference between the gambling odds and the election odds in practice is that you generally can gamble again and again in a short period of time, whereas elections happen every two years or less. Imagine that you can make only thirty bets in your lifetime, making one every two years, and the mathematics become more comparable.

Tyndmyr
Posts: 11213
Joined: Wed Jul 25, 2012 8:38 pm UTC

Re: 2018 Midterm Elections Coverage

Postby Tyndmyr » Wed Sep 19, 2018 9:21 pm UTC

ijuin wrote:The difference between the gambling odds and the election odds in practice is that you generally can gamble again and again in a short period of time, whereas elections happen every two years or less. Imagine that you can make only thirty bets in your lifetime, making one every two years, and the mathematics become more comparable.


It's a very different timescale, certainly.

That said, political events are happening nigh-constantly. It's enough to bet on routinely, and it means that better statistics ultimately grant a significant edge. There's a lot of room for improvement in the political process due to more accurate prediction. Even if it's only a few percent more accurate, that can make a good deal of difference. The more accurately you know the odds for races, the more efficiently you can allocate resources. There's strong potential for one national party to make gains in this way. Currently, I'd credit the Democrats with better data, and I think it does give them at least a slight edge.

User avatar
pogrmman
Posts: 524
Joined: Wed Jun 29, 2016 10:53 pm UTC
Location: Probably outside

Re: 2018 Midterm Elections Coverage

Postby pogrmman » Fri Sep 21, 2018 11:33 pm UTC

I’m watching the O Rourke/Cruz debate right now (because, you know, I’m in TX and am a voter) and Cruz is frankly pissing me off. He’s just basically ignored all the questions and all of O Rourke’s statements and returned to standard talking points. For instance, in the very first question about Dreamers, Cruz entirely ignored the part about Dreamers and takes about illegal immigrants in general — not focusing on the fact that they often were brought over by their parents instead of doing it themselves. They’re both walking all over the moderators, too — the moderators seem nervous, too.


Return to “News & Articles”

Who is online

Users browsing this forum: No registered users and 18 guests