re: Why Nate Bronze is full of shite: Using his own data

quote:

---

I don’t understand the hate for 538. Their body of work overall is as good as anyone’s if not better. 2016 doesn’t change that.

---

The point is he missed all the elections that actually meant something.

It's like saying "I'm at 80% predicting Bama games so I'm accurate!" predicting that Bama will beat Little Sisters of the Poor, then completely whiffing on the 20% of games that are actually in question.

quote:

---

think what is going on here is that poll have not caught up with reality.

People don't answer poll calls unless they are highly motivated. Trump voters knowing what happened in the past are not giving those people the time of day. I am not anyway

---

I think it's more than that though. These are no longer the days of landline with no caller idea where you answer the phone no matter what.

Scam calls, advanced caller is etc gives people a really good reason to ignore the call or text.

As well as, I think it takes certain personality trait to want to or not mind participating in polls, whether calls, texts, or online.

It's my belief that that trait trends towards soneone more self absorbed, with more time on their hands. I can't think of a better descriptor but you get the jist

Which all know which political group in referring to. And those people reside in the Republican party as well, so skewing data to party affiliation doesn't solve the problem.

Polls will never be as meaningful and accurate.

Not sure why pollsters don't see this.

quote:

---

All? He called every state right in 2012 didn’t he? And didn’t he do about that well in 2008?

---

see my post above. The further we get technologically, the less accurate polls will be.

quote:

---

hey vol, do you kinda think 538 is underrating gop chances of holding senate?

I think they only give gop 25% shot, but we have a poll out of iowa today showing ernst up 2. There are polls out of NC showing tillis tied or only slightly behind.

win just one of those, win bama, lose Maine, AZ, and CO, and you are at a 50-50 senate.

---

At first glance I thought that they were underrating them, but then I started to look at the map (23 GOP seats up vs. 12 D seats), plus this group's recent history of wave elections (Ds picked up 8 in 2008; GOP picked up 9 in 2014), the polling data, PLUS the fact that a 50/50 tie goes to the power who wins a presidency, and I think that the current 77% for D's seems quite reasonable. That's less than the Economists's 83%.

And according to Nate on the 538 podcast, Senate elections have some advantages due to: less correlation between races which means less systematic polling error than the presidential election; more reliable fundamentals outside of polling (e.g., incumbency, incumbements with multiple elections, partisanship, etc.); and just a much larger sample size (roughly 400 races since 1996 compared to 6 presidential elections).

quote:

---

but we have a poll out of iowa today showing ernst up 2.There are polls out of NC showing tillis tied or only slightly behind.

---

Ernst is looking like a toss-up, but frankly that's a bit surprising given how large her margin was in 2014, just her relative standing within the party (was a superstar at one point; not full on MAGA, but not a never-trumper and just more of a traditional R), and Iowa looking like it was becoming more reliably GOP.

That said, it looks likes there are 3 incumbents who are likely to lose (20% chance or worse): Jones in Alabama, Gardner in Colorado, and McSally in Arizona. So that puts the D's at 48 with a +1 advantage.

Furthermore, there are 6 races where the incumbent as a 65% chance or worse chance of winning; all of those have Republican incumbents. The Ds only need to win 2 of those if Biden wins, and Nate has the Ds favored in 4. That said, they aren't independent of one another (especially two races in Georgia), so it's not a huge margin.

Overall though, as would seem obvious, the House is the most likely to go (remain) blue (98% in 538's model; >99% in The Economist's model) the presidency is the next most likely (89% in 538's model; 95% in The Economist's model); and the Senate is the least likely (77% in 538's model; 83% in The Economist's model). And I think the available data (polling, electoral map, turnout projections, demographics) line up with these projections.

quote:

---

One thing that I keep on forgetting is that even if the senate goes blue, the margin would be pretty thin. Likely 52-48 or 51-49

I don’t see much bold left wing shite getting through with that make up

---

Plus there is a rural/conservative/republican bias in the Senate, so a lot of Democrats have to shift more towards that bias or they'll risk losing their seat (Manchin as one). So a small D margin will be enough to at least keep some functionality in the government but not enough to actually result in any major far-left changes, and HOPEFULLLY maybe end up with some bipartisan and somewhat effective legislation (stimulus, infrastructure, reasonable criminal justice reform).

quote:

---

One thing that I keep on forgetting is that even if the senate goes blue, the margin would be pretty thin. Likely 52-48 or 51-49

I don’t see much bold left wing shite getting through with that make up

---

The problem is that means they own the committees. They could rig an impeachment trial to be incredibly difficult for Trump to defend himself in when the inevitable second impeachment comes through since they would be setting the rules.

51 votes is enough to bring through whatever legislation they want if they keep their members loyal, which they will. There will no longer be a filibuster. If Biden is also President, you better believe the first hundred days of his administration will "fundamentally transform" the nation.

Explain to me then how he rates Trafalgar C- despite having a higher % races called correctly over some A+ like ny times sienna and less margin of error than some A rated pollsters

He ignores data and goes with muh feels

Look at that dude holy frick

That feels more like hope, I understand the idea of millions of Trump voters in hiding and “fooling pollsters” but some evidence would be nice.

That and their body of work (polls) speaks for itself. They’re right way more than they’re wrong. And Trump need them to be really wrong. An error like 16 and Biden still wins. And again, in 2018 they did very well.