When is 538 rolling out their maps?

[twenty42]

I thought they usually came out by the beginning of July.

I’m looking forward to seeing what changes Nate makes to his methodology after his titanic miss in 2016. 😂😂

[Hindsight was 2020]

Nate was the only one giving Trump a chance fwi

[twenty42]

Nate was the only one giving Trump a chance fwi

His map was pretty awful and his state-by-state results were way off. RCP’s map was much closer to the actual result.

[Sestak]

Nate was the only one giving Trump a chance fwi

His map was pretty awful and his state-by-state results were way off. RCP’s map was much closer to the actual result.

To be fair, RCP's prediction only looked good because they had a ridiculous amount of tossups (GA? ME-AL? NM? IA?)

[BudgieForce]

Nate was the only one giving Trump a chance fwi

His map was pretty awful and his state-by-state results were way off. RCP’s map was much closer to the actual result.

He can only do so much. State polling was favorable to Hillary so his models were too.

[twenty42]

Nate was the only one giving Trump a chance fwi

His map was pretty awful and his state-by-state results were way off. RCP’s map was much closer to the actual result.

To be fair, RCP's prediction only looked good because they had a ridiculous amount of tossups (GA? ME-AL? NM? IA?)

The no tossup map was actual + NV - MI - WI - PA. Not a slam dunk by any means, but at least they didn’t drink the NC/FL koolaid.

[BudgieForce]

Nate was the only one giving Trump a chance fwi

His map was pretty awful and his state-by-state results were way off. RCP’s map was much closer to the actual result.

To be fair, RCP's prediction only looked good because they had a ridiculous amount of tossups (GA? ME-AL? NM? IA?)

The no tossup map was actual + NV - MI - WI - PA. Not a slam dunk by any means, but at least they didn’t drink the NC/FL koolaid.

FL koolaid?

Florida was decided by a point. Thats the definition of a toss-up.

[Antonio the Sixth]

538 was calibrating the odds right based on the available information. It's not Nate's fault that people don't realize that having a 70% chance of winning means that you'll lose one time out of three.

[Adam Griffin]

I mean, 538's final prediction was 302-235 and 4 states totaling 75 EVs were within 1.2 points...had those 4 went the other way it would have been 307-231. From a fundamentals standpoint/considering most polls have a margin of error of 6 points or more margin-wise, that's not a bad call: the race was just incredibly close in a lot of places and state polls overall obviously had a bias.

[Skye]

State polling indicated Hillary was winning swing states which she actually lost on election day, so 538's model reflected that. It's not like Silver would just go full Trump mode and claim "Fake polls!" and change the model.

538 was actually one of the sites that was more bullish on Trump chances to win. Credit were credit is due.

[The Undefeatable Debbie Stabenow]

To return to the original topic: does anyone know? I follow Silver on Twitter but I don't think I've seen anything from him or anyone else regarding this.

[Sherrod Brown Shill]

To return to the original topic: does anyone know? I follow Silver on Twitter but I don't think I've seen anything from him or anyone else regarding this.

I think they said a couple months ago they were working on a Senate model in one of the podcasts. And I can't imagine them doing a Senate model but not a House model. I figure it will be soon, at the end of the primary season at the latest.

[Xing]

I´m sure they´re quaking in fear of doing a model because they know they´ll never be as good at predictions/analysis as the truly esteemed political scientists such as the OP.

[twenty42]

Nate was the only one giving Trump a chance fwi

His map was pretty awful and his state-by-state results were way off. RCP’s map was much closer to the actual result.

To be fair, RCP's prediction only looked good because they had a ridiculous amount of tossups (GA? ME-AL? NM? IA?)

The no tossup map was actual + NV - MI - WI - PA. Not a slam dunk by any means, but at least they didn’t drink the NC/FL koolaid.

FL koolaid?

Florida was decided by a point. Thats the definition of a toss-up.

Yes, it was a tossup, not Lean D as 538 and Politico were calling it.

[john cage bubblegum]

538 was calibrating the odds right based on the available information. It's not Nate's fault that people don't realize that having a 70% chance of winning means that you'll lose one time out of three.

Really, 538 did about as well as could be expected given the available polling.  State polling in the Midwest did poorly, so of course those states are going to be called incorrectly.  They had FL and NC as toss-ups because, that's what the polls showed. 

But the point is, 538 gave Trump about a 30% chance of winning, which in retrospect seems about right.  They correctly recognized that a polling error/last minute surge concentrated in a key region like the Midwest could give Trump the win.  And Clinton's polling lead in PA, as opposed to WI and MI, was actually quite narrow.

[Pericles]

Yeah, it's as they say "garbage in, garbage out". Nate can have a perfect model but if the polls going in are wrong the model is going to be wrong. He did the best he could in 2016 of showing the state of the race and the uncertainty.