538 Model Megathread

[pbrower2a]

Map showing likelihood of wins for Clinton and Trump, Johnson considered

chance of
win          sat

99%+       9
95-98.9    8
90-94.9    7
80-89.9    6
70-79.9    5
60-69.9    4
55-59.9    3
52-54.9    2

[Higgs]

Nate Silver may suck when it comes to primaries, but he is so far 100/100 in predicting how each state will go in a general election.

Here is the electoral map based on the currently projected odds:

As of right now Arizona, Georgia, Missouri, and North Carolina are the pure toss up states.

Lol none of those other than North Carolina are "pure toss up states". Missouri especially, that's a huge stretch.

[RJEvans]

Junk!

There is no reason to believe that the race is that tight in red states like TN and TX.

The data is right there.

Two polls with Trump at +9 in TN, gives Trump an 8-12 point victory.
Three polls with 3 and 7 point lead for Trump in TX gives him a 6-9 point victory.

This is not Nate's punditry, this is his model spitting out data.

[HAnnA MArin County]

Nonsense. Atlas told me that this race is gonna be much closer. Nate Silver has clearly been paid off by the $hills. He's in the tank for the neoliberal warmongering corporatist.

(Beautiful map though ).

[Queen Mum Inks.LWC]

Other than AZ, it looks ok ...

I'd give AZ to Trump in a close race, because the state's laws that will block Latinos from voting.

In the polls plus forecast, AZ does go to Trump though.

[Cruzcrew]

Utah at R+6 seems too dem optimistic and they actually give Gary Johnson an >8% chance of getting an electoral vote because of those junk swing state polls with him getting 10-15% on average.

[Chief Justice Keef]

Johnson will be lucky to break 3% nationally.

Can we stop dismissing Johnson? He has a legitimate shot at breaking 10% this year, polls are showing a trend around 6-9% already.

[SteveRogers]

Johnson will be lucky to break 3% nationally.

Can we stop dismissing Johnson? He has a legitimate shot at breaking 10% this year, polls are showing a trend around 6-9% already.

Yes, but if he doesn't keep climbing to the 15% that he needs to get into the debates, then it is equally likely that his support collapses in the last month or so of the campaign and he ends up with closer to 3%.

[Ljube]

Nate Silver predicts Trump will never win the Republican nomination!

[pbrower2a]

Making any definitive prediction before the conventions and debates is foolish

Things will change.  So what does it mean that at this point Minnesota has an 85.7% of winning Minnesota and Donald Trump has a 14.0% of winning Minnesota? For now, Nate Silver has made computer estimates of probability of Clinton, Johnson (practically nil), and Trump winning Minnesota.  Trump shows about 2800 results in which he wins, and Clinton shows about 17200 chances of winning Minnesota. Everything can go right for Donald Trump and he could get lucky and win Minnesota. But most likely the calendar runs out before Trump and the GOP has a chance to win the state. Maybe Donald Trump has a rally or two and they prove flops. He has then lost time better used elsewhere. Hillary Clinton plays a nickel defense that allows Donald Trump to lose slowly as a 9% lead shrinks to 5% on Election Day. That's a losing proposition for Donald Trump. Or Donald Trump quits buying ads and making appearances in Minnesota. An 85.7% chance for Clinton becomes a 90% chance, then a 94% chance, then a 96.7% chance, then a 97.2% chance, then effectively no chance in late October or early November. People have no surprise when the networks call Minnesota quickly on Election Night. After all, it is Minnesota, which hadn't gone for a Republican nominee since it was the 49th-best state for Richard Nixon in 1972. Minnesota has a high floor for Republicans in one of their better years (about 49%) and a low ceiling (about 58%) for Democrats.

Can things go so right that Republicans have a chance in Minnesota? Sure. The economy can go into a sudden and severe tailspin, ISIS can suddenly take over Iraq, Hillary Clinton can be caught in a sex scandal, a terrorist act might discredit an inattentive Obama administration... that all takes some imagination. In such a case, many other states swing toward Donald Trump. But the longer that time passes, the less likely such things can happen. Again, time runs out if none of those things happen, and the current estimate that Nate Silver has has gone from an 85.7% chance to a 99.9+% chance of Hillary Clinton winning Minnesota. An 8% lead in late June? That can disintegrate. It can also solidify, and it is practically impossible for an 8% lead to disintegrate in late October and early November.

Now let's try a state that is closer to an even chance of going one way or the other.  Hillary Clinton has about a 41% chance of winning Georgia and Donald Trump has about a 59% chance. That's nearly a coin toss. An 8% lead can disintegrate in 4 months. Obviously, Hillary Clinton does not win Georgia without winning North Carolina (similar demographics) and winning Florida decisively (which a recent poll suggests). Some things are possible. No, not a return to the political realities of the time when Jimmy Carter was Governor!

Georgia suburbs have been unusually strong, for American suburbs, in voting Republican in recent years. If those suburbs start voting like the suburbs of St. Louis, then the Republicans' chance of winning Georgia get wiped out. The white vote in Georgia went heavily R in the last two Presidential elections, perhaps out of distrust of you-know-who. You-know-who isn't on the ballot this time. Let the rural white vote become more like the rural white vote of Kentucky, and Democrats win Georgia. Georgia has a huge presence of active military personnel, and they pay attention to foreign policy. Bad foreign policy can get soldiers killed. Let Hillary Clinton seem too dovish on foreign policy, and she loses Georgia decisively. But let Donald Trump seem too reckless, and many soldiers and their spouses decide to vote for Hillary Clinton.

So guess what happens if Donald Trump tries to force a win in Michigan, which he now has about one chance in 11 of winning and maybe one chance in 33 in early October. He ends up neglecting states like Missouri, Georgia, Arizona, and maybe Indiana that really are in play. He loses a couple of those.  

I'm not saying that any of these scenarios play out. They can, and if they do they can change the reality of the Presidential race more decisively than the winner running out the clock in a political equivalent of the nickel defense.  The one ahead yields ground for time, but not enough to allow a victory to the other side. The team up 24-3 at the half can afford to let the team behind make some ground gains -- but any attempt at a long pass results in an interception that makes things even worse for the team behind. The nickel defense gives a team decidedly behind the choice of how to lose the game.

Hillary Clinton is in the position in which she can play a cautious game and win. Donald Trump may have to make desperate gambles that have a slight chance of winning or can make him lose the Electoral college 450-88 or so.          

[pbrower2a]

Map showing likelihood of wins for Clinton and Trump, Johnson considered

Note: this is likelihood and not margin. Margin may be related to likelihood at this stage.

chance of
win          sat

99%+       9
95-98.9    8
90-94.9    7
80-89.9    6
70-79.9    5
60-69.9    4
55-59.9    3
52-54.9    2

[Mr. Morden]

Clinton's closest states (polls only):

Arizona
North Carolina
Colorado
Ohio
Iowa
Florida <-- Tipping point
Virginia
New Hampshire
Nevada
Pennsylvania

Wouldn't that order make Virginia the tipping point?

[Seriously?]

Nate Silver predicts Trump will never win the Republican nomination!

[Seriously?]

Nate Silver may suck when it comes to primaries, but he is so far 100/100 in predicting how each state will go in a general election.

So what? Most of us that are honest on here could probably hit somewhere between 47 and 51 (states + DC) once the general election rolls around assuming the polling data is somewhat accurate and comprehensive.

Silver will weigh the polls a bit, but from a pure numbers perspective, if you follow the polling enough, you'll do pretty well.

Silver did miss in the off-year election by a seat or two in the Senate, IIRC.

[emailking]

Making any definitive prediction before the conventions and debates is foolish

An 80% chance is not exactly definitive.

[skoods]

The irony of the fact that there's 5 "Who will win in November" polls on Atlas every day and the same people post their predictions in those threads that are roasting Silver now for making a prediction before the convention/5 months early.

[Sir Mohamed]

The map looks realistic, though I doubt that Johnson will be anywhere near 7%. Hillary will break 50%. I suppose she’ll end up with 51-52% of the vote, the Trumpster at 44-45% and Johnson+Stein at around 4%.

[Beefalow and the Consumer]

Clinton's closest states (polls only):

Arizona
North Carolina
Colorado
Ohio
Iowa
Florida <-- Tipping point
Virginia
New Hampshire
Nevada
Pennsylvania

Wouldn't that order make Virginia the tipping point?

Heh.  You're right, my bad.  How'd I bork that one?

Clinton 270.

[Likely Voter]

If you sort the states by Trump's %chance to win, then NH is actually the tipping point at 32.8% vs 27.6% for VA. 

So Trump's narrow path ( sorted by % chance to win) is Romney + OH + CO + IA + FL + NH

[This account no longer in use.]

If you sort the states by Trump's %chance to win, then NH is actually the tipping point at 32.8% vs 27.6% for VA. 

So Trump's narrow path ( sorted by % chance to win) is Romney + OH + CO + IA + FL + NH

TNVol triggered.

[Gass3268]

Making any definitive prediction before the conventions and debates is foolish

The awesome thing about doing it now is you can watch the numbers change was we getting closer to election day.

[Nym90]

If you sort the states by Trump's %chance to win, then NH is actually the tipping point at 32.8% vs 27.6% for VA. 

So Trump's narrow path ( sorted by % chance to win) is Romney + OH + CO + IA + FL + NH

Delving a little deeper he has a better chance of winning ME-02 (currently about 40% according to polls-plus) than most of those states. The really narrow Trump path would be Romney plus OH, CO, IA, FL and ME-02 which gets him to 269 EVs and then the GOP House majority elects him President (or would they?).

Realistically though, I doubt Trump would win Maine's 2nd district if he can't win New Hampshire, although there are reasons to believe the former could be more favorable to him than the latter since it's more rural.

[Deblano]

Nate Silver, you better not  this up and jinx everything. You already made an ass of yourself with the Republican Primaries.

[pbrower2a]

Making any definitive prediction before the conventions and debates is foolish

An 80% chance is not exactly definitive.

Of course. In a best-of-seven playoff series, a team up 3-0 can still lose (assuming that the teams are really equal) the series. The chance of such is 1 in 16, or 6.25% of the time.

Four months remain, and as much can go wrong for Hillary Clinton as it can for Donald Trump. At this stage, if nothing happens except the passing of days, then Hillary Clinton wins. If the same margins existed a month from now, then Hillary Clinton's chances for winning rise while nothing big happens.

Silver's model cannot predict a collapse. It can only show it as it happens.  

[Likely Voter]

Perhaps Clinton going after NE-2 is their backup plan in case Trump wins ME-2. That would be pretty crazy outcome withe the race coming down to NE and ME and their weird rules.  You can bet the NE legislature would be kicking themselves for not going through with the idea of switching to WTA.