Final Presidential Forecast 2024

With Popular Vote Forecast

You can see my previous presidential forecast here which includes links to previous forecasts, analysis, and snark.

That Selzer poll with toplines Harris 47% and Trump 44% in Iowa is (probably) going to be an absolute blast for me to shove in the faces of the unscientific analysts who literally believe that her poll, if accurate, predicts that Harris will win by 3 in Iowa.

But we'll save that for after the election.

For now.

Forecast Comparison

RealCarlAllen: 67% Democrats win

Nate Silver: 49% Democrats win

FiveThirtyEight: 48% Democrats win

I don't care if Nate pulls a Pennsylvania Senate 2022 and moves his forecast 14 points overnight again. Not considering it.

Range of Outcomes

There are two ways I have illustrated the range of likely (and unlikely) outcomes in my forecast. Note the x-axis for Electoral Vote bins.

This second visual is the one comparable to prediction markets - it uses the same range of outcomes (with the exception of the furthest extremes, which I've combined into one bin)

I've explained ad nauseum that the Democrats, though only narrowly “ahead” in poll averages in the blue wall, are in a better position among undecideds/late deciders than others seem to believe. See previous forecast posts for more detail.

That tall bar for Harris is a bit misleading visually because three of the five most common outcomes pile into that 302-331 bin:

308, 319, 303.

276 and 292 follow closely behind. The 272-276 bin is almost exclusively that exact 276 outcome.

Trump's most common victories are a smattering of blue wall cracks: only Pennsylvania, Wisconsin and Pennsylvania, with some “all of the above” mixed in.

I disagree with other forecasters that a swing state sweep is Trump's most common victory, but that's a debate for a different post.

Battlegrounds and Beyond

These are the states that have win probabilities worth mentioning - from the swing states to the “stretch” states.

Remember, because win probabilities are correlated, that does not mean that my forecast thinks the most likely outcome is that Harris wins the states where she's favored, and loses where she's not. See the range of outcomes for that.

Popular Vote Forecast

See this post for an explanation of the visual

National Popular Vote (85% chance Harris wins)

---

The Swing States

Nevada (55% Chance Harris wins)

North Carolina (44% Chance Harris wins)

Georgia (38% Chance Harris wins)

Arizona (35% Chance Harris wins)

Wisconsin (66% Chance Harris wins)

Pennsylvania (70% Chance Harris wins)

Michigan (72% Chance Harris wins)

---

The Stretch States (and Districts)

Maine 2nd (22% Chance Harris wins)

Iowa (12% Chance Harris wins)

Florida (12% Chance Harris wins)

Nebraska 2nd (88% Chance Harris wins)

Minnesota (90% Chance Harris wins)

Texas (8% Chance Harris wins)

New Hampshire (93% Chance Harris wins)

Ohio (6% Chance Harris wins)

Virginia (95% Chance Harris wins)

Support my work and learn about how polls work (and how they don't work) by buying my book.

Thanks for the support.