Embedded Files
Out of 10,000 simulations of the election based on the base average assumptions by state shown below, using standard normal errors for nationwide and state-level results.
State-Based Presidential Election Prediction Model
State-Based Presidential Election Prediction Model
By Kurt Lehmann Last Updated: 10/29/2024
Results:
Map of Average Base Projections of Margin (for Trump)
Map of Average Base Projections of Margin (for Trump) Plus Electoral Votes
Map of Probability for Trump to Win Each State
Chance of Landslide Victory* for Trump:
Chance of Landslide Victory* for Trump:
10%
Chance of Landslide Victory* for Harris:
Chance of Landslide Victory* for Harris:
5%
Chance of Electoral Vote 269-269 Tie:
Chance of Electoral Vote 269-269 Tie:
0.7%
*Landslide Victory entails winning by over 100 Electoral College votes.
Electoral Vote Outcomes Range Distribution
Out of 10,000 simulations of the election with random normal errors applied, this chart shows how frequently each candidate earns an electoral vote total within the displayed 40-EV-wide ranges.
Trump earns between 270 and 309 electoral votes in almost 60% of simulations, whereas Kamala Harris earns that many only about 25% of the time. Trump's path to victory is very clear, and a slim battleground-state edge victory is now most likely. Harris is now clearly losing, but she has a very plausible path to victory yet.
Electoral Vote Margin Outcomes Distribution
Out of 10,000 simulations of the election with random normal errors applied, this chart shows how frequently each candidate earns an electoral vote margin of victory as displayed.
Trump holds the broad majority of winning scenarios, although Harris still holds a decent chance with more evenly distributed outcomes. In simulations where nationally polling underestimates Harris support by 2 or more points, she wins the electoral college about 84% of the time - her path is still very well viable!
Methodology:
This model has two components - the average base projects by state and the full simulation of final outcomes. Base projections are projections of the total vote margin (as a share of total votes) between the two candidates. They are represented as averages between total vote margins based on numerous methodologies, listed below:
Average polling data, adjusted for the recent polling error history (real outcome minus pre-election polling average) in each state.
Recent historical election trends in each state.
Google search trends of key terms (like "voting for Trump" or "Democratic Party") in select states, mainly battleground states.
Trends in voter registration data allocated by party in states that report party affiliation totals.
County-level demographic trends, like population growth, in select states - mainly battleground states.
The model simulation of final outcomes was based on the average base projections, but included normal standard distributed errors (one nationwide, applying to all 50-states, and one state-based, randomly chosen in each state) for each of the 10,000 iterations. This tells us probabilistic representations of various outcomes, like the final winner, the prospect of a 269-269 electoral vote tie, etc.
Google Sites
Report abuse