Probabilistic forecast of the 2020 US Presidential election from 538

A dataset containing the predicted probabilities for each major party candidate to receive different numbers of electoral votes in the 2020 US Presidential election according to FiveThirtyEight as of October 16, 2020.

Format

A data frame with 539 rows and 11 variables:

cycle

The election cycle: 2020

branch

The kind of race this forecast pertains to: "President"

model

The model type: "polls-plus"

modeldate

Date of the model run, as a Date object: "2020-10-16"

candidate_inc

Name of the incumbent: "Trump"

candidate_chal

Name of the challenger: "Biden"

evprob_inc

Chance that the incumbent wins total_ev electoral votes: numeric between 0 and 1

evprob_chal

Chance that the challenger wins total_ev electoral votes: numeric between 0 and 1

total_ev

Number of electoral votes in question: integer between 0 and 538

timestamp

Date and time the simulations were run, as a POSIXct object: "2020-10-16 21:08:11 EST"

simulations

Number of simulations run: 40000

Source

This dataset contains the electoral vote output of FiveThirtyEight's 2020 US Presidential Election Forecast as of Oct 16, 2020. The original dataset is licensed under the CC BY 4.0 license. The dataset was obtained from the FiveThirtyEight website, and its metadata description is based on the dataset description in the FiveThirtyEight Github repository. The dataset has been only been slightly modified: date formats were converted from strings into R date objects; empty columns related to third party candidates were dropped; and the data frame was sorted by total_ev.

Examples

library(ggplot2)

data(pres_pred_2020, package = "plinko")

ggplot(pres_pred_2020, aes(x = total_ev, y = evprob_chal)) +
  geom_col() +
  geom_vline(xintercept = 269) +
  labs(
    x = "Electoral votes for Biden",
    y = "Predicted probability",
    title = "Predictive distribution for Biden's electoral votes",
    subtitle = "FiveThirtyEight's model of the 2020 US election as of Oct 16, 2020"
  )