Title: | Data and Code Behind the Stories and Interactives at 'FiveThirtyEight' |
---|---|
Description: | Datasets and code published by the data journalism website 'FiveThirtyEight' available at <https://github.com/fivethirtyeight/data>. Note that while we received guidance from editors at 'FiveThirtyEight', this package is not officially published by 'FiveThirtyEight'. |
Authors: | Albert Y. Kim [aut, cre] , Chester Ismay [aut] , Jennifer Chunn [aut], Meredith Manley [ctb] , Maggie Shea [ctb], Starry Yujia Zhou [ctb], Andrew Flowers [ctb], Jonathan Bouchet [ctb], G. Elliott Morris [ctb], Adam Spannbauer [ctb], Pradeep Adhokshaja [ctb], Olivia Barrows [ctb], Jojo Miller [ctb], Jayla Nakayama [ctb], Ben Baumer [ctb] , Rana Gahwagy [ctb] , Natalia Iannucci [ctb] , Marium Tapal [ctb] , Irene Ryan [ctb], Alina Barylsky [ctb], Danica Miguel [ctb], Sunni Raleigh [ctb], Anna Ballou [ctb], Jane Bang [ctb], Jordan Moody [ctb], Kara Van Allen [ctb], Jessica Keast [ctb], Lizette Carpenter [ctb], Fatima Keita [ctb] |
Maintainer: | Albert Y. Kim <[email protected]> |
License: | MIT + file LICENSE |
Version: | 0.6.2.9000 |
Built: | 2024-11-04 19:54:26 UTC |
Source: | https://github.com/rudeboybert/fivethirtyeight |
The raw data behind the story "Why The GOP Is So Hell-Bent On Passing An Unpopular Health Care Bill" https://fivethirtyeight.com/features/why-the-gop-is-so-hell-bent-on-passing-an-unpopular-health-care-bill/.
ahca_polls
ahca_polls
A data frame with 15 rows representing polls and 7 variables:
Start date of the poll.
End date of the poll.
The entity that conducts and collects information from the poll.
The number of affirmative responses to the question at the pollster.
The number of negative responses to the question at the pollster.
The website associated with the polling question.
The polling question asked at the pollster.
See https://github.com/fivethirtyeight/data/blob/master/ahca-polls/README.md
# To convert data frame to tidy data (long) format, run: library(dplyr) library(tidyr) library(stringr) ahca_polls_tidy <- ahca_polls %>% pivot_longer(-c(start, end, pollster, text, url), names_to = "opinion", values_to = "count")
# To convert data frame to tidy data (long) format, run: library(dplyr) library(tidyr) library(stringr) ahca_polls_tidy <- ahca_polls %>% pivot_longer(-c(start, end, pollster, text, url), names_to = "opinion", values_to = "count")
The raw data behind the story "Should Travelers Avoid Flying Airlines That Have Had Crashes in the Past?" https://fivethirtyeight.com/features/should-travelers-avoid-flying-airlines-that-have-had-crashes-in-the-past/.
airline_safety
airline_safety
A data frame with 56 rows representing airlines and 9 variables:
airline
indicates that regional subsidiaries are included
available seat kilometers flown every week
Total number of incidents, 1985-1999
Total number of fatal accidents, 1985-1999
Total number of fatalities, 1985-1999
Total number of incidents, 2000-2014
Total number of fatal accidents, 2000-2014
Total number of fatalities, 2000-2014
Aviation Safety Network https://aviation-safety.net.
# To convert data frame to tidy data (long) format, run: library(dplyr) library(tidyr) library(stringr) airline_safety_tidy <- airline_safety %>% pivot_longer(-c(airline, incl_reg_subsidiaries, avail_seat_km_per_week), names_to = "type", values_to = "count") %>% mutate( period = str_sub(type, start=-5), period = str_replace_all(period, "_", "-"), type = str_sub(type, end=-7) )
# To convert data frame to tidy data (long) format, run: library(dplyr) library(tidyr) library(stringr) airline_safety_tidy <- airline_safety %>% pivot_longer(-c(airline, incl_reg_subsidiaries, avail_seat_km_per_week), names_to = "type", values_to = "count") %>% mutate( period = str_sub(type, start=-5), period = str_replace_all(period, "_", "-"), type = str_sub(type, end=-7) )
The raw data behind the story "Trump Might Be The First President To Scrap A National Monument" https://fivethirtyeight.com/features/trump-might-be-the-first-president-to-scrap-a-national-monument/.
antiquities_act
antiquities_act
A data frame with 344 rows representing acts and 9 variables (Note that 7 of the original rows failed to parse and are omitted here):
Current name of piece of land designated under the Antiquities Act
State(s) or territory where land is located
If included, original name of piece of land designated under the Antiquities Act
Current land management agency. NPS = National Parks Service, BLM = Bureau of Land Management, USFS = US Forest Service, FWS = US Fish and Wildlife Service, NOAA = National Oceanic and National Oceanic and Atmospheric Administration
Type of action taken on land
Date of action
Year of action
President or congress that issued action
Acres affected by action. Note that total current acreage is not included. National monuments that cover ocean are listed in square miles.
National Parks Conservation Association https://www.npca.org/ and National Parks Service Archeology Program https://www.nps.gov/history/archeology/sites/antiquities/MonumentsList.htm
The raw data behind the story "How Much Trouble Is Ted Cruz Really In?" https://fivethirtyeight.com/features/how-much-trouble-is-ted-cruz-really-in/.
august_senate_polls
august_senate_polls
A data frame with 594 rows representing senate polls, and 11 variables:
the election year
the state of the poll
the class of the senate
the start date of the poll
the end odate of the poll
the percent of support for the Democrat during the poll
the percent of support for the Republican during the poll
the result percent of support for the Democrat during the election
the result percent of support for the Republican during the election
the difference between the percent of support of one party during the poll and the result percent of support for the same party during the election
the absolute value of the error value
Emerson College’s poll of registered voters
The raw data behind the story "Joining The Avengers Is As Deadly As Jumping Off A Four-Story Building" https://fivethirtyeight.com/features/avengers-death-comics-age-of-ultron/.
avengers
avengers
A data frame with 173 rows representing characters and 21 variables:
The URL of the comic character on the Marvel Wikia
The full name or alias of the character
The number of comic books that character appeared in as of April 30
Is the member currently active on an avengers affiliated team?
The recorded gender of the character
Sometimes the character was given probationary status as an Avenger, this is the date that happened
The month and year the character was introduced as a full or reserve member of the Avengers
The year the character was introduced as a full or reserve member of the Avengers
2015 minus the year
The status of the avenger, if they were given "Honorary" Avenger status, if they are simply in the "Academy," or "Full" otherwise
TRUE if the Avenger died, FALSE if not.
TRUE if the Avenger returned from their first death, FALSE if they did not, blank if not applicable
TRUE if the Avenger died a second time after their revival, FALSE if they did not, blank if not applicable
TRUE if the Avenger returned from their second death, FALSE if they did not, blank if not applicable
TRUE if the Avenger died a third time after their second revival, FALSE if they did not, blank if not applicable
TRUE if the Avenger returned from their third death, FALSE if they did not, blank if not applicable
TRUE if the Avenger died a fourth time after their third revival, FALSE if they did not, blank if not applicable
TRUE if the Avenger returned from their fourth death, FALSE if they did not, blank if not applicable
TRUE if the Avenger died a fifth time after their fourth revival, FALSE if they did not, blank if not applicable
TRUE if the Avenger returned from their fifth death, FALSE if they did not, blank if not applicable
Descriptions of deaths and resurrections.
Deaths of Marvel comic book characters between the time they joined the Avengers and April 30, 2015, the week before Secret Wars #1.
The raw data behind the stories: "How To Spot A Front-Runner On The 'Bachelor' Or 'Bachelorette'" https://fivethirtyeight.com/features/the-bachelorette/, "Rachel's Season Is Fitting Neatly Into 'Bachelorette' History" https://fivethirtyeight.com/features/rachels-season-is-fitting-neatly-into-bachelorette-history/, and "Rachel Lindsay's 'Bachelorette' Season, In Three Charts" https://fivethirtyeight.com/features/rachel-lindsays-bachelorette-season-in-three-charts/.
bachelorette
bachelorette
A data frame with 887 rows representing the Bachelorette and Bachelor contestants and 23 variables:
Bachelor or Bachelorette.
Which season?
An identifier for the contestant in a given season.
Who was eliminated in week 1.
Who was eliminated in week 2.
Who was eliminated in week 3.
Who was eliminated in week 4.
Who was eliminated in week 5.
Who was eliminated in week 6.
Who was eliminated in week 7.
Who was eliminated in week 8.
Who was eliminated in week 9.
Who was eliminated in week 10.
Who was on which date in week 1.
Who was on which date in week 2.
Who was on which date in week 3.
Who was on which date in week 4.
Who was on which date in week 5.
Who was on which date in week 6.
Who was on which date in week 7.
Who was on which date in week 8.
Who was on which date in week 9.
Who was on which date in week 10.
Eliminates connote either an elimination (starts with "E") or a rose (starts with "R"). Eliminations supersede roses. "E" connotes a standard elimination, typically at a rose ceremony. "EQ" means the contestant quits. "EF" means the contestant was fired by production. "ED" connotes a date elimination. "EU" connotes an unscheduled elimination, one that takes place at a time outside of a date or rose ceremony. "R" means the contestant received a rose. "R1" means the contestant got a first impression rose. "D1" means a one-on-one date, "D2" means a 2-on-1, "D3" means a 3-on-1 group date, and so on. Weeks of the show are eliminated by rose ceremonies, and may not line up exactly with episodes.
https://bachelor-nation.fandom.com/wiki/Bachelor_Nation_Wiki and then missing seasons were filled in by ABC and FiveThirtyEight staffers.
The raw data behind the story "Dear Mona, Which State Has The Worst Drivers?" https://fivethirtyeight.com/features/which-state-has-the-worst-drivers/
bad_drivers
bad_drivers
A data frame with 51 rows representing the 50 states + D.C. and 8 variables:
State
Number of drivers involved in fatal collisions per billion miles
Percentage of drivers involved in fatal collisions who were speeding
Percentage of drivers involved in fatal collisions who were alcohol-impaired
Percentage of drivers involved in fatal collisions who were not distracted
Percentage of drivers involved in fatal collisions who had not been involved in any previous accidents
Car insurance premiums ($)
Losses incurred by insurance companies for collisions per insured driver ($)
National Highway Traffic Safety Administration 2012, National Highway Traffic Safety Administration 2009 & 2012, National Association of Insurance Commissioners 2010 & 2011.
The raw data behind the story "The Dollar-And-Cents Case Against Hollywood's Exclusion of Women" https://fivethirtyeight.com/features/the-dollar-and-cents-case-against-hollywoods-exclusion-of-women/.
bechdel
bechdel
A data frame with 1794 rows representing movies and 15 variables:
Year of release
Text to construct IMDB url. Ex: https://www.imdb.com/title/tt1711425
Movie test
bechdel test result (detailed, with discrepancies indicated)
bechdel test result (detailed): ok
= passes test, dubious
, men
= women only talk about men, notalk
= women don't talk to each other, nowomen
= fewer than two women
Bechdel Test PASS vs FAIL binary
Film budget
Domestic (US) gross
Total International (i.e., worldwide) gross
Bechdel Code
Budget in 2013 inflation adjusted dollars
Domestic gross (US) in 2013 inflation adjusted dollars
Total International (i.e., worldwide) gross in 2013 inflation adjusted dollars
A vignette of an analysis of this dataset using the tidyverse
can be found on CRAN or by running:
vignette("bechdel", package = "fivethirtyeightdata")
https://bechdeltest.com/ and https://www.the-numbers.com/. The original data can be found at https://github.com/fivethirtyeight/data/tree/master/bechdel.
The raw data behind the story "'Straight Outta Compton' Is The Rare Biopic Not About White Dudes" https://fivethirtyeight.com/features/straight-outta-compton-is-the-rare-biopic-not-about-white-dudes/. An analysis using this data was contributed by Pradeep Adhokshaja as a package vignette at https://fivethirtyeightdata.github.io/fivethirtyeightdata/articles/biopics.html.
biopics
biopics
A data frame with 761 rows representing movies and 14 variables:
Title of the film.
Text to construct IMDB url. Ex: https://www.imdb.com/title/tt1711425
Country of origin.
Year of release.
Gross earnings at U.S. box office.
Director of film.
The number of subjects featured in the film.
The actual name of the featured subject.
The occupation of subject or reason for recognition.
Indicates whether the subject's race was discernible based on background of self, parent, or grandparent.
Race of the subject.
Dummy variable that indicates person of color.
Sex of subject.
The actor or actress who played the subject.
The raw data behind the story "A Statistical Analysis of the Work of Bob Ross" https://fivethirtyeight.com/features/a-statistical-analysis-of-the-work-of-bob-ross/. An analysis using this data was contributed by Jonathan Bouchet as a package vignette at https://fivethirtyeightdata.github.io/fivethirtyeightdata/articles/bob_ross.html.
bob_ross
bob_ross
A data frame with 403 rows representing episodes and 71 variables:
Episode code
Season number
Episode number
Title of episode
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
Present (1) or not (0)
See https://github.com/fivethirtyeight/data/tree/master/bob-ross
# To convert data frame to tidy data (long) format, run: library(dplyr) library(tidyr) library(stringr) bob_ross_tidy <- bob_ross %>% pivot_longer(-c(episode, season, episode_num, title), names_to = "object", values_to = "present") %>% mutate(present = as.logical(present)) %>% arrange(episode, object)
# To convert data frame to tidy data (long) format, run: library(dplyr) library(tidyr) library(stringr) bob_ross_tidy <- bob_ross %>% pivot_longer(-c(episode, season, episode_num, title), names_to = "object", values_to = "present") %>% mutate(present = as.logical(present)) %>% arrange(episode, object)
The raw data behind the story "Two Years In, Turnover In Trump’s Cabinet Is Still Historically High" https://fivethirtyeight.com/features/two-years-in-turnover-in-trumps-cabinet-is-still-historically-high/.
cabinet_turnover
cabinet_turnover
A data frame with 312 rows representing cabinet members and 8 variables:
Surname of of sitting President
Cabinet Position
Appointee's full name
Date the appointee was sworn in
Date the appointee left office
Length of Tenure, in days
Days into administration that the appointee left office
Whether or not Cabinet member served in more than one administrations
from Federal Government Websites and News Reports
The raw data behind the story "Looking For Clues: Who Is Going To Run For President In 2016?" https://fivethirtyeight.com/features/2016-president-who-is-going-to-run/.
cand_events_20150114
cand_events_20150114
A data frame with 42 rows representing events attended in Iowa and New Hampshire by potential presidential primary candidates and 8 variables:
Potential presidential candidate
Political party
State of event
Name of event
Type of event
Date of event
Link to event
Snippet of event description
See https://github.com/fivethirtyeight/data/tree/master/potential-candidates
cand_state_20150114
, cand_events_20150130
, and cand_state_20150130
The raw data behind the story "Who Will Run For President: Romney Is Out" https://fivethirtyeight.com/features/romney-not-running-for-president/.
cand_events_20150130
cand_events_20150130
A data frame with 74 rows representing events attended by potential presidential primary candidates and 8 variables:
Potential presidential candidate
Political party
State of event
Name of event
Type of event
Date of event
Link to event
Snippet of event description
See https://github.com/fivethirtyeight/data/tree/master/potential-candidates
cand_state_20150130
, cand_events_20150114
, and cand_state_20150114
The raw data behind the story "Looking For Clues: Who Is Going To Run For President In 2016?" https://fivethirtyeight.com/features/2016-president-who-is-going-to-run/.
cand_state_20150114
cand_state_20150114
A data frame with 25 rows representing potential presidential primary candidates and 5 variables:
Potential presidential candidate
Political party
Date of event
Latest statement
Likelihood of running score, 1 = Not running, 5 = Definitely running
See https://github.com/fivethirtyeight/data/tree/master/potential-candidates
cand_events_20150114
, cand_events_20150130
, and cand_state_20150130
The raw data behind the story "Who Will Run For President: Romney Is Out" https://fivethirtyeight.com/features/romney-not-running-for-president/.
cand_state_20150130
cand_state_20150130
A data frame with 27 rows representing potential presidential primary candidates and 5 variables:
Potential presidential candidate
Political party
Date of event
Latest statement
Likelihood of running score, 1 = Not running, 5 = Definitely running
See https://github.com/fivethirtyeight/data/tree/master/potential-candidates
cand_events_20150130
, cand_events_20150114
, and cand_state_20150114
The raw data behind the story "The Ultimate Halloween Candy Power Ranking" https://fivethirtyeight.com/features/the-ultimate-halloween-candy-power-ranking/.
candy_rankings
candy_rankings
A data frame with 85 rows representing Halloween candy and 13 variables:
The name of the Halloween candy.
Does it contain chocolate?
Is it fruit flavored?
Is there caramel in the candy?
Does it contain peanuts, peanut butter or almonds?
Does it contain nougat?
Does it contain crisped rice, wafers, or a cookie component?
Is it a hard candy?
Is it a candy bar?
Is it one of many candies in a bag or box?
The percentile of sugar it falls under within the data set.
The unit price percentile compared to the rest of the set.
The overall win percentage according to 269,000 matchups.
See https://github.com/fivethirtyeight/data/tree/master/candy-power-ranking
# To convert data frame to tidy data (long) format, run: library(dplyr) library(tidyr) library(stringr) candy_rankings_tidy <- candy_rankings %>% pivot_longer(-c(competitorname, sugarpercent, pricepercent, winpercent), names_to = "characteristics", values_to = "present") %>% mutate(present = as.logical(present)) %>% arrange(competitorname)
# To convert data frame to tidy data (long) format, run: library(dplyr) library(tidyr) library(stringr) candy_rankings_tidy <- candy_rankings %>% pivot_longer(-c(competitorname, sugarpercent, pricepercent, winpercent), names_to = "characteristics", values_to = "present") %>% mutate(present = as.logical(present)) %>% arrange(competitorname)
The raw data behind the story "American Chess Is Great Again" https://fivethirtyeight.com/features/american-chess-is-great-again/.
chess_transfers
chess_transfers
A data frame with 932 rows representing international player transfers and 5 variables:
The corresponding website on the World Chess Federation page which details the transfers of a given year.
An numeric identifier for the chess player who transferred.
The current national federation of the chess player
The national federation from which the chess player has transferred.
The date at which the transfer took place.
World Chess Federation
The raw data behind the story "Why Classic Rock Isn't What It Used To Be" https://fivethirtyeight.com/features/why-classic-rock-isnt-what-it-used-to-be/.
classic_rock_raw_data
classic_rock_raw_data
A data frame with 37,673 rows representing song plays and 8 variables:
Song name
Artist name
Station callsign
Time of song play in seconds elapsed since January 1, 1970
Time of song play in date/time format
Unique ID for each song play
Song and artist name combined
See https://github.com/fivethirtyeight/data/tree/master/classic-rock
The raw data behind the story "Why Classic Rock Isn't What It Used To Be" https://fivethirtyeight.com/features/why-classic-rock-isnt-what-it-used-to-be/.
classic_rock_song_list
classic_rock_song_list
A data frame with 2230 rows representing unique songs and 7 variables:
Song name
Artist name
Release year as listed in SongFacts
Song and artist name combined
Logical variable of whether release year is included
Number of plays across all stations
Number of plays across all stations if a year was found
SongFacts and https://github.com/fivethirtyeight/data/tree/master/classic-rock
The raw data behind the story "The Economic Guide To Picking A College Major" https://fivethirtyeight.com/features/the-economic-guide-to-picking-a-college-major/.
college_all_ages
college_all_ages
A data frame with 173 rows representing majors (all ages) and 11 variables:
Major code, FO1DP in ACS PUMS
Major description
Category of major from Carnevale et al
Total number of people with major
Number employed (ESR == 1 or 2)
Employed at least 50 weeks (WKW == 1) and at least 35 hours (WKHP >= 35)
Number unemployed (ESR == 3)
Unemployed / (Unemployed + Employed)
25th percentile of earnings
Median earnings of full-time, year-round workers
75th percentile of earnings
See https://github.com/fivethirtyeight/data/blob/master/college-majors/readme.md.
college_grad_students
, college_recent_grads
The raw data behind the story "The Economic Guide To Picking A College Major" https://fivethirtyeight.com/features/the-economic-guide-to-picking-a-college-major/.
college_grad_students
college_grad_students
A data frame with 173 rows representing majors (graduate vs nongraduate students) and 22 variables:
Major code, FO1DP in ACS PUMS
Major description
Category of major from Carnevale et al
Total number of people with major
Sample size (unweighted) of full-time, year-round ONLY (used for earnings)
Number employed (ESR == 1 or 2)
Employed at least 50 weeks (WKW == 1) and at least 35 hours (WKHP >= 35)
Number unemployed (ESR == 3)
Unemployed / (Unemployed + Employed)
25th percentile of earnings
Median earnings of full-time, year-round workers
75th percentile of earnings
Total number of people with major
Number employed (ESR == 1 or 2)
Employed at least 50 weeks (WKW == 1) and at least 35 hours (WKHP >= 35)
Number unemployed (ESR == 3)
Unemployed / (Unemployed + Employed)
25th percentile of earnings
Median earnings of full-time, year-round workers
75th percentile of earnings
grad_total / (grad_total + nongrad_total)
(grad_median-nongrad_median)/nongrad_median
See https://github.com/fivethirtyeight/data/blob/master/college-majors/readme.md.
college_all_ages
, college_recent_grads
The raw data behind the story "The Economic Guide To Picking A College Major" https://fivethirtyeight.com/features/the-economic-guide-to-picking-a-college-major/.
college_recent_grads
college_recent_grads
A data frame with 173 rows representing majors (recent graduates) and 21 variables:
Rank by median earnings
Major code, FO1DP in ACS PUMS
Major description
Category of major from Carnevale et al
Total number of people with major
Sample size (unweighted) of full-time, year-round ONLY (used for earnings)
Men with major
Women with major
Proportion women
Number employed (ESR == 1 or 2)
Employed 35 hours or more
Employed less than 35 hours
Employed at least 50 weeks (WKW == 1) and at least 35 hours (WKHP >= 35)
Number unemployed (ESR == 3)
Unemployed / (Unemployed + Employed)
25th percentile of earnings
Median earnings of full-time, year-round workers
75th percentile of earnings
Number with job requiring a college degree
Number with job not requiring a college degree
Number in low-wage service jobs
See https://github.com/fivethirtyeight/data/blob/master/college-majors/readme.md. Note that
women-stem.csv
was a subset of the original recent-grads.csv
, so no data frame was created.
college_grad_students
, college_all_ages
The raw data behind the story "Elitist, Superfluous, Or Popular? We Polled Americans on the Oxford Comma" https://fivethirtyeight.com/features/elitist-superfluous-or-popular-we-polled-americans-on-the-oxford-comma/.
comma_survey
comma_survey
A data frame with 1129 rows representing respondents and 13 variables:
Respondent ID
Gender
Age
Household income bracket
Education level
Location (census region)
In your opinion, which sentence is more grammatically correct?
Prior to reading about it above, had you heard of the serial (or Oxford) comma?
How much, if at all, do you care about the use (or lack thereof) of the serial (or Oxford) comma in grammar?
How would you write the following sentence?
When faced with using the word "data", have you ever spent time considering if the word was a singular or plural noun?
How much, if at all, do you care about the debate over the use of the word "data" as a singular or plural noun?
In your opinion, how important or unimportant is proper use of grammar?
See https://github.com/fivethirtyeight/data/tree/master/comma-survey.
The raw data behind the story "Both Republicans And Democrats Have an Age Problem" https://fivethirtyeight.com/features/both-republicans-and-democrats-have-an-age-problem/.
congress_age
congress_age
A data frame with 18,635 rows representing members of Congress (House and Senate) and 13 variables:
Congress number.
Chamber of congress: House of Representatives or Senate.
bioguide
First name
Middle name
Last name
Suffix
Birthday
State abbreviation
Party abbreviation
Boolean variable of whether member was an incumbent.
Start date of session.
Age at start of session.
See https://github.com/fivethirtyeight/data/tree/master/congress-age
The raw data behind the story "How Many Americans Are Married To Their Cousins?" https://fivethirtyeight.com/features/how-many-americans-are-married-to-their-cousins/.
cousin_marriage
cousin_marriage
A data frame with 70 rows representing countries and 2 variables:
Country
Percent of marriages that are consanguineous
consang.net
The raw data behind the story "Every Guest Jon Stewart Ever Had On 'The Daily Show'" https://fivethirtyeight.com/features/every-guest-jon-stewart-ever-had-on-the-daily-show/.
daily_show_guests
daily_show_guests
A data frame with 2693 rows representing guests and 5 variables:
The year the episode aired
Their occupation or office, according to Google's Knowledge Graph or, if they're not in there, how Stewart introduced them on the program.
Air date of episode. Not unique, as some shows had more than one guest
A larger group designation for the occupation. For instance, us senators, us presidents, and former presidents are all under "politicians"
The person or list of people who appeared on the show, according to Wikipedia. The GoogleKnowledge_Occupation only refers to one of them in a given row.
Google Knowledge Graph, The Daily Show clip library, Wikipedia.
All datasets included in both fivethirtyeight and fivethirtyeightdata packages
datasets_master
datasets_master
A data frame with 9 variables:
Name of lazy-loaded data frame
Whether the (large) dataset is in the fivethirtyeightdata package
Title as it appears on FiveThirtyEight.com
Link to article on FiveThirtyEight.com
Main author
Second author (if any)
Third author (if any)
Date published
Tag for article
The raw data behind the stories: "We Researched Hundreds Of Races. Here’s Who Democrats Are Nominating" https://fivethirtyeight.com/features/democrats-primaries-candidates-demographics/ and "How’s The Progressive Wing Doing In Democratic Primaries So Far?" https://fivethirtyeight.com/features/the-establishment-is-beating-the-progressive-wing-in-democratic-primaries-so-far/.
dem_candidates
dem_candidates
A data frame with 811 rows representing Democratic candidates, and 32 variables:
All candidates who received votes in 2018’s Democratic primary elections for U.S. Senate, U.S. House and governor in which no incumbent ran. Supplied by Ballotpedia.
The state in which the candidate ran. Supplied by Ballotpedia.
The body of government for which the candidate ran. Supplied by Ballotpedia.
If applicable, congressional district number for which the candidate ran. Supplied by Ballotpedia.
The office for which the candidate ran. Supplied by Ballotpedia.
Whether it was a “regular” or “special” election. Supplied by Ballotpedia.
The date on which the primary was held. Supplied by Ballotpedia.
Whether the candidate lost (“Lost”) the primary or won/advanced to a runoff (“Advanced”). Supplied by Ballotpedia.
“None” if there was no runoff; “On the Ballot” if the candidate advanced to a runoff but it hasn’t been held yet; “Advanced” if the candidate won the runoff; “Lost” if the candidate lost the runoff. Supplied by Ballotpedia.
“On the Ballot” if the candidate won the primary or runoff and has advanced to November; otherwise, “None.” Supplied by Ballotpedia.
The FiveThirtyEight partisan lean of the district or state in which the election was held. Partisan leans are calculated by finding the average difference between how a state or district voted in the past two presidential elections and how the country voted overall, with 2016 results weighted 75 percent and 2012 results weighted 25 percent.
The percentage of the vote received by the candidate in his or her primary. In states that hold runoff elections, we looked only at the first round (the regular primary). In states that hold all-party primaries (e.g., California), a candidate’s primary percentage is the percentage of the total Democratic vote they received. Unopposed candidates and candidates nominated by convention (not primary) are given a primary percentage of 100 but were excluded from our analysis involving vote share. Numbers come from official results posted by the secretary of state or local elections authority; if those were unavailable, we used unofficial election results from the New York Times.
“Yes” if the candidate won his or her primary and has advanced to November; “No” if he or she lost.
“White” if we identified the candidate as non-Hispanic white; “Nonwhite” if we identified the candidate as Hispanic and/or any nonwhite race; blank if we could not identify the candidate’s race or ethnicity. To determine race and ethnicity, we checked each candidate’s website to see if he or she identified as a certain race. If not, we spent no more than two minutes searching online news reports for references to the candidate’s race.
If the candidate’s website says that he or she served in the armed forces, we put “Yes.” If the website is silent on the subject (or explicitly says he or she didn’t serve), we put “No.” If the field was left blank, no website was available.
If the candidate’s website says that he or she is LGBTQ (including indirect references like to a same-sex partner), we put “Yes.” If the website is silent on the subject (or explicitly says he or she is straight), we put “No.” If the field was left blank, no website was available.
We used Ballotpedia, VoteSmart and news reports to research whether the candidate had ever held elected office before, at any level. We put “Yes” if the candidate has held elected office before and “No” if not.
We used Federal Election Committee fundraising data (for federal candidates) and state campaign-finance data (for gubernatorial candidates) to look up how much each candidate had invested in his or her own campaign, through either donations or loans. We put “Yes” if the candidate donated or loaned a cumulative $400,000 or more to his or her own campaign before the primary and “No” for all other candidates.
If the candidate identifies on his or her website that he or she has a background in the fields of science, technology, engineering or mathematics, we put “Yes.” If not, we put “No.” If the field was left blank, no website was available.
We put “Yes” if the candidate mentions working for the Obama administration or campaign on his or her website, or if the candidate shows up on this list of Obama administration members and campaign hands running for office. If not, we put “No.”
“Yes” if the candidate was placed on the DCCC’s Red to Blue list before the primary, was endorsed by the DSCC before the primary, or if the DSCC/DCCC aired pre-primary ads in support of the candidate. (Note: according to the DGA’s press secretary, the DGA does not get involved in primaries.) “No” if the candidate is running against someone for whom one of the above things is true, or if one of those groups specifically anti-endorsed or spent money to attack the candidate. If those groups simply did not weigh in on the race, we left the cell blank.
“Yes” if the candidate was endorsed by Emily’s List before the primary. “No” if the candidate is running against an Emily-endorsed candidate or if Emily’s List specifically anti-endorsed or spent money to attack the candidate. If Emily’s List simply did not weigh in on the race, we left the cell blank.
“Yes” if the candidate received the Gun Sense Candidate Distinction from Moms Demand Action/Everytown for Gun Safety before the primary, according to media reports or the candidate’s website. “No” if the candidate is running against an candidate with the distinction. If Moms Demand Action simply did not weigh in on the race, we left the cell blank.
“Yes” if the candidate was endorsed by Joe Biden before the primary. “No” if the candidate is running against a Biden-endorsed candidate or if Biden specifically anti-endorsed the candidate. If Biden simply did not weigh in on the race, we left the cell blank.
“Yes” if the candidate was endorsed by Elizabeth Warren before the primary. “No” if the candidate is running against a Warren-endorsed candidate or if Warren specifically anti-endorsed the candidate. If Warren simply did not weigh in on the race, we left the cell blank.
“Yes” if the candidate was endorsed by Bernie Sanders before the primary. “No” if the candidate is running against a Sanders-endorsed candidate or if Sanders specifically anti-endorsed the candidate. If Sanders simply did not weigh in on the race, we left the cell blank.
“Yes” if the candidate was endorsed by Our Revolution before the primary, according to the Our Revolution website. “No” if the candidate is running against an Our Revolution-endorsed candidate or if Our Revolution specifically anti-endorsed or spent money to attack the candidate. If Our Revolution simply did not weigh in on the race, we left the cell blank.
“Yes” if the candidate was endorsed by Justice Democrats before the primary, according to the Justice Democrats website, candidate website or news reports. “No” if the candidate is running against a Justice Democrats-endorsed candidate or if Justice Democrats specifically anti-endorsed or spent money to attack the candidate. If Justice Democrats simply did not weigh in on the race, we left the cell blank.
“Yes” if the candidate was endorsed by the Progressive Change Campaign Committee before the primary, according to the PCCC website, candidate website or news reports. “No” if the candidate is running against a PCCC-endorsed candidate or if the PCCC specifically anti-endorsed or spent money to attack the candidate. If the PCCC simply did not weigh in on the race, we left the cell blank.
“Yes” if the candidate was endorsed by Indivisible before the primary, according to the Indivisible website, candidate website or news reports. “No” if the candidate is running against an Indivisible-endorsed candidate or if Indivisible specifically anti-endorsed or spent money to attack the candidate. If Indivisible simply did not weigh in on the race, we left the cell blank.
“Yes” if the candidate was endorsed by the Working Families Party before the primary, according to the WFP website, candidate website or news reports. “No” if the candidate is running against a WFP-endorsed candidate or if the WFP specifically anti-endorsed or spent money to attack the candidate. If the WFP simply did not weigh in on the race, we left the cell blank.
“Yes” if the candidate was endorsed by VoteVets before the primary, according to the VoteVets website, candidate website or news reports. “No” if the candidate is running against a VoteVets-endorsed candidate or if VoteVets specifically anti-endorsed or spent money to attack the candidate. If VoteVets simply did not weigh in on the race, we left the cell blank.
“Yes” if a No Labels-affiliated group (Citizens for a Strong America Inc., Forward Not Back, Govern or Go Home, United for Progress Inc. or United Together) spent money in support of the candidate in the primary. “No” if the candidate is running against an candidate supported by a No Labels-affiliated group or if a No Labels-affiliated group specifically anti-endorsed or spent money to attack the candidate. If No Labels simply did not weigh in on the race, we left the cell blank.
This data was also used in "We Looked At Hundreds Of Endorsements. Here’s Who Democrats Are Listening To" published on 2008-08-14 https://fivethirtyeight.com/features/the-establishment-is-beating-the-progressive-wing-in-democratic-primaries-so-far/
Ballotpedia, New York Times, and candidate websites. See also https://github.com/fivethirtyeight/data/blob/master/primary-candidates-2018/README.md
The raw data behind the story "Some Democrats Who Could Step Up If Hillary Isn't Ready For Hillary" https://fivethirtyeight.com/features/some-democrats-who-could-step-up-if-hillary-isnt-ready-for-hillary/.
democratic_bench
democratic_bench
A data frame with 67 rows representing members of the Democratic Party and 3 variables:
Candidate
Amount the candidate was expected to raise
Amount the candidate actually raised
See https://github.com/fivethirtyeight/data/tree/master/democratic-bench.
The raw data behind the story "Dear Mona Followup: Where Do People Drink The Most Beer, Wine And Spirits?" https://fivethirtyeight.com/features/dear-mona-followup-where-do-people-drink-the-most-beer-wine-and-spirits/.
drinks
drinks
A data frame with 193 rows representing countries and 5 variables:
country
Servings of beer in average serving sizes per person
Servings of spirits in average serving sizes per person
Servings of wine in average serving sizes per person
Total litres of pure alcohol per person
World Health Organization, Global Information System on Alcohol and Health (GISAH), 2010.
# To convert data frame to tidy data (long) format, run: library(dplyr) library(tidyr) library(stringr) drinks_tidy <- drinks %>% pivot_longer(cols = ends_with("servings"), names_to = "type", values_to = "servings") %>% mutate( type = str_sub(type, start=1, end=-10) ) %>% arrange(country, type)
# To convert data frame to tidy data (long) format, run: library(dplyr) library(tidyr) library(stringr) drinks_tidy <- drinks %>% pivot_longer(cols = ends_with("servings"), names_to = "type", values_to = "servings") %>% mutate( type = str_sub(type, start=1, end=-10) ) %>% arrange(country, type)
The raw data behind the story "How Baby Boomers Get High" https://fivethirtyeight.com/features/how-baby-boomers-get-high/. It covers usage of 13 drugs in the past 12 months across 17 age groups.
drug_use
drug_use
A data frame with 17 rows representing age groups and 28 variables:
Age group
Number of people surveyed
Percentage who used alcohol
Median number of times a user used alcohol
Percentage who used marijuana
Median number of times a user used marijuana
Percentage who used cocaine
Median number of times a user used cocaine
Percentage who used crack
Median number of times a user used crack
Percentage who used heroin
Median number of times a user used heroin
Percentage who used hallucinogens
Median number of times a user used hallucinogens
Percentage who used inhalants
Median number of times a user used inhalants
Percentage who used pain relievers
Median number of times a user used pain relievers
Percentage who used oxycontin
Median number of times a user used oxycontin
Percentage who used tranquilizer
Median number of times a user used tranquilizer
Percentage who used stimulants
Median number of times a user used stimulants
Percentage who used meth
Median number of times a user used meth
Percentage who used sedatives
Median number of times a user used sedatives
National Survey on Drug Use and Health from the Substance Abuse and Mental Health Data Archive https://www.icpsr.umich.edu/icpsrweb/content/SAMHDA/index.html.
# To convert data frame to tidy data (long) format, run: library(dplyr) library(tidyr) library(stringr) use <- drug_use %>% select(age, n, ends_with("_use")) %>% pivot_longer(-c(age, n), names_to = "drug", values_to = "use") %>% mutate(drug = str_sub(drug, start=1, end=-5)) freq <- drug_use %>% select(age, n, ends_with("_freq")) %>% pivot_longer(-c(age, n), names_to = "drug", values_to = "freq") %>% mutate(drug = str_sub(drug, start=1, end=-6)) drug_use_tidy <- left_join(x=use, y=freq, by = c("age", "n", "drug")) %>% arrange(age)
# To convert data frame to tidy data (long) format, run: library(dplyr) library(tidyr) library(stringr) use <- drug_use %>% select(age, n, ends_with("_use")) %>% pivot_longer(-c(age, n), names_to = "drug", values_to = "use") %>% mutate(drug = str_sub(drug, start=1, end=-5)) freq <- drug_use %>% select(age, n, ends_with("_freq")) %>% pivot_longer(-c(age, n), names_to = "drug", values_to = "freq") %>% mutate(drug = str_sub(drug, start=1, end=-6)) drug_use_tidy <- left_join(x=use, y=freq, by = c("age", "n", "drug")) %>% arrange(age)
This folder contains the data behind the story 'Election Update: The House Districts That Swing The Most (And Least) With The National Mood' https://fivethirtyeight.com/features/election-update-the-house-districts-that-swing-the-most-and-least-with-the-national-mood/
elasticity_by_district
elasticity_by_district
A dataset with 435 rows representing congressional districts and 2 variables
congressional district
pvi
The original dataset only has 2 columns: "district" and "elasticity". I separated the "district" columns into two. For example, in row 1 of the dataset, the original "district" = "MI-5", and I separated it into “state" = "Michigan" and "district_number" = "5". In addition, I used the full names for all states instead of abbreviations.
An elasticity score measures how sensitive a state or district it is to changes in the national political environment.
This folder contains the data behind the story 'Election Update: The House Districts That Swing The Most (And Least) With The National Mood' https://fivethirtyeight.com/features/election-update-the-house-districts-that-swing-the-most-and-least-with-the-national-mood/
elasticity_by_state
elasticity_by_state
A dataset with 435 rows representing each state and the District of Columbia and 2 variables
state
pvi
I used the full names for all states instead of abbreviations.
An elasticity score measures how sensitive a state or district it is to changes in the national political environment.
The raw data behind the story "Blatter's Reign At FIFA Hasn't Helped Soccer's Poor" https://fivethirtyeight.com/features/blatters-reign-at-fifa-hasnt-helped-soccers-poor/.
elo_blatter
elo_blatter
A data frame with 191 rows representing countries and 5 variables:
FIFA member country
The team's Elo in 1998
The team's Elo in 2015
Confederation to which country belongs
The country's purchasing power parity GDP as of 2006
The country's 2006 population
Source for gdp06
Source for popu06
See https://github.com/fivethirtyeight/data/tree/master/elo-blatter.
The raw data behind the story "Pols And Polls Say The Same Thing: Jeb Bush Is A Weak Front-Runner" https://fivethirtyeight.com/features/pols-and-polls-say-the-same-thing-jeb-bush-is-a-weak-front-runner/. This data includes something we call "endorsement points," an attempt to quantify the importance of endorsements by weighting each one according to the position held by the endorser: 10 points for each governor, 5 points for each senator and 1 point for each representative
endorsements
endorsements
A data frame with 109 rows representing candidates and 9 variables:
Election year
Political party
Candidate running in primary
Weighted endorsements through June 30th of the year before the primary
Percentage of total weighted endorsement points for the candidate's political party through June 30th of the year before the primary
Money raised through June 30th of the year before the primary
Percentage of total money raised by the candidate's political party through June 30th of the year before the primary
Percentage of votes won in the primary
Did the candidate win the primary?
See https://github.com/fivethirtyeight/data/tree/master/endorsements-june-30
The raw data behind the story "The 2020 Endorsement Primary - Which Democratic candidates are receiving the most support from prominent members of their party?" https://projects.fivethirtyeight.com/2020-endorsements/democratic-primary/.
endorsements_2020
endorsements_2020
A data frame with 1000 rows representing endorsements and 13 variables:
date of the endorsement
position of the endorser
city of the endorser
state of the endorser
name of the endorser
name of the endorsee
party of the endorser
source link of the endorsement
order of the endorsement
category of the endorsement
body of the endorsement
district
points the endorsement counts for
2020 endorsement tracker. Methodology: https://fivethirtyeight.com/methodology/how-our-presidential-endorsement-tracker-works/.
The raw data behind the story "Be Suspicious Of Online Movie Ratings, Especially Fandango's" https://fivethirtyeight.com/features/fandango-movies-ratings/. contains every film that has a Rotten Tomatoes rating, a RT User rating, a Metacritic score, a Metacritic User score, and IMDb score, and at least 30 fan reviews on Fandango.
fandango
fandango
A data frame with 146 rows representing movies and 23 variables:
The film in question
Year of film
The Rotten Tomatoes Tomatometer score for the film
The Rotten Tomatoes user score for the film
The Metacritic critic score for the film
The Metacritic user score for the film
The IMDb user score for the film
The number of stars the film had on its Fandango movie page
The Fandango ratingValue for the film, as pulled from the HTML of each page. This is the actual average score the movie obtained.
The Rotten Tomatoes Tomatometer score for the film , normalized to a 0 to 5 point system
The Rotten Tomatoes user score for the film , normalized to a 0 to 5 point system
The Metacritic critic score for the film, normalized to a 0 to 5 point system
The Metacritic user score for the film, normalized to a 0 to 5 point system
The IMDb user score for the film, normalized to a 0 to 5 point system
The Rotten Tomatoes Tomatometer score for the film , normalized to a 0 to 5 point system and rounded to the nearest half-star
The Rotten Tomatoes user score for the film , normalized to a 0 to 5 point system and rounded to the nearest half-star
The Metacritic critic score for the film, normalized to a 0 to 5 point system and rounded to the nearest half-star
The Metacritic user score for the film, normalized to a 0 to 5 point system and rounded to the nearest half-star
The IMDb user score for the film, normalized to a 0 to 5 point system and rounded to the nearest half-star
The number of user votes the film had on Metacritic
The number of user votes the film had on IMDb
The number of user votes the film had on Fandango
The difference between the presented Fandango_Stars and the actual Fandango_Ratingvalue
The data from Fandango was pulled on Aug. 24, 2015.
The raw data behind the story "How To Break FIFA" https://fivethirtyeight.com/features/how-to-break-fifa/.
fifa_audience
fifa_audience
A data frame with 3652 rows representing guests and 6 variables:
FIFA member country
Confederation to which country belongs
Country's share of global population (percentage)
Country's share of global world cup TV Audience (percentage)
Country's GDP-weighted audience share (percentage)
See https://github.com/fivethirtyeight/data/tree/master/fifa
# To convert data frame to tidy data (long) format, run: library(dplyr) library(tidyr) library(stringr) fifa_audience_tidy <- fifa_audience %>% pivot_longer(-c(country, confederation), names_to = "type", values_to = "share") %>% mutate(type = str_sub(type, start=1, end=-7)) %>% arrange(country)
# To convert data frame to tidy data (long) format, run: library(dplyr) library(tidyr) library(stringr) fifa_audience_tidy <- fifa_audience %>% pivot_longer(-c(country, confederation), names_to = "type", values_to = "share") %>% mutate(type = str_sub(type, start=1, end=-7)) %>% arrange(country)
The data behind the story "Our Guide To The Exuberant Nonsense Of College Fight Songs" https://projects.fivethirtyeight.com/college-fight-song-lyrics/.
fight_songs
fight_songs
A data frame with 65 rows representing college fight songs, and 23 variables:
school name
school college football conference
song title
song author(s)
year the song was written; some years are unknown
TRUE if song was written by a student, FALSE if not
TRUE if song is an official fight song according to the university, FALSE if not
TRUE if song was chosen as part of a contest, FALSE if not
beats per minute
duration of the song in seconds
TRUE if song says 'fight', FALSE if not
number of time song says 'fight'
TRUE if song says 'victory', FALSE if not
TRUE if song says 'win' or 'won', FALSE if not
TRUE if song says 'victory', 'win', or 'won'
TRUE if song says 'rah', FALSE if not
TRUE if song uses nonsense syllables, FALSE if not
TRUE if song mentions school colors, FALSE if not
TRUE if song refers to a group of men, boys, sons, etc., FALSE if not
TRUE if song mentions opponents, FALSE if not
TRUE if song spells something out, FALSE if not
total number of tropes in song
Spotify id for song
Spotify https://www.spotify.com/us/
An R package that provides access to the code and data sets published by FiveThirtyEight https://github.com/fivethirtyeight/data. Note that while we received guidance from editors at 538, this package is not officially published by 538. You can explore all datasets here: https://fivethirtyeight-r.netlify.app/articles/fivethirtyeight.html
# Example usage: library(fivethirtyeight) head(bechdel) # All information about any data set can be found in the help file: ?bechdel # To view a list of all data sets: data(package = "fivethirtyeight")
# Example usage: library(fivethirtyeight) head(bechdel) # All information about any data set can be found in the help file: ?bechdel # To view a list of all data sets: data(package = "fivethirtyeight")
The raw data behind the story "41 Percent Of Fliers Think You're Rude If You Recline Your Seat" https://fivethirtyeight.com/features/airplane-etiquette-recline-seat/.
flying
flying
A data frame with 1040 rows representing respondents and 27 variables:
RespondentID
Gender
Age
Height
Do you have any children under 18?
Household income bracket
Education Level
Location (census region)
How often do you travel by plane?
Do you ever recline your seat when you fly?
Under normal circumstances, does a person who reclines their seat during a flight have any obligation to the person sitting behind them?
Is it rude to recline your seat on a plane?
Given the opportunity, would you eliminate the possibility of reclining seats on planes entirely?
Is it rude to ask someone to switch seats with you in order to be closer to friends?
Is it rude to ask someone to switch seats with you in order to be closer to family?
Is it rude to wake a passenger up if you are trying to go to the bathroom?
Is it rude to wake a passenger up if you are trying to walk around?
In general, is it rude to bring a baby on a plane?
In general, is it rude to knowingly bring unruly children on a plane?
In a row of three seats, who should get to use the two arm rests?
In a row of two seats, who should get to use the middle arm rest?
Who should have control over the window shade?
Is it rude to move to an unsold seat on a plane?
Generally speaking, is it rude to say more than a few words to the stranger sitting next to you on a plane?
On a 6 hour flight from NYC to LA, how many times is it acceptable to get up if you're not in an aisle seat?
Have you ever used personal electronics during take off or landing in violation of a flight attendant's direction?
Have you ever smoked a cigarette in an airplane bathroom when it was against the rules?
SurveyMonkey survey
The raw data behind the story "The FiveThirtyEight International Food Association's 2014 World Cup" https://fivethirtyeight.com/features/the-fivethirtyeight-international-food-associations-2014-world-cup/. For all the countries below, the response to the following question is presented: "Please rate how much you like the traditional cuisine of X"
5: I love this country's traditional cuisine. I think it's one of the best in the world.
4: I like this country's traditional cuisine. I think it's considerably above average.
3: I'm OK with this county's traditional cuisine. I think it's about average.
2: I dislike this country's traditional cuisine. I think it's considerably below average.
1: I hate this country's traditional cuisine. I think it's one of the worst in the world.
N/A: I'm unfamiliar with this country's traditional cuisine.
food_world_cup
food_world_cup
A data frame with 1373 rows representing respondents and 48 variables:
Respondent ID
Generally speaking, how would you rate your level of knowledge of cuisines from different parts of the world?
How much, if at all, are you interested in cuisines from different parts of the world?
Gender
Age
Household income bracket
Education Level
Location (census region)
Cuisine of Algeria
Cuisine of Argentina
Cuisine of Australia
Cuisine of Belgium
Cuisine of Bosnia & Herzegovina
Cuisine of Brazil
Cuisine of Cameroon
Cuisine of Chile
Cuisine of China
Cuisine of Colombia
Cuisine of Costa Rica
Cuisine of Croatia
Cuisine of Cuba
Cuisine of Ecuador
Cuisine of England
Cuisine of Ethiopia
Cuisine of France
Cuisine of Germany
Cuisine of Ghana
Cuisine of Greece
Cuisine of Honduras
Cuisine of India
Cuisine of Iran
Cuisine of Ireland
Cuisine of Italy
Cuisine of Ivory Coast
Cuisine of Japan
Cuisine of Mexico
Cuisine of Nigeria
Cuisine of Portugal
Cuisine of Russia
Cuisine of South Korea
Cuisine of Spain
Cuisine of Switzerland
Cuisine of Thailand
Cuisine of the Netherlands
Cuisine of Turkey
Cuisine of the United States
Cuisine of Uruguay
Cuisine of Vietnam
See https://github.com/fivethirtyeight/data/tree/master/food-world-cup
The raw data behind the story 'How FiveThirtyEight's 2018 Midterm Forecasts Did' https://fivethirtyeight.com/features/how-fivethirtyeights-2018-midterm-forecasts-did/
forecast_results_2018
forecast_results_2018
A dataframe with 1518 rows representing forecast results (as of December 3, 2018) and 11 variables:
cycle of the election
branch of the election
election forecast for the gubernatorial race
the date of the forecast
version of the election forecast
the probability of winning for the Democrat
the probability of winning for the Republican
the predicted political affiliation of the forecast
whether the Democrat won
whether the Republican won
if a race was uncalled
FiveThirtyEight's 2018 House Forecast https://projects.fivethirtyeight.com/2018-midterm-election-forecast/house/
The raw data behind the story "We Watched 906 Foul Balls To Find Out Where The Most Dangerous Ones" https://fivethirtyeight.com/features/we-watched-906-foul-balls-to-find-out-where-the-most-dangerous-ones-land/.
foul_balls
foul_balls
A data frame with 906 rows representing foul balls and 7 variables:
the two teams that played
date of the most foul heavy day at each stadium
fly, grounder, line drive, popup, batter hits self
recorded velocity of each hit
zone predicted the foul ball would land in gauging angles
actual zone the ball landed it confirmed by camera angles
zone used for analysis
Information on the Zones from the 538 original article: Zones 1, 2 and 3 are the areas behind home plate and the dugouts. Zones 4 and 5 make up most of the foul territory outside the baselines up until the foul pole. Zones 6 and 7 include the areas beyond the foul poles.
Baseball Savant https://baseballsavant.mlb.com/.
The raw data behind the story "Are Democrats Winning The Race For Congress?" https://projects.fivethirtyeight.com/congress-generic-ballot-polls/.
generic_polllist
generic_polllist
A data frame with 934 rows representing polls and 21 variables:
No description provided.
No description provided.
Start date of the poll.
End date of the poll.
The organization that conducted the poll (rather than the organization that paid for or sponsored it)
No description provided.
No description provided.
A = ALL ADULTS, RV = REGISTERED VOTERS, LV = LIKELY VOTERS, V = VOTERS
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
See https://github.com/fivethirtyeight/data/blob/master/congress-generic-ballot/README.md
The raw data behind the story "Are Democrats Winning The Race For Congress?" https://projects.fivethirtyeight.com/congress-generic-ballot-polls/.
generic_topline
generic_topline
A data frame with 751 rows representing polls and 9 variables:
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
See https://github.com/fivethirtyeight/data/blob/master/congress-generic-ballot/README.md
The raw data behind the story "The Media Really Started Paying Attention to Puerto Rico When Trump Did" https://fivethirtyeight.com/features/the-media-really-started-paying-attention-to-puerto-rico-when-trump-did/: Google Trends Data.
google_trends
google_trends
A data frame with 37 rows representing dates and 5 variables:
Date
US Google search interest on the specified date for Hurricane Harvey
US Google search interest on the specified date for Hurricane Irma
US Google search interest on the specified date for Hurricane Maria
US Google search interest on the specified date for Hurricane Jose
Google search interest is measured in search term popularity relative to peak popularity in the given region and time period (with 100 as peak popularity)
Google Trends https://trends.google.com/trends/
mediacloud_hurricanes
, mediacloud_states
, mediacloud_online_news
,
mediacloud_trump
, tv_hurricanes
, tv_hurricanes_by_network
,
tv_states
The raw data behind the story 'Forecasting the races for governor' https://projects.fivethirtyeight.com/2018-midterm-election-forecast/governor/
governor_national_forecast
governor_national_forecast
A dataframe with 150 rows representing national-level results of the classic, lite, and deluxe gubernatorial forecasts since Oct. 11, 2018. and 11 variables
date of the forecast
the party of the forecast
the model of the forecast
the probability of the corresponding party winning
the mean of the number of seats
the median number of seats
the top 10 percentile of number of seats
the top 90 percentile of number of seats
unknown
the margin of p10_seats
the margin of p90_seats
The original dataset included a meaningless column called "state", and all variables under this column was "US". So this column was removed.
FiveThirtyEight’s House, Senate And Governor Models Methodology: https://fivethirtyeight.com/methodology/how-fivethirtyeights-house-and-senate-models-work/
The raw data behind the story 'Forecasting the races for governor' https://projects.fivethirtyeight.com/2018-midterm-election-forecast/governor/
governor_state_forecast
governor_state_forecast
A dataframe with 7743 rows representing state-level results of the classic, lite, and deluxe gubernatorial forecasts since Oct. 11, 2018. and 10 variables
date of the forecast
state of the forecast
name of the candidate
party of the candidate
whether the candidate is incumbent
the model of the forecast
the probability of the corresponding party winning
the voteshare of the corresponding party
the top 10 percentile of the voteshare
the top 00 percentile of the voteshare
The original dataset included two empty column "district" and "special",which were removed.
FiveThirtyEight’s House, Senate And Governor Models Methodology: https://fivethirtyeight.com/methodology/how-fivethirtyeights-house-and-senate-models-work/
The raw data behind the story "Higher Rates Of Hate Crimes Are Tied To Income Inequality" https://fivethirtyeight.com/features/higher-rates-of-hate-crimes-are-tied-to-income-inequality/.
hate_crimes
hate_crimes
A data frame with 51 rows representing US states and DC and 13 variables:
State name
State abbreviation
Median household income, 2016
Share of the population that is unemployed (seasonally adjusted), Sept. 2016
Share of the population that lives in metropolitan areas, 2015
Share of adults 25 and older with a high-school degree, 2009
Share of the population that are not U.S. citizens, 2015
Share of white residents who are living in poverty, 2015
Gini Index, 2015
Share of the population that is not white, 2015
Share of 2016 U.S. presidential voters who voted for Donald Trump
Hate crimes per 100,000 population, Southern Poverty Law Center, Nov. 9-18, 2016
Average annual hate crimes per 100,000 population, FBI, 2010-2015
See https://github.com/fivethirtyeight/data/tree/master/hate-crimes
library(ggplot2) ggplot(hate_crimes, aes(x = share_vote_trump, y = hate_crimes_per_100k_splc)) + geom_text(aes(label = state_abbrev)) + geom_smooth(se = FALSE, method = "lm") + labs(x = "Proportion of votes for Donald Trump", y = "Hate crimes per 100k during Nov 9-18, 2016 (SPLC)", title = "Relationship between Trump support & hate crimes")
library(ggplot2) ggplot(hate_crimes, aes(x = share_vote_trump, y = hate_crimes_per_100k_splc)) + geom_text(aes(label = state_abbrev)) + geom_smooth(se = FALSE, method = "lm") + labs(x = "Proportion of votes for Donald Trump", y = "Hate crimes per 100k during Nov 9-18, 2016 (SPLC)", title = "Relationship between Trump support & hate crimes")
The raw data behind the story "Hip-Hop Is Turning On Donald Trump" https://projects.fivethirtyeight.com/clinton-trump-hip-hop-lyrics/.
hiphop_cand_lyrics
hiphop_cand_lyrics
A data frame with 377 rows representing hip-hop songs referencing POTUS candidates in 2016 and 8 variables:
Candidate referenced
Song name
Artist name
Positive, negative or neutral
Theme of lyric
Date of album release
Lyrics
Genius link
Genius https://genius.com/
The raw data behind the story "The NCAA Bracket: Checking Our Work" https://fivethirtyeight.com/features/the-ncaa-bracket-checking-our-work/.
hist_ncaa_bball_casts
hist_ncaa_bball_casts
A data frame with 253 rows representing NCAA men's basketball tournament games and 6 variables:
See https://fivethirtyeight.com/features/the-ncaa-bracket-checking-our-work/
The raw data behind the story "How The FiveThirtyEight Senate Forecast Model Works" https://fivethirtyeight.com/features/how-the-fivethirtyeight-senate-forecast-model-works/.
hist_senate_preds
hist_senate_preds
A data frame with 207 rows representing US state elections and 5 variables:
Election
Year of election
Last name
Probability of winning election per FiveThirtyEight Election Day forecast
'Win' or 'Loss'
See https://github.com/fivethirtyeight/data/tree/master/forecast-methodology
The raw data behind the story 'Forecasting the race for the House' https://projects.fivethirtyeight.com/2018-midterm-election-forecast/house/
house_national_forecast
house_national_forecast
A dataframe with 588 rows representing district-level results of the classic, lite, and deluxe house forecasts since 2018/08/01 and 11 variables.
date of the forecast
the party of the forecast
the model of the forecast
the probability of the corresponding party winning
the mean of the number of seats
the median number of seats
the top 10 percentile of number of seats
the top 90 percentile of number of seats
unknown
the margin of p10_seats
the margin of p90_seats
The original dataset included a meaningless column called "state", and all variables under this column was "US". So this column was removed.
FiveThirtyEight’s House, Senate And Governor Models Methodology: https://fivethirtyeight.com/methodology/how-fivethirtyeights-house-and-senate-models-work/
Raw data behind this story "Do Americans Support Impeaching Trump?" https://projects.fivethirtyeight.com/impeachment-polls/
impeachment_polls
impeachment_polls
A data frame with 388 rows of polling data and 24 variables:
Poll start date, the first date responses were collected
Poll end date, the last date responses were collected
entity/organization that created poll, collected and published data
sponsor of pollster
number of respondents for each
categorical variable with 3 categories: a, rv, lv – value unknown
true/false logical – value unknown
poll question
category of poll question with 5 categories: impeach and remove, begin proceedings, begin inquiry, reasons, impeach
yes/no logical – value unknown
Percent of respondents in sample who answered "Yes" to the poll question
Percent of respondents in sample who answered "No" to the poll question
Percent of respondents in sample who did not answer "Yes" or "No" to the poll question
number of Republican respondents in sample
Percent of Republican respondents who answered "yes"
Percent of Republican respondents who answered "no"
number of Democrat respondents in sample
Percent of Democrat respondents who answered "yes"
Percent of Democrat respondents who answered "no"
number of Independent respondents in sample
Percent of Independent respondents who answered "yes"
Percent of Independent respondents who answered "no"
URL links to poll websites
any notes relating to polls in sample
data from https://github.com/fivethirtyeight/data/tree/master/impeachment-polls.
The raw data behind the story "Where Are America's Librarians?" https://fivethirtyeight.com/features/where-are-americas-librarians/.
librarians
librarians
A data frame with 371 rows representing areas in the US and 9 variables:
Bureau of Labor Statistics
The raw data behind the story "The Definitive Analysis Of 'Love Actually,' The Greatest Christmas Movie Of Our Time" https://fivethirtyeight.com/features/some-people-are-too-superstitious-to-have-a-baby-on-friday-the-13th/. The adjacency matrix of which actors appear in the same scene together.
love_actually_adj
love_actually_adj
A data frame with 14 rows representing actors and 15 variables:
The raw data behind the story "The Definitive Analysis Of 'Love Actually,' The Greatest Christmas Movie Of Our Time" https://fivethirtyeight.com/features/the-definitive-analysis-of-love-actually-the-greatest-christmas-movie-of-our-time/. A table of the central actors in "Love Actually" and which scenes they appear in.
love_actually_appearance
love_actually_appearance
A data frame with 71 rows representing scenes and 15 variables:
# To convert data frame to tidy data (long) format, run: library(dplyr) library(tidyr) library(stringr) love_actually_appearance_tidy <- love_actually_appearance %>% pivot_longer(-scenes, names_to = "actor", values_to = "appears") %>% arrange(scenes)
# To convert data frame to tidy data (long) format, run: library(dplyr) library(tidyr) library(stringr) love_actually_appearance_tidy <- love_actually_appearance %>% pivot_longer(-scenes, names_to = "actor", values_to = "appears") %>% arrange(scenes)
The raw data behind the story ""Mad Men" Is Ending. What's Next For The Cast?" https://fivethirtyeight.com/features/mad-men-is-ending-whats-next-for-the-cast/.
mad_men
mad_men
A data frame with 248 rows representing performers on TV shows and 15 variables:
The name of the actor, according to IMDb. This is not a unique identifier - two performers appeared in more than one program
The television show where this actor appeared in more than half the episodes
The year the television show began
The year the television show ended, "PRESENT" if the show remains on the air as of May 10.
Why the actor is no longer on the program: "END" if the show has concluded, "LEFT" if the show remains on the air.
The year the character left the show. Equal to "Show End" if the performer stayed on until the final season.
2015 minus CharEnd
The number of leading roles in films the performer has appeared in since and including "CharEnd", according to OpusData
The number of leading roles in films the performer has appeared in since and including "CharEnd", according to OpusData
The number of seasons of television of which the performer appeared in at least half the episodes since and including "CharEnd", according to OpusData
#LEAD + #Shows + 0.25*(#SUPPORT)
"Score" divided by "Years Since"
The list of films counted in #LEAD
The list of films counted in #SUPPORT
The seasons of shows counted in #Shows
IMDB https://imdb.com
The raw data behind the story "Dear Mona, How Many Flight Attendants Are Men?" https://fivethirtyeight.com/features/dear-mona-how-many-flight-attendants-are-men/.
male_flight_attend
male_flight_attend
A data frame with 320 rows representing job categories and 2 variables:
Category of job
Percentage of workforce that are male
IPUMS 2012 https://usa.ipums.org/usa/
This folder contains the data behind the story: "What Do Men Think It Means To Be A Man?" https://fivethirtyeight.com/features/what-do-men-think-it-means-to-be-a-man/
masculinity_survey
masculinity_survey
A dataset with 189 rows representing answers and 12 variables:
the survey question
the survey response
the ratio of overall participants who selected this response
the ratio of participants age 18 to 34 who selected this response
the ratio of participants age 35 to 64 who selected this response
the ratio of participants age 65 or over who selected this response
the ratio of overall white participants who selected this response
the ratio of overall non-white participants who selected this response
the ratio of participants who have child(ren) who selected this response
the ratio of participants who do not have children who selected this response
the ratio of straight participants who selected this response
the ratio of non-straight participants who selected this response
The original 'masculinity-survey.csv' contains the results of a survey of 1,615 adult men conducted by SurveyMonkey in partnership with FiveThirtyEight and WNYC Studios from May 10-22, 2018. The modeled error estimate for this survey is plus or minus 2.5 percentage points. The percentages have been weighted for age, race, education, and geography using the Census Bureau’s American Community Survey to reflect the demographic composition of the United States age 18 and over. Crosstabs with less than 100 respondents have been left blank because responses would not be statistically significant. I made heavy editions in Excel to make the dataset easily usable in R.
The original survey responses and original datasets can be found here: https://github.com/fivethirtyeight/data/tree/master/masculinity-survey
library(dplyr) library(ggplot2) library(tidyr) library(stringr) # Data wrangling masculinity_tidy <- masculinity_survey %>% # Narrow down rows to those pertaining to first question of survey: filter(question == 'In general, how masculine or "manly" do you feel?') %>% # Eliminate columns not relating to sexual orientaiton: select(-c(age_18_34, age_35_64, age_65_over, white_yes, white_no, children_yes, children_no, overall)) %>% # Convert data frame to tidy data (long) format: pivot_longer(-c(question, response), names_to = "sexuality", values_to = "ratio_by_sexuality") # Visualize results ggplot(data = masculinity_tidy, aes(x = response, y = ratio_by_sexuality, fill = sexuality)) + geom_bar(stat="identity", position = 'dodge') + labs(x = "Response", y = "Proportion", labs = "Sexuality", title = "In general, how masculine or 'manly' do you feel?")
library(dplyr) library(ggplot2) library(tidyr) library(stringr) # Data wrangling masculinity_tidy <- masculinity_survey %>% # Narrow down rows to those pertaining to first question of survey: filter(question == 'In general, how masculine or "manly" do you feel?') %>% # Eliminate columns not relating to sexual orientaiton: select(-c(age_18_34, age_35_64, age_65_over, white_yes, white_no, children_yes, children_no, overall)) %>% # Convert data frame to tidy data (long) format: pivot_longer(-c(question, response), names_to = "sexuality", values_to = "ratio_by_sexuality") # Visualize results ggplot(data = masculinity_tidy, aes(x = response, y = ratio_by_sexuality, fill = sexuality)) + geom_bar(stat="identity", position = 'dodge') + labs(x = "Response", y = "Proportion", labs = "Sexuality", title = "In general, how masculine or 'manly' do you feel?")
The raw data behind the story "Beto O'Rourke Ignored Cable News - And It Ignored Him" https://fivethirtyeight.com/features/beto-orourke-ignored-cable-news-and-it-ignored-him/
media_mentions_cable media_mentions_online
media_mentions_cable media_mentions_online
2 dataframes about 2020 presidential candidate media mentions
An object of class spec_tbl_df
(inherits from tbl_df
, tbl
, data.frame
) with 954 rows and 6 columns.
A data frame with 972 rows representing weeks of cable coverage and 7 variables:
start date for the week of coverage
candidate's name
number of 15-second clips in that week that mention the specified candidate
number of 15-second clips in that week that mention any candidates
total number of 15-second clips that week across the three networks
percentage of clips in which that specific candidate is mentioned out of all clips mentioning any candidate for that week (matched_clips / all_candidate_clips)
query used for the GDELT Television API
A data frame with 954 rows representing weeks and 6 variables:
start date for the week of coverage
candidate's name
number of stories in that week that mention the specified candidate
number of stories in that week that mention any candidate
percentage of stories in which that specific candidate is mentioned out of all stories mentioning any candidate for that week (matched_stories / all_candidate_stories)
query for Media Cloud
The GDELT Television API https://blog.gdeltproject.org/gdelt-2-0-television-api-debuts/, which processes the data from the TV News Archive https://archive.org/details/tv.
Two collections in the Media Cloud https://mediacloud.org/ database U.S. Top Online News https://sources.mediacloud.org/#/collections/58722749 and U.S. Top Digital Native News https://sources.mediacloud.org/#/collections/57078150
The raw data behind the story "The Media Really Started Paying Attention to Puerto Rico When Trump Did" https://fivethirtyeight.com/features/the-media-really-started-paying-attention-to-puerto-rico-when-trump-did/: Mediacloud Hurricanes Data.
mediacloud_hurricanes
mediacloud_hurricanes
A data frame with 38 rows representing dates and 5 variables:
Date
The number of sentences in online news which mention Hurricane Harvey on the specified date
The number of sentences in online news which mention Hurricane Irma
The number of sentences in online news which mention Hurricane Maria
The number of sentences in online news which mention Hurricane Jose
Mediacloud https://mediacloud.org/
mediacloud_states
, mediacloud_online_news
,
mediacloud_trump
, tv_hurricanes
, tv_hurricanes_by_network
,
tv_states
, google_trends
The raw data behind the story "The Media Really Started Paying Attention to Puerto Rico When Trump Did" https://fivethirtyeight.com/features/the-media-really-started-paying-attention-to-puerto-rico-when-trump-did/: Mediacloud Top Online News Data.
mediacloud_online_news
mediacloud_online_news
A data frame with 49 rows representing media outlets and 2 variables:
Name of media outlet source included in Media Cloud's "U.S. Top Online News" collection
URL of corresponding media outlet source
Mediacloud https://mediacloud.org/
mediacloud_hurricanes
, mediacloud_states
,
mediacloud_trump
, tv_hurricanes
, tv_hurricanes_by_network
,
tv_states
, google_trends
The raw data behind the story "The Media Really Started Paying Attention to Puerto Rico When Trump Did" https://fivethirtyeight.com/features/the-media-really-started-paying-attention-to-puerto-rico-when-trump-did/: Mediacloud States Data.
mediacloud_states
mediacloud_states
A data frame with 51 rows representing dates and 4 variables:
Date
The number of sentences in online news which mention Texas on the specified date
The number of sentences in online news which mention Puerto Rico
The number of sentences in online news which mention Florida
Mediacloud https://mediacloud.org/
mediacloud_hurricanes
, mediacloud_online_news
,
mediacloud_trump
, tv_hurricanes
, tv_hurricanes_by_network
,
tv_states
, google_trends
The raw data behind the story "The Media Really Started Paying Attention to Puerto Rico When Trump Did" https://fivethirtyeight.com/features/the-media-really-started-paying-attention-to-puerto-rico-when-trump-did/: Mediacloud Trump Data.
mediacloud_trump
mediacloud_trump
A data frame with 51 rows representing dates and 7 variables:
Date
The number of headlines that mention Puerto Rico on the given date
The number of headlines that mention Puerto Rico and either President or Trump
The number of headlines that mention Florida
The number of headlines that mention Florida and either President or Trump
The number of headlines that mention Texas
The number of headlines that mention Texas and either President or Trump
Mediacloud https://mediacloud.org/
mediacloud_hurricanes
, mediacloud_states
,
mediacloud_online_news
,tv_hurricanes
,
tv_hurricanes_by_network
, tv_states
, google_trends
The raw data behind the story "The Best MLB All-Star Teams Ever" https://fivethirtyeight.com/features/the-best-mlb-all-star-teams-ever/.
mlb_as_play_talent
mlb_as_play_talent
A data frame with 3930 rows representing Major League Baseball players in given seasons and 15 variables:
Player's ID at Baseball-Reference.com
The season in question
Order of All-Star Game for the season (in years w/ multiple ASGs; set to 0 when only 1 per year)
Game ID at Baseball-Reference.com
League of All-Star team
Position (according to baseball convention; 1=pitcher, 2=catcher, etc.) if starter
Estimate of offensive talent, in runs above league average per 600 plate appearances
Estimate of fielding talent, in runs above league average per 600 plate appearances
Estimate of pitching talent, in runs above league average per 200 innings pitched
Number of plate appearances in the All-Star Game itself
Number of innings pitched in the All-Star Game itself
Expected offensive runs added above average (from talent) based on PA in ASG, scaled to a 9-inning game
Expected defensive runs added above average (from talent) based on PA in ASG, scaled to a 9-inning game
Expected pitching runs added above average (from talent) based on IP in ASG, scaled to a 9-inning game
Expected runs added above average (from talent) based on PA/IP in ASG, scaled to a 9-inning game
https://www.baseball-reference.com/ , http://chadwick-bureau.com, Fangraphs
The raw data behind the story "The Best MLB All-Star Teams Ever" https://fivethirtyeight.com/features/the-best-mlb-all-star-teams-ever/.
mlb_as_team_talent
mlb_as_team_talent
A data frame with 172 rows representing Major League Baseball seasons and 16 variables:
The season in question
Order of All-Star Game for the season (in years w/ multiple ASGs; set to 0 when only 1 per year)
Game ID at Baseball-Reference.com
League of All-Star team
Total runs of offensive talent above average per game (36 plate appearances)
Total runs of fielding talent above average per game (36 plate appearances)
Total runs of pitching talent above average per game (9 innings)
MLB average runs scored/game that season
Expected runs scored per game based on talent (MLB R/G + team OFF talent)
Expected runs allowed per game based on talent (MLB R/G - team DEF talent- team PIT talent)
Unadjusted Pythagorean talent rating; PYTH =(RSPG^1.83)/(RSPG^1.83+RAPG^1.83)
Estimate of relative league quality where 2015 MLB = 1.00
Strength of schedule faced; adjusts an assumed .500 SOS downward based on timeline adjustment
Adjusted Pythagorean record; =(SOS*unadj_Pyth)/((2*unadj_Pyth*SOS)-SOS-unadj_Pyth+1)
Best player according to combo of actual PA/IP and talent
2nd-best player according to combo of actual PA/IP and talent
https://www.baseball-reference.com/ , http://chadwick-bureau.com, Fangraphs
The raw data behind the story 'Both Parties Think The Mueller Report Was Fair. They Just Completely Disagree On What It Says.' https://fivethirtyeight.com/features/both-parties-think-the-mueller-report-was-fair-they-just-disagree-on-what-it-says/
mueller_approval_polls
mueller_approval_polls
A dataset with 65 rows representing every job approval poll of Robert Mueller that we could find from when Mueller was appointed as special council on May 17, 2017 through May 3, 2019 and 12 variables
the start date of the poll
the end date of the poll
the name of the pollster
the size of the poll sample
unknown
the text of the poll question
the number of approval in the poll
the number of disapproval in the poll
the number of unsure in the poll
the number of approval from Republican
the number of approval from Democrat
the url of the poll
Polls, Washington Post / ABC and Washington Post / Schar School Polls
The raw data behind the story "A Handful Of Cities Are Driving 2016's Rise In Murder" https://fivethirtyeight.com/features/a-handful-of-cities-are-driving-2016s-rise-in-murders/.
murder_2015_final
murder_2015_final
A data frame with 83 rows representing large US cities and 5 variables:
Name of city
Name of state
Total murders in 2014
Total murders in 2015
2015 - 2014
Unknown
The raw data behind the story "A Handful Of Cities Are Driving 2016's Rise In Murder" https://fivethirtyeight.com/features/a-handful-of-cities-are-driving-2016s-rise-in-murders/.
murder_2016_prelim
murder_2016_prelim
A data frame with 79 rows representing large US cities and 7 variables:
Name of city
Name of state
Number of murders in 2015
Number of murder in 2016 (as of as_of
date)
2016 - 2015
Source of data
2016 murders up to this date
Listed as source
variable in dataset
The raw data behind the story "Projecting The Top 50 Players In The 2015 NBA Draft Class" https://fivethirtyeight.com/features/projecting-the-top-50-players-in-the-2015-nba-draft-class/. An analysis using this data was contributed by G. Elliott Morris as a package vignette at https://fivethirtyeightdata.github.io/fivethirtyeightdata/articles/NBA.html.
nba_draft_2015
nba_draft_2015
A data frame with 1090 rows representing National Basketball Association players/prospects and 9 variables:
Player name
The player's position going into the draft
The player's identification code
The year the player was eligible for the NBA draft
The model's projected statistical plus/minus over years 2-5 of the player's NBA career
Probability of becoming a superstar player (1 per draft, SPM >= +3.3)
Probability of becoming a starting-caliber player (10 per draft, SPM >= +0.5)
Probability of becoming a role player (25 per draft, SPM >= -1.4)
Probability of becoming a bust (everyone else, SPM < -1.4)
See https://fivethirtyeight.com/features/projecting-the-top-50-players-in-the-2015-nba-draft-class/
The raw data behind the story "A Better Way to Evaluate NBA Defense" https://fivethirtyeight.com/features/a-better-way-to-evaluate-nba-defense/.
nba_draymond
nba_draymond
A data frame with 3009 rows representing DRAYMOND ratings (Defensive Rating Accounting for Yielding Minimal Openness by Nearest Defender) for every player since the 2013-14 season with 4 variables:
The second year of the season; for example, 2018-2019 season would be listed as 2019
Name of the player
Number of possessions a player during the season
Defensive Rating Accounting for Yielding Minimal Openness by Nearest Defender
see https://github.com/fivethirtyeight/data/tree/master/nba-draymond
The raw data behind all nba predictions, including the story "The Complete History of the NBA" https://projects.fivethirtyeight.com/complete-history-of-the-nba
nba_elo_latest
nba_elo_latest
An object of class spec_tbl_df
(inherits from tbl_df
, tbl
, data.frame
) with 1230 rows and 24 columns.
A data frame with 1230 rows representing game played during the most current season of the NBA, and 24 variables:
Date
the season in which the game was played
True if the game was played on neutral territory, False if not
True if the game was played in a playoff, False if not
name of first team
name of second team
Team 1 Elo rating before game
Team 2 Elo rating before game
Team 1's probability of winning based on Elo rating
Team 2's probability of winning based on Elo rating
Team 1 Elo rating after the game
Team 2 Elo rating after the game
the score of team 1
the score of team 2
The raw data behind the story "Accurately Counting NBA Tattoos Isn't Easy, Even If You're Up Close" https://fivethirtyeight.com/features/accurately-counting-nba-tattoos-isnt-easy-even-if-youre-up-close/.
nba_tattoos
nba_tattoos
A data frame with 636 rows representing National Basketball Association players and 2 variables:
Name of player
TRUE corresponds to player having tattoos, FALSE corresponds to not
Ethan Swan https://nbatattoos.tumblr.com/
The raw data behind the story 'The Rise And Fall Of Women's NCAA Tournament Dynasties' https://fivethirtyeight.com/features/louisiana-tech-was-the-uconn-of-the-80s/
ncaa_w_bball_tourney
ncaa_w_bball_tourney
A dataset with 2092 rows representing every team that has participated in the NCAA Division I Women's Basketball Tournament since it began in 1982 and 19 variables
the year of the game which the team participated in
the school of the participating team
The '(OR)' seeding designation in 1983 notes the eight teams that played an opening-round game to become the No. 8 seed in each region.
the conference record of the team (if available)
number of winning in conference record
number of losses in conference record
percent of winning in conference record
number of winning in regular-season record
number of losses in regular-season record
percent of winning in regular-season record
Whether the school qualified with an automatic bid (by winning its conference or conference tournament) or an at-large bid.
Whether the school played its first-round tournament games on its home court.
number of winning in tournament record
number of losses in tournament record
The round of the final game for each team. OR=opening-round loss (1983 only); 1st=first-round loss; 2nd=second-round loss; RSF=loss in the Sweet 16; RF=loss in the Elite Eight; NSF=loss in the national semifinals; N2nd=national runner-up; Champ=national champions
number of winning in full record
number of losses in full record
percent of winning in full record
NCAA
The raw data behind the story "How Every NFL Team’s Fans Lean Politically" https://fivethirtyeight.com/features/how-every-nfl-teams-fans-lean-politically/: Google Trends Data.
nfl_fandom_google
nfl_fandom_google
a data frame with 207 rows representing designated market areas and 9 variables:
Designated Market Area
The percentage of search traffic in the media market region related to the NFL over the past 5 years
The percentage of search traffic in the region related to the NBA over the past 5 years
The percentage of search traffic in the region related to the MLB over the past 5 years
The percentage of search traffic in the region related to NASCAR over the past 5 years
The percentage of search traffic in the region related to the CBB over the past 5 years
The percentage of search traffic in the region related to the CFB over the past 5 years
The percentage of voters in the region who voted for Trump in the 2016 Presidential Election
Google Trends https://trends.google.com/trends/.
# To convert data frame to tidy data (long) format, run: library(dplyr) library(tidyr) nfl_fandom_google_tidy <- nfl_fandom_google %>% pivot_longer(-c("dma", "trump_2016_vote"), names_to = "sport", values_to = "search_traffic") %>% arrange(dma)
# To convert data frame to tidy data (long) format, run: library(dplyr) library(tidyr) nfl_fandom_google_tidy <- nfl_fandom_google %>% pivot_longer(-c("dma", "trump_2016_vote"), names_to = "sport", values_to = "search_traffic") %>% arrange(dma)
The raw data behind the story "How Every NFL Team’s Fans Lean Politically" https://fivethirtyeight.com/features/how-every-nfl-teams-fans-lean-politically/: SurveyMonkey Data.
nfl_fandom_surveymonkey
nfl_fandom_surveymonkey
a data frame with 33 rows representing teams and 25 variables:
NFL team
Total number of poll respondents who ranked the given team in their top 3 favorites
Number of Asian, democrat poll respondents who ranked the given team in their top 3 favorites
Number of Black, democrat poll respondents who ranked the given team in their top 3 favorites
Number of Hispanic, democrat poll respondents who ranked the given team in their top 3 favorites
Number of democrat poll respondents who identified their race as "other" (not Asian, Black, Hispanic, or White) and ranked the given team in their top 3 favorites
Number of White, democrat poll respondents who ranked the given team in their top 3 favorites
Total number of democrat poll respondents who ranked the given team in their top 3 favorites
Number of Asian, independent poll respondents who ranked the given team in their top 3 favorites
Number of Black, independent poll respondents who ranked the given team in their top 3 favorites
Number of Hispanic, independent poll respondents who ranked the given team in their top 3 favorites
Number of independent poll respondents who identified their race as "other" (not Asian, Black, Hispanic, or White) and ranked the given team in their top 3 favorites
Number of White, independent poll respondents who ranked the given team in their top 3 favorites
Total number of independent poll respondents who ranked the given team in their top 3 favorites
Number of Asian, republican poll respondents who ranked the given team in their top 3 favorites
Number of Black, republican poll respondents who ranked the given team in their top 3 favorites
Number of Hispanic, republican poll respondents who ranked the given team in their top 3 favorites
Number of republican poll respondents who identified their race as "other" (not Asian, Black, Hispanic, or White) and ranked the given team in their top 3 favorites
Number of White, republican poll respondents who ranked the given team in their top 3 favorites
Total number of republican poll respondents who ranked the given team in their top 3 favorites
Percent of fans (who ranked the team in their top 3 favorite NFL teams) who are republicans
Percent of fans who are democrats
Percent of fans who are independent
Percent of fans who are White
Percent of fans who are not White
See https://github.com/fivethirtyeight/data/blob/master/nfl-fandom/NFL_fandom_data-surveymonkey.csv
# To convert data frame to tidy data (long) format, run: library(dplyr) library(tidyr) nfl_fandom_surveymonkey_tidy <- nfl_fandom_surveymonkey %>% pivot_longer(-c("team", "total_respondents", "gop_percent", "dem_percent", "ind_percent", "white_percent", "nonwhite_percent"), names_to = "race_party", values_to = "percent") %>% arrange(team)
# To convert data frame to tidy data (long) format, run: library(dplyr) library(tidyr) nfl_fandom_surveymonkey_tidy <- nfl_fandom_surveymonkey %>% pivot_longer(-c("team", "total_respondents", "gop_percent", "dem_percent", "ind_percent", "white_percent", "nonwhite_percent"), names_to = "race_party", values_to = "percent") %>% arrange(team)
The raw data behind the story "The Rams Are Dead To Me, So I Answered 3,352 Questions To Find A New NFL Team" https://fivethirtyeight.com/features/the-rams-are-dead-to-me-so-i-answered-3352-questions-to-find-a-new-team/.
nfl_fav_team
nfl_fav_team
A data frame with 32 rows representing National Football League teams and 17 variables:
Name of NFL team
Fan relations - Courtesy by players, coaches and front offices toward fans, and how well a team uses technology to reach them
Ownership - Honesty; loyalty to core players and the community
Players - Effort on the field, likability off it
Future wins - Projected wins over next 5 seasons
Bandwagon Factor - Are the team's next 5 years likely to be better than their previous 5?
Tradition - Championships/division titles/wins in team's entire history
Bang for the buck - Wins per fan dollars spent
Behavior - Suspensions by players on team since 2007, with extra weight to transgressions vs. women
Proximity to New York City
Proximity to St. Louis
Affordability - Price of tickets, parking and concessions
Small Market - Size of market in terms of population, where smaller is better
Stadium experience - Quality of venue; fan-friendliness of environment; frequency of game-day promotions
Coaching - Strength of on-field leadership
Uniform - Stylishness of uniform design, according to Uni Watch's Paul Lukas
Big Market - Size of market in terms of population, where bigger is better
https://www.allourideas.org/nflteampickingsample
The raw data behind the story "The NFL's Uneven History Of Punishing Domestic Violence" https://fivethirtyeight.com/features/nfl-domestic-violence-policy-suspensions/.
nfl_suspensions
nfl_suspensions
A data frame with 269 rows representing National Football League players and 7 variables:
first initial.last name
team at time of suspension
number of games suspended (one regular season = 16 games)
personal conduct, substance abuse, performance enhancing drugs or in-game violence
description of suspension
year of suspension
news source
https://en.wikipedia.org/wiki/List_of_players_and_coaches_suspended_by_the_NFL, https://www.spotrac.com/fines-tracker/nfl/suspensions/
The raw data behind the story "Who Goes To Meaningless NFL Games And Why?" https://fivethirtyeight.com/features/who-goes-to-meaningless-nfl-games-and-why/.
nfltix_div_avgprice
nfltix_div_avgprice
A data frame with 108 rows representing National Football League games and 3 variables:
NFL divisional game info
NFL division
Average ticket price
StubHub
The raw data behind the story "Who Goes To Meaningless NFL Games And Why?" https://fivethirtyeight.com/features/who-goes-to-meaningless-nfl-games-and-why/.
nfltix_usa_avg
nfltix_usa_avg
A data frame with 32 rows representing National Football League teams and 2 variables:
Name of NFL team
Average ticket price
StubHub
The raw data behind the story "The Football Hall Of Fame Has A Receiver Problem" https://fivethirtyeight.com/features/the-football-hall-of-fame-has-a-receiver-problem/.
nflwr_aging_curve
nflwr_aging_curve
A data frame with 24 rows representing National Football League wide receiver ages and 3 variables:
Beginning age
Ending age
Change in TRY per game from one age-year to next
Unknown
The raw data behind the story "The Football Hall Of Fame Has A Receiver Problem" https://fivethirtyeight.com/features/the-football-hall-of-fame-has-a-receiver-problem/.
nflwr_hist
nflwr_hist
A data frame with 6496 rows representing National Football League wide receivers and 6 variables:
Player identification code at https://www.pro-football-reference.com/
The player's name
Career True Receiving Yards
Adjusted Net Yards Per Attempt (relative to average) of player's career teams, weighted by TRY w/ each team
The amount by which career_ranypa
exceeds what would be expected from his QBs' (age-adjusted) performance without the receiver
The number of yards per game by which a player would outgain an average receiver on the same team, after adjusting for teammate quality and age
See https://fivethirtyeight.com/features/the-football-hall-of-fame-has-a-receiver-problem/
The raw data behind the story "You Can't Trust What You Read About Nutrition" https://fivethirtyeight.com/features/you-cant-trust-what-you-read-about-nutrition/.
nutrition_pvalues
nutrition_pvalues
A data frame with 27716 rows representing Regression fits for p-hacking and 3 variables:
Name of food (response/dependent variable)
Name of characteristic (predictor/independent variable)
P-value from regression fit
See https://fivethirtyeight.com/features/you-cant-trust-what-you-read-about-nutrition/
This directory contains the data for FiveThirtyEight's partisan lean, which is used in our [House] https://projects.fivethirtyeight.com/2018-midterm-election-forecast/house [Senate] https://projects.fivethirtyeight.com/2018-midterm-election-forecast/senate and [Governor] https://projects.fivethirtyeight.com/2018-midterm-election-forecast/governor/ forecasts.
partisan_lean_district
partisan_lean_district
A dataset with 435 rows representing votes and 4 variables
the state of the vote
the district_number of the vote
the party of the vote
the Cook Partisan Voting Index of the vote
The original dataset only has 2 columns: "district" and "pvi_538". I separated each of the 2 columns into two. For example, in row 1 of the dataset, the original "district" = "AK-1", and I separated it into "state" = "Arkansas" and "district_number" = "1"; the original "pvi_538" = "R+15.21", and I separated it into “pvi_party" = "R" and "pvi_amount" = "15.21". In addition, I used the full names for all states instead of abbreviations.
Partisan lean is the average difference between how a state or district votes and how the country votes overall, with 2016 presidential election results weighted 50 percent, 2012 presidential election results weighted 25 percent and results from elections for the state legislature weighted 25 percent.
This directory contains the data for FiveThirtyEight's partisan lean, which is used in our [House] https://projects.fivethirtyeight.com/2018-midterm-election-forecast/house [Senate] https://projects.fivethirtyeight.com/2018-midterm-election-forecast/senate and [Governor] https://projects.fivethirtyeight.com/2018-midterm-election-forecast/governor/ forecasts.
partisan_lean_state
partisan_lean_state
A dataset with 50 rows representing states and 3 variables
the state
the party of the vote
the Cook Partisan Voting Index of the vote
The original dataset only has 2 columns: "state" and "pvi_538". I separated the "pvi_538" columns into two. For example, in row 1 of the dataset, the original "pvi_538" = "R+27", and I separated it into “pvi_party" = "R" and "pvi_amount" = "27".
Partisan lean is the average difference between how a state or district votes and how the country votes overall, with 2016 presidential election results weighted 50 percent, 2012 presidential election results weighted 25 percent and results from elections for the state legislature weighted 25 percent.
The raw data behind the story "The Dallas Shooting Was Among The Deadliest For Police In U.S. History" https://fivethirtyeight.com/features/the-dallas-shooting-was-among-the-deadliest-for-police-in-u-s-history/.
police_deaths
police_deaths
A data frame with 22800 rows representing Police officers/dogs who lost their lives and 7 variables:
Name of person/canine who died
Cause of death
Date of event
Year of event
TRUE if canine, FALSE if human
Name of police department
State of police department
Officer Down Memorial Page https://www.odmp.org/
The raw data behind the story "Where Police Have Killed Americans In 2015" https://fivethirtyeight.com/features/where-police-have-killed-americans-in-2015/.
police_killings
police_killings
A data frame with 467 rows representing People who died from interactions with police and 34 variables:
Name of deceased
Age of deceased
Gender of deceased
Race/ethnicity of deceased
Month of killing
Day of incident
Year of incident
Address/intersection where incident occurred
City where incident occurred
State where incident occurred
Latitude, geocoded from address
Longitude, geocoded from address
State FIPS code
County FIPS code
Tract ID code
Combined tract ID code
Combined county ID code
Tract description
Agency involved in incident
Cause of death
How/whether deceased was armed
Tract population
Share of pop that is non-Hispanic white
Share of pop that is black (alone, not in combination)
Share of pop that is Hispanic/Latino (any race)
Tract-level median personal income
Tract-level median household income
County-level median household income
'h_income' / 'county_income'
Household income, quintile within county
Household income, quintile nationally
Tract-level poverty rate (official)
Tract-level unemployment rate
Share of 25+ pop with BA or higher
See https://github.com/fivethirtyeight/data/tree/master/police-killings
The raw data behind the story "Most Police Don't Live In The Cities They Serve" https://fivethirtyeight.com/features/most-police-dont-live-in-the-cities-they-serve/.
police_locals
police_locals
A data frame with 75 rows representing cities and 8 variables:
U.S. city
Number of police officers serving that city
Percentage of the total police force that lives in the city
Percentage of white (non-Hispanic) police officers who live in the city
Percentage of non-white police officers who live in the city
Percentage of black police officers who live in the city
Percentage of Hispanic police officers who live in the city
Percentage of Asian police officers who live in the city
The dataset includes the cities with the 75 largest police forces, with the exception of Honolulu for which data is not available. All calculations are based on data from the U.S. Census.
The Census Bureau numbers are potentially going to differ from other counts for three reasons:
The census category for police officers also includes sheriffs, transit police and others who might not be under the same jurisdiction as a city's police department proper. The census category won't include private security officers.
The census data is estimated from 2006 to 2010; police forces may have changed in size since then.
There is always a margin of error in census numbers; they are estimates, not complete counts.
Note: Missing values means that there are fewer than 100 police officers of that race serving that city.
See https://github.com/fivethirtyeight/data/tree/master/police-locals
# To convert data frame to tidy data (long) format, run: library(dplyr) library(tidyr) police_locals_tidy <- police_locals %>% pivot_longer(all:asian, names_to = "race", values_to = "perc_in")
# To convert data frame to tidy data (long) format, run: library(dplyr) library(tidyr) police_locals_tidy <- police_locals %>% pivot_longer(all:asian, names_to = "race", values_to = "perc_in")
The raw data behind the story "The Last 10 Weeks Of 2016 Campaign Stops In One Handy Gif" https://fivethirtyeight.com/features/the-last-10-weeks-of-2016-campaign-stops-in-one-handy-gif/.
pres_2016_trail
pres_2016_trail
A data frame with 177 rows representing 2016 Republican and Democratic candidate campaign trail stops and 5 variables:
Clinton or Trump
The date of the event
The location of the event
Latitude of the event location
Longitude of the event location
https://hillaryspeeches.com/, https://www.conservativedailynews.com/
The raw data behind the story "Sitting Presidents Give Way More Commencement Speeches Than They Used To" https://fivethirtyeight.com/features/sitting-presidents-give-way-more-commencement-speeches-than-they-used-to/.
pres_commencement
pres_commencement
A data frame with 154 rows representing speeches and 8 variables:
Number of president (33 is Harry Truman, the 33rd president; 44 is Barack Obama, the 44th president)
Name of president
Description of commencement speech
Date speech was delivered
City where speech was delivered
State where speech was delivered
Name of building in which speech was delivered
Room in which speech was delivered
American Presidency Project, Gerhard Peters and John T. Woolley https://www.presidency.ucsb.edu
The raw data behind the story "Do Pulitzers Help Newspapers Keep Readers?" https://fivethirtyeight.com/features/do-pulitzers-help-newspapers-keep-readers/.
pulitzer
pulitzer
A data frame with 50 rows representing newspapers and 7 variables:
Newspaper
Daily Circulation in 2004
Daily Circulation in 2013
Percent change in Daily Circulation from 2004 to 2013
Number of Pulitzer Prize winners and finalists from 1990 to 2003
Number of Pulitzer Prize winners and finalists from 2004 to 2014
Number of Pulitzer Prize winners and finalists from 1990 to 2014
See https://fivethirtyeight.com/features/do-pulitzers-help-newspapers-keep-readers/
The raw data behind the story "Can You Rule Riddler Nation?" https://fivethirtyeight.com/features/can-you-rule-riddler-nation/. Analysis of the submitted solutions can be found at: https://fivethirtyeight.com/features/can-you-save-the-drowning-swimmer/
riddler_castles
riddler_castles
A data frame with 1387 rows representing submissions and 11 variables:
Number of troops out of 100 send to castle 1
Number of troops out of 100 send to castle 2
Number of troops out of 100 send to castle 3
Number of troops out of 100 send to castle 4
Number of troops out of 100 send to castle 5
Number of troops out of 100 send to castle 6
Number of troops out of 100 send to castle 7
Number of troops out of 100 send to castle 8
Number of troops out of 100 send to castle 9
Number of troops out of 100 send to castle 10
Why did you choose your troop deployment?
See https://github.com/fivethirtyeight/data/tree/master/riddler-castles
# To convert data frame to tidy data (long) format, run library(dplyr) library(tidyr) library(stringr) riddler_castles_tidy<-riddler_castles %>% pivot_longer(castle1:castle10, names_to = "castle" , values_to = "soldiers") %>% mutate(castle = as.numeric(str_replace(castle, "castle","")))
# To convert data frame to tidy data (long) format, run library(dplyr) library(tidyr) library(stringr) riddler_castles_tidy<-riddler_castles %>% pivot_longer(castle1:castle10, names_to = "castle" , values_to = "soldiers") %>% mutate(castle = as.numeric(str_replace(castle, "castle","")))
The raw data behind the story "The Battle For Riddler Nation, Round 2" https://fivethirtyeight.com/features/the-battle-for-riddler-nation-round-2/. Analysis of the submitted solutions can be found at: https://fivethirtyeight.com/features/how-much-should-you-bid-for-that-painting/
riddler_castles2
riddler_castles2
A data frame with 932 rows representing submissions and 11 variables:
Number of troops out of 100 send to castle 1
Number of troops out of 100 send to castle 2
Number of troops out of 100 send to castle 3
Number of troops out of 100 send to castle 4
Number of troops out of 100 send to castle 5
Number of troops out of 100 send to castle 6
Number of troops out of 100 send to castle 7
Number of troops out of 100 send to castle 8
Number of troops out of 100 send to castle 9
Number of troops out of 100 send to castle 10
Why did you choose your troop deployment?
See https://github.com/fivethirtyeight/data/tree/master/riddler-castles
# To convert data frame to tidy data (long) format, run library(dplyr) library(tidyr) library(stringr) riddler_castles_tidy<-riddler_castles2 %>% pivot_longer(castle1:castle10, names_to = "castle" , values_to = "soldiers") %>% mutate(castle = as.numeric(str_replace(castle, "castle","")))
# To convert data frame to tidy data (long) format, run library(dplyr) library(tidyr) library(stringr) riddler_castles_tidy<-riddler_castles2 %>% pivot_longer(castle1:castle10, names_to = "castle" , values_to = "soldiers") %>% mutate(castle = as.numeric(str_replace(castle, "castle","")))
The raw data behind the story "Pick A Number, Any Number" https://fivethirtyeight.com/features/pick-a-number-any-number/
riddler_pick_lowest
riddler_pick_lowest
A data frame with 3660 rows representing dates and 1 variable:
Guessed number
People showing their work
This folder contains data behind the story 'Is The Russia Investigation Really Another Watergate?' https://projects.fivethirtyeight.com/russia-investigation/
russia_investigation
russia_investigation
A dataset with 194 rows representing every special investigation since the Watergate probe began in 1973 and 13 variables
Unique id for each investigation
Start date of the investigation
End date of the investigation
Length, in days, of the investigation. Days will be negative if the charge occurred before the investigation began.
Name of the person charged (if applicable). Will be blank if there were no charges.
Length, in days, from the start of the investigation to the date the person was charged (if applicable). Days will be negative if the charge occurred before the investigation began.
Result of charge (if applicable)
Date the person plead guilty or was convicted (if applicable)
Length, in days, from the start of the investigation to the date the person plead guilty or was convicted (if applicable)
Whether or not the relevant person's conviction was overturned
Whether or not the relevant person's charge was pardoned
Whether or not the relevant person's charge was a U.S. resident
President at the center of the investigation
Information for this story is drawn from an original data set of special counsel, independent counsel and special prosecutor investigations from 1973 to 2019. The data set was created by consulting historical sources, including final reports generated by independent counsels, special counsels and special prosecutors; reports in Congressional Quarterly; and contemporaneous news stories. Secondary historical sources were also consulted, including a 2006 Congressional Research Service report about independent counsel investigations and a history of the Watergate investigation by Stanley Kutler. Data about pardons was obtained from the Office of the Pardon Attorney. Indicted organizations were excluded from our analysis. The data set, which is available on Github, includes the names of all people charged as part of these investigations, as well as the outcome of their cases and the dates of major actions in their cases.
2006 Congressional Research Service report: https://digital.library.unt.edu/ark:/67531/metadc815038/m2/1/high_res_d/98-19_2006Jun08.pdf
dataset in GitHub: https://github.com/fivethirtyeight/data/tree/master/russia-investigation
The raw data behind the story "The Rock Isn't Alone: Lots Of People Are Worried About 'The Big One'" https://fivethirtyeight.com/features/the-rock-isnt-alone-lots-of-people-are-worried-about-the-big-one/.
san_andreas
san_andreas
A data frame with 1013 rows representing respondents and 11 variables:
In general, how worried are you about earthquakes?
How worried are you about the "Big One," a massive, catastrophic earthquake?
Do you think the "Big One" will occur in your lifetime?
Have you ever experienced an earthquake?
Have you or anyone in your household taken any precautions for an earthquake (packed an earthquake survival kit, prepared an evacuation plan, etc.)?
How familiar are you with the San Andreas Fault line?
How familiar are you with the Yellowstone Supervolcano?
Age
Gender
How much total combined money did all members of your HOUSEHOLD earn last year?
US Region
See https://github.com/fivethirtyeight/data/tree/master/san-andreas
The raw data behind the story "The (Very) Long Tail Of Hurricane Recovery" https://projects.fivethirtyeight.com/sandy-311/
sandy_311
sandy_311
A data frame with 1783 rows representing dates and 25 variables:
Date
No description provided.
The number of emergency hotline (311) calls made to the Administration for Children's Services related to Hurricane Sandy on the given date
The number of emergency hotline (311) calls made to Building Protection Systems, Inc related to Hurricane Sandy
The number of emergency hotline (311) calls made to the Community Affairs Unit related to Hurricane Sandy
The number of emergency hotline (311) calls made to the City Hall related to Hurricane Sandy
The number of emergency hotline (311) calls made to the Department of Environmental Protection related to Hurricane Sandy
The number of emergency hotline (311) calls made to the Department of Buildings related to Hurricane Sandy
The number of emergency hotline (311) calls made to the Department of Education related to Hurricane Sandy
The number of emergency hotline (311) calls made to the Department of Finance related to Hurricane Sandy
The number of emergency hotline (311) calls made to the Department of Health and Mental Hygiene related to Hurricane Sandy
The number of emergency hotline (311) calls made to the Department of Parks and Recreation related to Hurricane Sandy
The number of emergency hotline (311) calls made to the Federal Emergency Management Agency related to Hurricane Sandy
The number of emergency hotline (311) calls made to the Department of Housing Preservation and Development related to Hurricane Sandy
The number of emergency hotline (311) calls made to the Human Resources Administration related to Hurricane Sandy
The number of emergency hotline (311) calls made to the Mayor's Fund to Advance NYC related to Hurricane Sandy
The number of emergency hotline (311) calls made to the Mayor's Office of Special Enforcement related to Hurricane Sandy
The number of emergency hotline (311) calls made to Emergency Management related to Hurricane Sandy
The number of emergency hotline (311) calls made to the New York City Housing Authority related to Hurricane Sandy
The number of emergency hotline (311) calls made to NYC Service related to Hurricane Sandy
The number of emergency hotline (311) calls made to the New York Police Department related to Hurricane Sandy
The number of emergency hotline (311) calls made to the NYC Department of Labor related to Hurricane Sandy
The number of emergency hotline (311) calls made to Small Business Services related to Hurricane Sandy
The number of emergency hotline (311) calls made to NYS Emergency Management related to Hurricane Sandy
The total number of emergency hotline (311) calls made related to Hurricane Sandy
Data from NYC Open Data https://data.cityofnewyork.us/City-Government/311-Call-Center-Inquiry/tdd6-3ysr, Agency acronyms from the Data Dictionary. See also https://github.com/fivethirtyeight/data/tree/master/sandy-311-calls
# To convert data frame to tidy data (long) format, run: library(dplyr) library(tidyr) sandy_311_tidy <- sandy_311 %>% pivot_longer(-c("date", "total"), names_to = "agency", values_to = "num_calls") %>% arrange(date) %>% select(date, agency, num_calls, total) %>% rename(total_calls = total) %>% mutate(agency = as.factor(agency))
# To convert data frame to tidy data (long) format, run: library(dplyr) library(tidyr) sandy_311_tidy <- sandy_311 %>% pivot_longer(-c("date", "total"), names_to = "agency", values_to = "num_calls") %>% arrange(date) %>% select(date, agency, num_calls, total) %>% rename(total_calls = total) %>% mutate(agency = as.factor(agency))
This file contains links to the data behind FiveThirtyEight's 'Senate forecasts' https://projects.fivethirtyeight.com/2018-midterm-election-forecast/senate/
senate_national_forecast
senate_national_forecast
A dataframe with 450 rows representing national-level results of the classic, lite, and deluxe Senate forecasts since Aug. 1, 2018 and 11 variables
date of the forecast
the party of the forecast
the model of the forecast
the probability of the corresponding party winning
the mean of the number of seats
the median number of seats
the top 10 percentile of number of seats
the top 90 percentile of number of seats
unknown
the margin of p10_seats
the margin of p90_seats
The original dataset included a meaningless column called "state", and all variables under this column was "US". So this column was removed.
FiveThirtyEight’s House, Senate And Governor Models Methodology: https://fivethirtyeight.com/methodology/how-fivethirtyeights-house-and-senate-models-work/
The raw data behind the story "Early Senate Polls Have Plenty to Tell Us About November" https://fivethirtyeight.com/features/early-senate-polls-have-plenty-to-tell-us-about-november/.
senate_polls
senate_polls
A data frame with 107 rows representing a poll and 4 variables:
Year
Final poll margin
Early presidential approval rating
Early poll margin
See https://github.com/fivethirtyeight/data/tree/master/early-senate-polls
This file contains links to the data behind FiveThirtyEight's 'Senate forecasts' https://projects.fivethirtyeight.com/2018-midterm-election-forecast/senate/
senate_seat_forecast
senate_seat_forecast
A dataframe with 28353 rows representing seat-level results of the classic, lite, and deluxe Senate forecasts since Aug. 1, 2018 and 12 variables
date of the forecast
state of the forecast
class of the forecast
unknown
name of the candidate
party of the candidate
whether the candidate is incumbent
the model of the forecast
the probability of the corresponding party winning
the voteshare of the corresponding party
the top 10 percentile of the voteshare
the top 00 percentile of the voteshare
FiveThirtyEight’s House, Senate And Governor Models Methodology: https://fivethirtyeight.com/methodology/how-fivethirtyeights-house-and-senate-models-work/
The raw data behind the stories "Club Soccer Predictions" https://projects.fivethirtyeight.com/soccer-predictions/ and "Global Club Soccer Rankings" https://projects.fivethirtyeight.com/soccer-predictions/global-club-rankings/.
spi_global_rankings
spi_global_rankings
A data frame with 453 rows representing soccer rankings and 7 variables:
The name of the soccer club.
The name of the league to which the club belongs.
A club's current global ranking.
A club's previous global ranking
Offensive rating for a given team (the higher the value the stronger the team's offense).
Defensive rating for a given team (the lower the value the stronger the team's defense).
A club's SPI score.
See https://github.com/fivethirtyeight/data/blob/master/soccer-spi/README.md
State name, abbreviation, US Census designated division & region.
state_info
state_info
A data frame with 51 rows representing airlines and 4 variables:
State name
State abbreviation
US Census designated division. Values for division
are nested within region
US Census designated region
US Census Bureau https://en.wikipedia.org/wiki/List_of_regions_of_the_United_States#Interstate_regions.
library(dplyr) # Number of states in each division state_info %>% count(division) # Number of states in each region state_info %>% count(region)
library(dplyr) # Number of states in each division state_info %>% count(division) # Number of states in each region state_info %>% count(region)
The raw data behind the story "What America’s Governors Are Talking About" https://fivethirtyeight.com/features/what-americas-governors-are-talking-about/
state_index state_words
state_index state_words
2 data frames about the 50 U.S Governors' Speeches
An object of class spec_tbl_df
(inherits from tbl_df
, tbl
, data.frame
) with 2223 rows and 9 columns.
A data frame with 50 rows representing the 50 U.S. states and 5 variables:
the state
the name of the state's governor
the party of the state's governor
the filename of the speech in the speeches folder at https://github.com/rudeboybert/fivethirtyeight/tree/master/data-raw/state-of-the-state/speeches
a link to an official/media source for the speech
A data frame with 2,223 rows representing phrases and 9 variables:
one-, two-, and three-word phrases spoken repeatedly
thematic categories for the phrases
number of Democratic speeches containing the phrase
number of Republican speeches containing the phrase
total number of speeches containing the phrase
percent of the 23 Democratic speeches containing the phrase
percent of the 27 Republican speeches containing the phrase
the chi-square test statistic for statistical significance
p-value for chi^2 test
The chi-square test statistic https://scikit-learn.org/stable/modules/generated/sklearn.feature_selection.chi2.html#sklearn.feature_selection.chi2
The raw data behind the story "How Americans Like Their Steak" https://fivethirtyeight.com/features/how-americans-like-their-steak/.
steak_survey
steak_survey
A data frame with 550 rows representing respondents and 15 variables:
Respondent ID
not sure
Is respondent a smoker?
Is respondent a drinker?
Is respondent a gambler?
Is respondent a skydiver?
not sure
not sure
not sure
Preferred steak preparation
Is respondent female?
Age
Household income
Education level
Region of US
See https://fivethirtyeight.com/features/how-americans-like-their-steak/
The raw data behind the story "A Complete Catalog Of Every Time Someone Cursed Or Bled Out In A Quentin Tarantino Movie" https://fivethirtyeight.com/features/complete-catalog-curses-deaths-quentin-tarantino-films/. An analysis using this data was contributed by Olivia Barrows, Jojo Miller, and Jayla Nakayama as a package vignette at https://fivethirtyeightdata.github.io/fivethirtyeightdata/articles/tarantino_swears.html.
tarantino
tarantino
A data frame with 1894 rows representing curse/death instances and 4 variables:
Film title
Whether the event was a profane word (TRUE) or a death (FALSE)
The specific profane word, if the event was a word
The number of minutes into the film the event occurred
See https://github.com/fivethirtyeight/data/tree/master/tarantino
The raw data behind the story "Why Some Tennis Matches Take Forever" https://fivethirtyeight.com/features/why-some-tennis-matches-take-forever/.
tennis_events_time
tennis_events_time
A data frame with 205 rows representing tournaments and 5 variables:
Name of event
Court surface used at the event
Seconds added per point for this event on this surface in years shown, from regression model controlling for players, year and other factors
Start year for data used from this tournament in regression
End year for data used from this tournament in regression
See https://github.com/fivethirtyeight/data/tree/master/tennis-time
tennis_players_time
and tennis_serve_time
The raw data behind the story "Why Some Tennis Matches Take Forever" https://fivethirtyeight.com/features/why-some-tennis-matches-take-forever/.
tennis_players_time
tennis_players_time
A data frame with 218 rows representing players and 2 variables:
Player Name
Weighted average of seconds added per point as loser and winner of matches, 1991-2015, from regression model controlling for tournament, surface, year and other factors
See https://github.com/fivethirtyeight/data/tree/master/tennis-time
tennis_events_time
and tennis_serve_time
The raw data behind the story "Why Some Tennis Matches Take Forever" https://fivethirtyeight.com/features/why-some-tennis-matches-take-forever/.
tennis_serve_time
tennis_serve_time
A data frame with 120 rows representing serves and 7 variables:
Name of player serving at 2015 French Open
Time in seconds between end of marked point and next serve, timed by stopwatch app
Opponent, receiving serve
Score in the current game during the timed interval between points
Set number, out of five
Score in games within the set
Date
See https://github.com/fivethirtyeight/data/tree/master/tennis-time
tennis_events_time
and tennis_players_time
The raw data behind the story "For A Trump Nominee, Neil Gorsuch’s Record Is Surprisingly Moderate On Immigration" https://fivethirtyeight.com/features/for-a-trump-nominee-neil-gorsuchs-record-is-surprisingly-moderate-on-immigration/.
tenth_circuit
tenth_circuit
A data frame with 954 rows representing cases and 13 variables:
Name of the case
Date of decision
Case citation, as listed in the Federal Reporter Series
Case citation, Westlaw format
Issue number, in cases divided into multiple issues
Weight per issue (total weight per case equals one)
Name of first judge
Name of second judge
Name of third judge
Vote of first judge. 1 = liberal, 0 = conservative.
Vote of second judge. 1 = liberal, 0 = conservative.
Vote of third judge. 1 = liberal, 0 = conservative.
Category of case, immigration or discrimination
In immigration cases, partial relief to immigration petitioner is coded as liberal because the petitioner typically seeks just one core remedy (e.g., withholding of removal, adjustment of status, or asylum); in discrimination cases, partial relief is coded as multiple issues because the plaintiff often seeks separate remedies under multiple claims (e.g., disparate treatment, retaliation, etc.) and different sources of law.
See https://github.com/fivethirtyeight/data/tree/master/tenth-circuit
The raw data behind the story: "How Popular is Donald Trump" https://projects.fivethirtyeight.com/trump-approval-ratings/: Approval Poll Dataset
trump_approval_poll
trump_approval_poll
A data frame with 3051 rows representing individual polls and 20 variables:
The subgroup the poll falls into as defined by the type of people being polled (all polls, voters, adults)
The date the polling began
The date the polling concluded
The polling group which produced the poll
The grade for President Trump that the respondents' approval ratings correspond to
The sample size of the poll
The type of people being polled (a for adults, lv for likely voters, rv for registered voters)
The weight fivethirtyeight gives the poll when determining approval ratings based on historical accuracy of the pollster
The percentage of respondents who approve of the president
The percentage of respondents who disapprove of the president
The percentage of respondents who approve of the president adjusted for systematic tendencies of the polling firm
The percentage of respondents who approve of the president adjusted for systematic tendencies of the polling firm
True if there are multiple versions of the poll, False if there are not
TRUE if the poll was tracked, FALSE if not
Poll result URL
Poll ID number
ID number for the question being polled
Date the poll was created
Date and time the poll was compiled
Variables "model_date", "influence", and "president" were deleted because each observation contained the same value for these variables: January 5, 2018; 0; and Donald Trump respectively.
https://projects.fivethirtyeight.com/trump-approval-data/approval_polllist.csv and https://projects.fivethirtyeight.com/trump-approval-data/approval_topline.csv
The raw data behind the story: "How Popular is Donald Trump" https://projects.fivethirtyeight.com/trump-approval-ratings/: Approval Trend Dataset.
trump_approval_trend
trump_approval_trend
A data frame with 1044 rows representing poll trends and 11 variables:
The subgroup the poll falls into as defined by the type of people being polled (all polls, voters, adults)
The date the model was created
Estimated approval ratings
Higher bound of the estimated approval percentage
Lower bound of the estimated approval percentage
Estimated disapproval percentage
Higher bound of the estimated disapproval percentage
Lower bound of the estimated disapproval percentage
Date and time the model was compiled
The Variable "president" was removed because all values were "Donald Trump"
https://projects.fivethirtyeight.com/trump-approval-data/approval_topline.csv
This folder contains the data behind the stories: 'What Trump's Legal Battles Tell Us About Presidential Power' https://fivethirtyeight.com/features/what-trumps-legal-battles-tell-us-about-presidential-power/; 'Why It Might Be Impossible To Overturn A Presidential Pardon' https://fivethirtyeight.com/features/why-it-might-be-impossible-to-overturn-a-presidential-pardon/; 'Will The Supreme Court Fast-Track Cases Involving Trump?' https://fivethirtyeight.com/features/will-the-supreme-court-fast-track-cases-involving-trump/; 'Why One of Trump’s Biggest Legal Threats Is New York’s Attorney General' https://fivethirtyeight.com/features/why-one-of-trumps-biggest-legal-threats-is-new-yorks-attorney-general/; 'Should Judges Pay Attention To Trump’s Tweets?' https://fivethirtyeight.com/features/should-judges-pay-attention-to-trumps-tweets/; 'Trump Is Losing The Legal Fight Against Sanctuary Cities, But It May Still Pay Off Politically' https://fivethirtyeight.com/features/trump-is-losing-the-legal-fight-against-sanctuary-cities-but-it-may-still-pay-off-politically/; 'Will Trump’s Latest Lawsuits Keep Congress From Investigating Future Presidents?' https://fivethirtyeight.com/features/will-trumps-latest-lawsuits-keep-congress-from-investigating-future-presidents/;
trump_lawsuits
trump_lawsuits
A dataset with 57 rows representing lawsuits and 16 variables
Current docket number
Date lawsuit was originally filed
Case name (current)
Names of plaintiffs (if more than five, "et al" for plaintiffs who are not in case name)
Names of defendants (if more than five, "et al" for defendants who are not in case name)
Court the lawsuit is currently in front of
Other courts the case has appeared before
Where the case is being heard | 1 = Federal; 2 = State
Names of the judges the case is currently before
PACER code for nature of lawsuit (Not relevant for criminal cases) https://pacer.uscourts.gov/help/faqs/what-nature-suit-code
Whether the case is related to action before Trump was president, his personal conduct as president, or a policy action as president | 1 = Case directed at pre-presidency action; 2 = Case directed at personal action of Trump as president; 3 = Case directed at policy action of Trump as president
The capacity in which Trump is implicated | 1 = Case directed at Trump personally; 2 = Case directed at action of Trump administration; 3 = Trump as plaintiff; 4 = Trump administration as plaintiff; 5 = Case directed at Trump associate; 6 = Other
Criminal vs. civil | 1 = Criminal; 2 = Civil
Key topic area raised in the case (i.e. emoluments, First Amendment, DACA, etc). Categories created based on key policy topic area or legal issue. Calls are subjective and based on reporting and may change.
Original docket number, if case has been appealed or changed jurisdiction
Whether the case, or the part of the case connected to Trump, is ongoing. | 1 = Case is ongoing; 2 = Case or part of case connected to Trump is closed
Approval Polls
The raw data behind the story "How Trump Hacked The Media" https://fivethirtyeight.com/features/how-donald-trump-hacked-the-media/.
trump_news
trump_news
A data frame with 286 rows representing lead stories and 3 variables:
Date of lead story about Donald Trump.
Story classification
Memeorandum https://www.memeorandum.com/.
The raw data behind the story "The World's Favorite Donald Trump Tweets" https://fivethirtyeight.com/features/the-worlds-favorite-donald-trump-tweets/. Tweets posted on twitter by Donald Trump (@realDonaldTrump). An analysis using this data was contributed by Adam Spannbauer as a package vignette at https://fivethirtyeightdata.github.io/fivethirtyeightdata/articles/trump_twitter.html.
trump_twitter
trump_twitter
A data frame with 448 rows representing tweets and 3 variables:
Twitter https://twitter.com/realdonaldtrump
The raw data behind the story "What the World Thinks of Trump" https://fivethirtyeight.com/features/what-the-world-thinks-of-trump/: Trump World Issues Dataset
trumpworld_issues
trumpworld_issues
A data frame with 185 rows representing countries and 6 variables:
The country whose population is being polled
The difference in the number of respondents from the given country who approve and who disapprove of the issue (Trump proposal) in question (approve-disapprove)
The number of respondents from the given country who approve of the issue (Trump proposal)
The number of respondents who disapprove of the issue
undefined
The specific trump policy proposal being posed. Specifically: 1: Withdraw support for international climate change agreements 2: Build a wall on the border between the U. S. and Mexico 3: Withdraw U.S. support from the Iran nuclear weapons agreement 4: Withdraw U.S. support for major trade agreements 5: Introduce tighter restrictions on those entering the U.S. from some majority-Muslim countries
Pew Research Center https://www.pewresearch.org/fact-tank/2017/07/17/9-charts-on-how-the-world-sees-trump/
The raw data behind the story "What the World Thinks of Trump" https://fivethirtyeight.com/features/what-the-world-thinks-of-trump/: Trump World Polls Dataset.
trumpworld_polls
trumpworld_polls
A data frame with 32 rows representing years and 40 variables:
Year the poll was conducted
The average percentage people who answered the poll question positively (support the president or have a favorable view of the U.S.)
The percentage of people from Canada who answered the poll question positively
The percentage of people from France who answered the poll question positively
The percentage of people from Germany who answered the poll question positively
The percentage of people from Greece who answered the poll question positively
The percentage of people from Hungary who answered the poll question positively
The percentage of people from Italy who answered the poll question positively
The percentage of people from Netherlands who answered the poll question positively
The percentage of people from Poland who answered the poll question positively
The percentage of people from Spain who answered the poll question positively
The percentage of people from Sweden who answered the poll question positively
The percentage of people from the U.K. who answered the poll question positively
The percentage of people from Russia who answered the poll question positively
The percentage of people from Australia who answered the poll question positively
The percentage of people from India who answered the poll question positively
The percentage of people from Indonesia who answered the poll question positively
The percentage of people from Japan who answered the poll question positively
The percentage of people from the Philippines who answered the poll question positively
The percentage of people from South Korea who answered the poll question positively
The percentage of people from Vietnam who answered the poll question positively
The percentage of people from Israel who answered the poll question positively
The percentage of people from Jordan who answered the poll question positively
The percentage of people from Lebanon who answered the poll question positively
The percentage of people from Tunisia who answered the poll question positively
The percentage of people from Turkey who answered the poll question positively
The percentage of people from Ghana who answered the poll question positively
The percentage of people from Kenya who answered the poll question positively
The percentage of people from Nigeria who answered the poll question positively
The percentage of people from Senegal who answered the poll question positively
The percentage of people from South Africa who answered the poll question positively
The percentage of people from Tanzania who answered the poll question positively
The percentage of people from Argentina who answered the poll question positively
The percentage of people from Brazil who answered the poll question positively
The percentage of people from Chile who answered the poll question positively
The percentage of people from Colombia who answered the poll question positively
The percentage of people from Mexico who answered the poll question positively
The percentage of people from Peru who answered the poll question positively
The percentage of people from Venezuela who answered the poll question positively
The item being polled. Specifically, whether respondents: 1) Have a favorable view of the U.S. or 2) Trust the U.S. President when it comes to foreign affairs
Pew Research Center https://www.pewresearch.org/fact-tank/2017/07/17/9-charts-on-how-the-world-sees-trump/
# To convert data frame to tidy data (long) format, run: library(dplyr) library(tidyr) trumpworld_polls_tidy <- trumpworld_polls %>% pivot_longer(-c("year", "avg", "question"), names_to = "country", values_to = "percent_positive")
# To convert data frame to tidy data (long) format, run: library(dplyr) library(tidyr) trumpworld_polls_tidy <- trumpworld_polls %>% pivot_longer(-c("year", "avg", "question"), names_to = "country", values_to = "percent_positive")
The raw data behind the story "The Media Really Started Paying Attention to Puerto Rico When Trump Did" https://fivethirtyeight.com/features/the-media-really-started-paying-attention-to-puerto-rico-when-trump-did/: TV Hurricanes Data.
tv_hurricanes
tv_hurricanes
A data frame with 37 rows representing dates and 5 variables:
Date
The percent of sentences in TV news that mention Hurricane Harvey on the given date
The percent of sentences in TV news that mention Hurricane Irma
The percent of sentences in TV news that mention Hurricane Maria
The percent of sentences in TV news that mention Hurricane Irma
Internet TV News Archive https://archive.org/details/tv and Television Explorer
mediacloud_hurricanes
, mediacloud_states
,
mediacloud_online_news
, mediacloud_trump
,
tv_hurricanes_by_network
, tv_states
, google_trends
The raw data behind the story "The Media Really Started Paying Attention to Puerto Rico When Trump Did" https://fivethirtyeight.com/features/the-media-really-started-paying-attention-to-puerto-rico-when-trump-did/: TV Hurricanes by Network Data.
tv_hurricanes_by_network
tv_hurricanes_by_network
A data frame with 84 rows representing dates and 6 variables:
Date
The hurricane in question
The percent of sentences on the BBC News TV channel on the given date that mention the hurricane in question
The percent of sentences on CNN News that mention the hurricane in question
The percent of sentences on Fox News that mention the hurricane in question
The percent of sentences on MSNBC News that mention the hurricane in question
Internet TV News Archive https://archive.org/details/tv and Television Explorer
mediacloud_hurricanes
, mediacloud_states
, mediacloud_online_news
,
mediacloud_trump
, tv_hurricanes
, tv_states
, google_trends
The raw data behind the story "The Media Really Started Paying Attention to Puerto Rico When Trump Did" https://fivethirtyeight.com/features/the-media-really-started-paying-attention-to-puerto-rico-when-trump-did/: TV States Data.
tv_states
tv_states
A data frame with 52 rows representing dates and 4 variables:
Date
The percent of sentences in TV News on the given day that mention Florida
The percent of sentences in TV News on the given day that mention Texas
The percent of sentences in TV News on the given day that mention Puerto Rico
Internet TV News Archive https://archive.org/details/tv and Television Explorer
mediacloud_hurricanes
, mediacloud_states
, mediacloud_online_news
,
mediacloud_trump
, tv_hurricanes
, tv_hurricanes_by_network
, google_trends
The raw data behind: "Mayweather Is Defined By The Zero Next To His Name" https://fivethirtyeight.com/features/mayweather-is-defined-by-the-zero-next-to-his-name/
undefeated
undefeated
A data frame with 2125 rows representing boxing matches and 4 variables:
Name of boxer
URL with the boxer's record
Date of the match
Number of cumulative wins for the boxer including the match at the specified date
Box Rec
The raw data behind the story "The Most Common Unisex Names In America: Is Yours One Of Them?" https://fivethirtyeight.com/features/there-are-922-unisex-names-in-america-is-yours-one-of-them/.
unisex_names
unisex_names
A data frame with 919 rows representing names and 5 variables:
First names from the Social Security Administration
Total number of living Americans with the name
Percentage of people with the name who are male
Percentage of people with the name who are female
Gap between male_share and female_share
Social Security Administration https://www.ssa.gov/oact/babynames/limits.html. See https://github.com/fivethirtyeight/data/tree/master/unisex-names.
The raw data behind the story "Some People Are Too Superstitious To Have A Baby On Friday The 13th" https://fivethirtyeight.com/features/some-people-are-too-superstitious-to-have-a-baby-on-friday-the-13th/.
US_births_1994_2003
US_births_1994_2003
A data frame with 3652 rows representing dates and 6 variables:
Year
Month
Day
POSIX date
Abbreviation of day of week
Number of births
Centers for Disease Control and Prevention's National Center for Health Statistics
The raw data behind the story "Some People Are Too Superstitious To Have A Baby On Friday The 13th" https://fivethirtyeight.com/features/some-people-are-too-superstitious-to-have-a-baby-on-friday-the-13th/.
US_births_2000_2014
US_births_2000_2014
A data frame with 5479 rows representing dates and 6 variables:
Year
Month
Day
POSIX date
Abbreviation of day of week
Number of births
Social Security Administration
The raw data behind the story "Where People Go To Check The Weather" https://fivethirtyeight.com/features/weather-forecast-news-app-habits/.
weather_check
weather_check
A data frame with 928 rows representing respondents and 9 variables:
Respondent ID
Do you typically check a daily weather report?
How do you typically check the weather?
If they responded "A specific website or app" when asked how they typically check the weather, they were asked to write-in the app or website they used.
If you had a smartwatch (like the soon to be released Apple Watch), how likely or unlikely would you be to check the weather on that device?
Age
Gender
How much total combined money did all members of your HOUSEHOLD earn last year?
US Region
The source of the data is a Survey Monkey Audience poll commissioned by FiveThirtyEight and conducted from April 6 to April 10, 2015. See https://github.com/fivethirtyeight/data/tree/master/weather-check
The raw data behind the story "2019 Women’s World Cup Predictions" https://projects.fivethirtyeight.com/2019-womens-world-cup-predictions/
wwc_2019_forecasts wwc_2019_matches
wwc_2019_forecasts wwc_2019_matches
2 dataframes about the 2019 Women's World Cup matches and teams
An object of class spec_tbl_df
(inherits from tbl_df
, tbl
, data.frame
) with 52 rows and 18 columns.
A data frame with 192 rows representing 2019 Women's World Cup team match-by-match projections, and 21 variables:
Date match was played
Team
Assigned group for the group stage
Soccer power index
SPI offensive rating
SPI defensive rating
Simulated number of wins
Simulated number of ties
Simulated number of losses
Simulated difference between goals_scored and goals_against
The number of goals that a team is expected to score against an average team on a neutral field
The number of goals that a team is expected to concede against an average team on a neutral field
Chance of winning group stage game 1
Chance of winning group stage game 2
Chance of winning group stage game 3
Chance of winning group stage game 4
Chance of playing in the round of 16
Chance of playing in the quarter-finals
Chance of playing in the semi-finals
Chance of playing in the finals
Chance of winning the tournament
2019 Women's World Cup Predictions A data frame with 52 rows representing Women's World Cup matches, and 18 variables:
Date match was played
Team 1
Team 2
Soccer power index of team 1
Soccer power index of team 2
Probability that team 1 will win match
Probability that team 2 will win match
Probability that the teams will tie the match
Projected number of goals scored by team 1
Projected number of goals scored by team 2
Actual number of goals scored by team 1
Actual number of goals scored by team 2
Shot-based expected goals for team 1
Shot-based expected goals for team 2
Non-shot expected goals for team 1
Non-shot expected goals for team 2
Goals scored by team 1 accounting for the conditions under which each goal was scored
Goals scored by team 2 accounting for the conditions under which each goal was scored
https://projects.fivethirtyeight.com/soccer-api/international/2019/wwc_forecasts.csv
https://projects.fivethirtyeight.com/soccer-api/international/2019/wwc_matches.csv