Election Prediction: Big Data, Algorithms, and Baseball
With less than 50 days left until the Nov. 8 election, the stage is set with local, state and national media producing a blizzard of extensive campaign coverage.
In the print media, there is a uniformity in the way election stories are structured. Most often note how the race fits into the “big picture”; how much money the contestants raised; and what the polls say.
During this century, another element has become the norm with the convergence of “Big Data” and algorithmic modeling: voter assessments and forecasts from analysts exploring historical statistics, analyzing voter data and aggregating polls to make predictive judgments based on an ever-expanding field of factors. and relationships.
Until the late 1970s, predicting the outcome of an election was a parlor game for academics, political scientists, and pundits like Ouija boards, tea leaves, and the “Washington Ruler.” which said if the city’s NFL team won its last home. game before Election Day, the president’s party would win. The ‘rule’ picked every presidential winner between 1940 and 2000, but since then has only been right once.
Models for the “modern era of election forecasting” were established in 1978 by a Yale University professor’s model based on economics and tenure, and in 1979 by a Yale University professor’s model. the University of Kentucky establishing the relationships between presidential approval ratings and subsequent votes.
In the late 1990s, the American Political Science Association’s magazine PS: Political Science & Politics and the International Journal of Forecasting annually introduced new election forecasting models developed by universities, analysts, consultants, non-profit organizations, media and the gambling industry, based on “econometrics”, public opinion, public service, party unity, scandals, poll aggregation and models of historical votes, to provide “fluid intelligence” to voters, media, experts, candidates and campaigns.
But it wasn’t until a sportscaster used the same tools developed to project the performance of baseball players that election predictions gained ambient credibility.
In late 2007, Nate Silver, a consultant at KPMG Chicago and co-author of “Baseball Prospectus,” began writing statistical analyzes of election data for Daily Kos. Using this “baseball” model, he correctly predicted presidential outcomes in 49 states in 2008 and all 50 in 2012.
Silver’s FiveThirtyEight site is among dozens online that provide voter assessments, such as Cook Political Report, and race projections, such as Larry Sabato’s Crystal Ball, Inside Elections, Electoral-Vote, Politico, RealClearPolitics and 270 to Win. Here are four oft-cited sites that claim to be independent, non-partisan, and do not issue endorsements:
— Five Thirty Eight: Founded in 2008 as a poll aggregator with an interpretation blog by Silver, the name reflects the number of voters in the US Electoral College. In 2010, The New York Times published the site. ESPN acquired it in 2013, moving it to another Disney network, ABC News, in 2018.
Since 2014, FiveThirtyEight has focused on poll analysis, politics, economics, science, popular culture and sports blogging. It ranks pollsters, weighing the accuracy of their polls.
After correctly predicting the results in 49 states in the 2008 presidential election and in every state in 2012, like most forecasters, FiveThirtyEight got it wrong in the 2016 election, assessing the chances of victory for the former President Donald Trump at 28%. In 2020, he correctly predicted results in 48 states, but overestimated President Joe Biden’s margins of victory in Wisconsin, Michigan and Pennsylvania.
As of Sept. 21, FiveThirtyEight gives Democrats a 71% chance of taking control of the Senate, winning up to four seats, and Republicans an equal 71% chance of securing a majority in the House, totaling up to 33 seats. in the November elections.
— Cook’s political relationship with Amy Walter: Founded by analyst Charlie Cook in 1984, it is an “online newsletter” analyzing electorates and campaigns for presidential, House, Senate and gubernatorial elections since 2004.
In 2021 it changed its name to ‘The Cook Political Report with Amy Walter’ after Walter took over as editor, publisher and owner.
The site is known for its Cook Partisan Voting Index (PVI), which rates all 435 congressional districts with a scale of seven categories: Solid Democratic, Likely Democratic, Lean Democratic, Toss-Up, Lean Republican, Likely Republican, and Solid Republican.
PVI correctly predicted that Republicans would take control of the Senate in 2014, but incorrectly projected that Hillary Clinton would win in 2016. In 2020, he predicted Biden would win with 290 electoral votes, underestimating his margin of victory.
In 2022, out of 35 Senate races, Cook notes four “tossups” and six competitive. Nine of the 14 Senate seats held by Democrats and 16 of 21 held by Republicans are considered “sure” or “probable” wins for the incumbent’s party.
Cook’s “tossup” Senate seats are held by Democrats from Arizona, Georgia and Nevada, and Republican Sen. Ron Johnson (R-Wis.). Democratic incumbents from Colorado and New Hampshire are running in “Lean Democratic” races. Republicans are defending an open seat they won in 2016 in Pennsylvania, which Cook calls “Lean Democratic.”
Cook classifies 162 congressional precincts as “solid Democrats” and 30 as “Democrats” or “probable.” It ranks 188 congressional districts as “solid Republican” and 24 as “leaning” or “likely” Republican. The GOP will win a House majority, Cook projects, but by what margin the outcome of 31 “tossup” races depends, including in 22 districts now held by Democrats.
— Larry Sabato’s crystal ball: Created in 2002 by Larry Sabato, director of the Center for Politics at the University of Virginia, the site focuses on trends in the presidential, congressional and gubernatorial races with analyzes that include key points for readers .
Senate and House race odds are depicted in maps and charts showing which candidate won the last presidential election there, among other metrics.
In 2008, Crystal Ball predicted winners in 421 of 435 House races, 34 of 35 Senate races, and all 11 gubernatorial races. His projection that Barack Obama would win 364 Electoral College votes was slightly less than the end result. In 2012, he predicted Obama’s re-election, but underestimated his margin of victory and mistakenly said the Senate would remain unchanged – the Democrats won two seats.
In the 2016 election, Crystal Ball predicted that Clinton would win easily with 322 electoral votes, a 50-50 Senate, and Democrats winning 13 House seats to eat away at the GOP majority. This, of course, did not happen. “We ruined everything,” Sabato said at the time.
In 2020, Crystal Ball predicted the Biden-Trump winner in 49 states — missing in North Carolina — and winners in all but five congressional districts. It was the only site to predict Trump would win Florida, and one of the few to say Biden would win Georgia.
In a September “struggling seats” analysis of the 2022 midterms, Crystal Ball argues that Biden’s low approval ratings put Democrats in the worst position they’ve been in since 2010. He predicts that the GOP will take control of both houses in November, gaining up to 42 seats in the House and at least one seat in the Senate.
— Inside Elections with Nathan L. Gonzales: Founded by Roll Call columnist Stuart Rothenberg as “The Rothenberg Political Report” in 1989, Roll Call and CNN election analyst Nathan L. Gonzales took it over in 2015 and added his name in 2017 Rothenberg remains editor-in-chief.
The site produces newsletters that analyze and handicap House, Senate, gubernatorial, and presidential elections in 24 annual reports that feature interviews with candidates, data, election history and trends, and public and private polls on individual races.
In 2016, Inside Elections also predicted a win for Clinton, predicting she would get at least 332 electoral votes. He also overstated Biden’s victory in 2020.
In 2022, Inside Elections predicts the GOP will win 12 to 30 House seats, the Senate will remain split, and Republicans will win 20 of 36 gubernatorial races.