The election bets were right – again. The gambling market successful predicted the 2024 presidential race and called it in favor of the president-elect Donald Trump. More than $3 billion was wagered on the race, including more than $2.6 billion on Polymarket alone.
While some have scoffed On this prospect, proponents of the predictive power of the gambling markets were vindicated again on Tuesday when Trump defeated Vice President Kamala Harris and returned to the White House.
It appears that Trump will win a decisive victory in the Electoral College (312-226), far from ‘the nearest presidential election in history” that many experts predicted. The bettors didn’t think the race would be that close as they gave him a 20.8-point betting lead on Election Day.
The gamblers usually got it right
The bettors predicted a Trump victory for 30 consecutive days leading up to the election, giving him a 28.9-point lead a week ago. Two weeks ago, punters predicted he would win all seven swing states. After a number of polls were released in Harris’ favor in Wisconsin and Michigan, as well as a poll showing her in the lead in solid red Iowa, bettors went scaled back their expectations.
As it turned out, scaling back their expectations was the only mistake the gamblers made, as the results mirrored their previous prediction almost identically. Harris took a last-minute lead in the betting markets in Wisconsin and Michigan, but lost both states.
Electoral markets tend to do that in favor of the Democrats in general, but less than the polls or predictive models such as FiveThirtyEight. They tend to be more “surprised” when a Republican candidate wins, which makes it no surprise that Trump outperformed the betting in 2024.
The betting markets not only predicted the presidency, but were also right about the Senate. Bettors predicted that the Republican Party would regain control of the Senate on Election Day with more than 80% certainty. Republicans had been heavily favored to win the Senate since July.
Bettors predicted a 52-48 majority for Republicans in the Senate, and as of Wednesday the GOP has secured 52 seats in the Senate. However, it is more than likely that Republicans will also take at least one of Pennsylvania and Nevada.
According to ElectionBettingOdds.com founder Maxim Lott, this is not surprising. Considering that there were five Senate elections that gave Republican underdogs about a 20% to 30% chance of winning, he told the Washington Examiner that statistically at least one of those underdogs should win.
The majority in the House of Representatives has not yet been determined because it is too close to call, but bettors had that race as a coin flip. In the final week of the election cycle, control of the House of Representatives changed hands numerous times, with neither party gaining a significant lead. Republicans had the best odds of winning the House on Halloween with a 55.6% chance of winning, while Democrats got a 52% chance on Election Day.
How the betting markets changed during the race
Once the dust settled on the 2024 primaries and everyone knew, or thought they knew, who the two major candidates would be, Trump became the slight betting favorite over then-Democratic nominee President Joe Biden.
Trump extended his betting lead by a dominant margin of 48.2 points shortly after he was shot in the ear in July, which followed Biden’s June 27 debate.collapse.”
On July 3, Harris’s chances of becoming the next president surpassed Biden’s, even though he had not yet dropped out of the race. Bettors accurately predicted she would replace Biden, even as pundits and the president himself were adamant he would stay in the race.
Once Harris was announced as the new Democratic nominee, there was an overwhelming amount of positive media coverage characterizing her as a “joyful warrior” narrowed the odds, leading to her exchanging the lead with Trump several times between August and September 2024.
The first major momentum shift in the Trump-Harris matchup occurred after the September 10 presidential debate. Trump’s debate performance reduced his chances, and Harris continued to lead the betting markets for 26 straight days as she mostly avoided interviews with journalists.
Lott, of ElectionBettingOdds.com, told the Washington Examiner that he noticed another shift following the October 1 vice presidential debate between newly elected Vice President J.D. Vance and Governor Tim Walz (D-MN). Days later, on October 6, Trump took a slight lead in the odds. The next day Harris appeared 60 minutes in an interview that was criticized by both sides of the aisle, and the former president saw his chances rise.
After taking the betting odds lead on October 6, Trump extended that lead over the next three weeks until a flurry of polls showing momentum for Harris in Wisconsin, Michigan and Iowa caused his margin to fall from 28.9 points to 9.9 points in a matter of days before correcting back to 20.8 points.
Trump went on to outperform expectations and polls en route to victory in the Electoral College and the popular vote. His outperformance in the polls follows a trend of all Republican presidential candidates, except Mitt Romney, have done so since 2004.
History shows a track record of success
It’s no surprise that the bettors got it right again, as historical records show that the bets are extremely efficient in predicting a candidate’s chance of winning an election. For example, candidates given a 75% chance of winning will win about 75% of the time, the data shows.
Between 2016 and 2022 Election betting followed The chances of 807 different candidates during a number of elections. They found that when gamblers gave a candidate a 90% to 100% chance of winning, that candidate won 99.4% of the time. Candidates with a probability between 80% and 90% won 93.8% of the time.
(Image courtesy of Maxim Lott)
Between 70% and 80%, that percentage drops to 74.4%. Between 60% and 70% it is 68.8%.
In the margin between 50% and 60%, this is 56%. Between 40% and 50% it is 44%. Between 30% and 40% it is 34.9%. Between 20% and 30% it is 23.8%. Between 10% and 20% it is 4.5%, and between 0% and 10% it is 1%.
In total, 88.9% of the candidates favored by the election betting won their election out of the 807 candidates followed. Trump was not only favored, but also given about a 60% chance of beating Harris.
CLICK HERE TO READ MORE FROM THE WASHINGTON EXAMINATOR
However, the data shows that there is an “underdog bias” in races where one candidate is heavily favored over the other. This bias produces a higher predicted probability of upset in races where the outcome is essentially a foregone conclusion.
Only three bets underdogs have won presidential elections since the 1940s, and two of them were narrow underdogs. John F. Kennedy won at +110, Jimmy Carter won at +100, and Trump won in 2016 despite +375 odds. Even in 2016, bettors gave Trump more than a 20% chance of winning, while many of the best predictive models had him in single-digit territory.