The United States’ selects its 47th president as final votes are cast on Tuesday, November 5th.
As of Tuesday nearing 7 p.m. ET, odds to win the presidential election for Donald Trump are -150, or 56.90%, while Kamala Harris is +120, or 43.10%. These odds are via BetMGM UK.
Below, you will find my breakdown of the race as well as my prediction and how I'm betting Trump vs. Harris.
If you're in the U.S., you can bet the election at a social sportsbook like Rebet. Learn more about sweepstakes betting here (hint: it's very similar to real betting). They're offering election odds, with Trump favored at -170 as of Tuesday night.
Presidential Election Analysis
The most common way to evaluate presidential elections is through election forecasting. Poll-based forecasts include:
Each of these forecasts have this race as a toss-up, with no model favoring either candidate by more than 53%. In fact, an average of these four models puts the election at almost exactly 50/50 odds.
However, election polling comes with plenty of error. Famous examples include both 2016 and 2020 where polls underestimated Donald Trump in both elections. Trump went on to win in 2016, while Joe Biden took home a closer victory than polls suggested in 2020.
That said, polling errors can underestimate Democrats as well. Barack Obama in 2012 is the most recent example in a U.S. presidential election. But even two years ago the 2022 midterm elections saw a polling error underestimate Democrats in the first post-Dobbs election.
So in addition to poll-based models, I’ve taken a deep dive into each key swing state’s early voting data, and looked at other telling factors that in recent presidential elections have uncovered where the polling errors may lie.
Poll-Based Models & Early Voting Data
District-level polling was a key indicator in both 2016 and 2020 that Democrats may not fare as well as state or national polling suggested. Also, high quality polling of Iowa by Ann Selzer showed Donald Trump likely to do better in Iowa than other polls suggested.
Well, both district-level polling and Saturday’s Selzer poll of Iowa show Democrat overperformance compared to most national and state-level polling.
The Washington jungle primary has been a strong indicator of the national environment in recent years, and a recent analysis by Split Ticket gives key indications that Democrats may overperform.
Other special elections in 2023-24 have shown a swing of around 6-7 points toward Democrats compared to each district’s partisan lean over a 69-election sample.
The candidate that has led in net favorable rating has won 16 of the last 17 elections. The only time that didn’t happen? In 2016, when Donald Trump won. However, Hillary Clinton still won the popular vote and Clinton herself was quite unpopular.
And going beyond poll-based modeling, a recent machine learning model developed by the Brown Political Review uses over 100 variables to forecast the election, including polls. Their model has Harris at about a 70% chance to win the presidency.
Swing States Will Once Again Be Key
When diving into each state’s early vote, I believe Harris is in a strong position in Michigan. Wisconsin looks somewhat in toss-up territory, but I also lean Harris in Pennsylvania.
Nevada, Georgia, and North Carolina early voting results all look stronger to me than polling forecasts suggest.
And Arizona could be the closest state, which still comes in above poll-based forecasts expectations for Harris.
Presidential Election Prediction
I will point out, I’ve taken any political bias out of this as I always do when betting. Whether I’m betting NFL, college basketball, NASCAR, or anything else, I’m not letting favoritism creep into my bets. In 2016 I made money backing Trump, while in 2020 I profited especially on the live betting front when I saw early returns favor Biden.
In all, I can’t ignore the abundance of non-polling evidence pointing to a Harris win, especially as the majority of these exact same indicators pointed toward Trump overperformance relative to the polls in 2016 and 2020.
That’s just not the case here, and in a toss-up scenario based off the polls, I’m going to side with the other key indicators.
There’s even some evidence of a major shift states with higher-than-average populations of older white women and college-educated voters. States like Iowa, Kansas, Ohio, parts of Pennsylvania and New Hampshire all fit this bill and have had polls just in the last week showing a Harris overperformance. If so, there’s potential for a large polling error.
That said, I won’t predict a huge blowout.
But I will predict a Harris win with at least 282 electoral votes.
She does this by winning the swing states of MI, PA, GA, and NV. I also think WI and NC are leans toward Harris, while AZ is a pure toss-up. There’s a non-zero chance of a wave in which she starts to grab some of TX, FL, OH, or IA based off high-quality regional and district polling.
I also predict a 3% or higher national popular vote margin, which is a little less than halfway between Clinton’s +2.1% and Biden’s +4.5% margin.
My best election bet at each major regulated betting site is as follows:
- Democrats winning party (+140 at Bet365 Canada)
- Harris to win (13/10 at BetMGM UK)
- Harris to win electoral college and popular vote (44% implied at Kalshi)
Election Odds in Other Key States
- Arizona: Trump entered Election Day as about a 2-1 favorite, but is that price too steep? Early voting data and district level polling is suggesting statewide and national polling is more likely underestimating Harris in this cycle.
- Florida: Trump is a huge favorite in the Sunshine State, and a Harris win would be the result of a major polling miss nationally.
- Georgia: There's a lot to still be decided in Georgia, a red state shifting blue over the last eight years, with turnout in both red and blue counties already up from 2020 in early voting.
- Iowa: Trump dropped from a massive favorite to a slightly less massive favorite over the weekend thanks to some prominent pollsters releasing their data. There's still some betting value on the board, but you may have to lay a steep price.
- Michigan: Joe Biden took Michigan in 2020 and Harris seems poised for another victory here as a slight favorite entering the day.
- Nevada: Mail-in ballots in Clark County (where Las Vegas is located) will play a huge role in NV — is it enough to give the edge to Harris, who is a small underdog entering Tuesday?
- New Hampshire: Harris is a big favorite here, but in the event of a broad polling miss, there's some betting value on a prop here.
- North Carolina: Trump is favored in the Tar Heel State, but should he be? This race will likely come down to which side goes to the independent voters.
- Ohio: Trump is a massive favorite in Ohio, which has gone red the last two elections after going blue for Barack Obama the previous two.
- Pennsylvania: The most important state in the election, PA is close to 50/50 entering Election Day.
- Texas: Trump is a huge favorite in Texas and only a big polling miss would create any changes there.
- Wisconsin: Joe Biden won Wisconsin in 2020 by just more than 20,000 votes, and WI is setting up for another razor-thin race in 2024.
