Election night 2024, hour by hour: how fast Polymarket priced Trump’s win, state by state, and what it means for tonight

Two years ago tonight's question was the presidency. On election night 2024, Polymarket's market on the winner was the most traded prediction market in history at the time, with more than $3.6 billion in volume across its outcomes. We rebuilt that night hour by hour from Polymarket's public price history to see how fast the market worked out the result, and what that suggests for the midterms that close their polls this evening.

The hourly data are in a CSV file.

Key findings

1. The market moved long before the call. Trump's price passed 90% at about 11 p.m. ET and 95% by midnight. The Associated Press called the race at 5:34 a.m. ET the next morning.
2. The first big move came after 9 p.m. Between 6 and 9 p.m. ET the price wandered between 58% and 71% while early states reported. It jumped as North Carolina and Georgia firmed up.
3. States settled in a clear order. North Carolina passed 95% first, at about 10 p.m. ET; Georgia an hour later; Pennsylvania and Arizona around midnight; Wisconsin at 1 a.m.; Michigan and Nevada around 2 a.m.
4. The market's pre-election favourite won, and a Republican sweep of the presidency, Senate and House, which Polymarket's balance-of-power market priced at 36.5% before the polls closed, also came true.

The night in prices

The day itself was calm: Trump traded around 60–62% from noon to 6 p.m. ET, in line with where the market had been for days, and very little information reached traders until the polls began to close. The first polls closed at 6 p.m., and in the next hour his price actually dipped to 58% as the first returns came in.

Then the market turned. By 8 p.m. Trump was at 71%. The decisive stretch was 9 p.m. to midnight: 71% → 83% → 92% → 97%. By 2 a.m. the price was above 99%. The AP's call came hours later, at 5:34 a.m. ET.

Harris's side of the night

The Harris contract mirrored Trump's, as it must in a two-way race where the two prices add up to roughly $1. It peaked at about 42% in the hour to 7 p.m. ET, as the first returns came in, then fell to 30% by 8 p.m., 20% by 10 p.m., 7% by 11 p.m. and about 3% by midnight. Anyone holding Harris at 6 p.m., at about 38%, had seen most of their stake disappear within five hours of the polls closing.

Why the market moved before the votes were counted

At 11 p.m. ET in 2024, with Trump at 92%, most swing states were far from finished counting. How could the market be so sure?

• Traders compare results with expectations, county by county. If a candidate is beating their 2020 margin in the counties that have reported, and those counties are representative, the final result can be estimated long before the count is complete.
• Early states are informative about later ones. A shift in North Carolina and Georgia among particular groups of voters is likely to show up in Pennsylvania, Michigan and Wisconsin too. Traders priced the later states off the earlier ones.
• Prices aggregate many analysts. Election-night models from media outlets and independent analysts, plus traders' own spreadsheets, all feed into the same price. No single source has to be right.

The risk in all of this is the counting order. In states that report one kind of vote first, early leads can be misleading. In 2024 that risk ran briefly against Trump at 7 p.m.

The swing states

Source: Polymarket price history and event volume for each state's winner market. Times are Eastern, at hourly resolution.

Three patterns stand out:

• Eastern states first. North Carolina and Georgia close their polls at 7:00–7:30 p.m. ET and count quickly. They were the first states the market treated as decided, and their results set the tone for the national price.
• Underdogs flipped late. Wisconsin and Michigan both started the night with Trump as the underdog, at 46% and 37%. The market moved him to favourite as the count went on, crossing 95% at 1 a.m. and 2 a.m.
• Slow counters stay open. Arizona was priced at 95% by midnight but did not pass 99% within the next day: the state counts mailed ballots for days, and some uncertainty remained until its count was well advanced.

What the market got right, and what it did not

The market's favourite won the presidency and all seven swing states went the same way. But a 60% favourite going into the night is not a lock, and the market was not certain: at 7 p.m. ET it briefly priced Trump at 58%. Anyone who sold Trump in that hour, reading the earliest returns as good news for Harris, lost money within two hours.

Polymarket's "balance of power" market had also priced the outcome. Before the polls closed it gave 36.5% to a Republican sweep of the presidency, Senate and House, well ahead of a Democratic president with a Republican Senate and Democratic House at 24.5% and a Democratic sweep at 15.5%. The Republican sweep happened.

What a 60% favourite looks like

Going into the night, Trump was a 60% favourite. That is not a strong favourite: it means the market expected him to lose about four times in ten. When he won, it did not prove the market "right" any more than a Harris win would have proved it wrong.

What the night does show is how quickly a 60–40 question turns into a 97–3 question once votes are counted. The uncertainty before the polls close is mostly about who will turn out and how they will vote. After the first few hours of results, most of that uncertainty is gone, even though the count is far from finished. Prices reflect that: they move fastest in the first three to five hours after the polls close and very little after that.

Lessons for tonight

The 2026 midterms are different: there is no single national race, and the headline markets are the Senate and House control markets. But the 2024 night suggests what to expect.

1. Expect the first hour to mislead. Early returns in 2024 briefly pushed the market the wrong way. In 2026 the first states to report Senate races, Georgia and South Carolina at 7 p.m. ET, count early votes quickly; read their first numbers with care.
2. Expect the decisive move between 8 p.m. and midnight ET. That is when most competitive states finish closing and count the bulk of their votes.
3. Expect the markets to be far ahead of the calls. In 2024 the price passed 95% more than five hours before the AP call. Tonight, a race priced at 95% at 10 p.m. may not be called until the morning.
4. Expect slow states to stay open. Arizona and Nevada count mail ballots for days. Alaska, in 2026, is likely to do the same.

2024 and 2026 compared

Sources: Polymarket; our eve-of-election snapshot for 2026.

The structure of tonight's markets means the decisive moves may be spread across more markets and more hours. A Senate majority can hinge on two or three races in different time zones, and a seat-count market will keep moving long after control is clear.

FAQ

When did the markets "call" the 2024 election? Polymarket's price for Trump passed 90% at about 11 p.m. ET and 95% by midnight on election night. The AP called the race at 5:34 a.m. ET.

Which swing state was decided first? North Carolina, where Trump's price passed 95% at about 10 p.m. ET.

Did the market ever favour Harris on the night? No. Her price peaked at about 42% in the first hour after the polls began to close.

Where can I get the data? In our CSV file, with hourly prices from noon on 5 November to 7 a.m. on 6 November 2024.

Method

• Source: Polymarket's public price history for the "Presidential Election Winner 2024" market (the Donald Trump and Kamala Harris outcomes) and each swing state's winner market, at hourly resolution.
• Prices are the last recorded price in each hour. At hourly resolution, a "first hour at 95%" can be up to an hour later than the first trade at that level.
• Times are converted to Eastern Standard Time (UTC−5), which was in force on 5 November 2024.
• Volume is Polymarket's reported total volume for each market. The 2024 presidential event's total of more than $3.6 billion includes all candidate outcomes, among them many that were never competitive.
• The AP call time is as reported by the Associated Press.

Download the data

The CSV file contains the hourly price of Trump and Harris to win the presidency from 12 p.m. ET on 5 November to 7 a.m. ET on 6 November 2024, with Trump's price in each of the seven swing states.

For tonight, see our election-night guide and the eve-of-election snapshot of every 2026 race.

Data: Polymarket public price history. Not investment advice.