Biggest Political Betting Upsets: What Punters Won

One UK punter backed Donald Trump to win the US presidency at 150/1 – three days after Trump entered the Republican race in June 2015. If they staked £100, they walked away with £15,000. Brexit paid out at 3/1 on the day of the referendum, when the entire market was pricing a 76% chance of Remain winning. Jeremy Corbyn won the Labour leadership in 2015 from odds of 100/1. These aren’t flukes. Each one was a market that got the probability badly wrong – and the punters who sensed it first made the kind of returns that sports betting almost never delivers.

Here’s how it happened, what the markets missed, and what it means for anyone watching the odds on Angela Rayner, Keir Starmer and Reform UK right now.

Brexit: £100 Million Wagered, 76% Chance of Remain – Wrong

The Brexit referendum in June 2016 became the single biggest political betting event in British history at the time. Betfair alone took £45.8 million in wagers. Total betting across all UK operators exceeded £100 million. That eclipsed the £40 million placed on the 2012 US presidential race and the £35 million on the 2014 World Cup final.

And the market was almost unanimously wrong.

Right up to polling day, Betfair priced Remain at roughly 76% – odds of around 1/4. Ladbrokes and Paddy Power were similar. The “Leave” vote was a 3/1 outsider on the morning of June 23. The polls were close. The bookmakers were not. By early morning on June 24, Leave had won 52% to 48%.

The punters who backed Leave at 3/1 collected. Ladbrokes alone paid out £1.2 million on Leave backers. The market had anchored itself to the assumption that Britain wouldn’t do something so dramatic – and kept anchoring even as polling data shifted toward Leave in the final weeks.

Columbia University statistics professor Andrew Gelman later identified what went wrong: the market became self-reinforcing. Pollsters looked at the Betfair odds showing Remain ahead and discounted their own polls. The bookmakers looked at the polling consensus and kept their Remain price short. Each group treated the other’s view as independent confirmation of something that was actually circular. When the result came in, both were caught simultaneously.

📊 Key Stat: 3/1 – the odds available on a Leave vote on the morning of the Brexit referendum, June 23 2016, implying roughly a 25% probability. Leave won 52% to 48%. (Source: Ladbrokes, June 2016)

“Some very tidy profits would have been made by the streetwise punters who sensed the true public sentiment – right up through the night, exit polls were indicating a defeat for Leave.” – Sporting Life political betting guide

---

The political market that looks most mispriced right now is where the biggest returns sit. The table below covers the platforms with the deepest UK political markets in 2026:

---

Trump at 150/1: The Bet That Made History

Donald Trump entered the 2016 Republican primary race on June 16, 2015. Three days later, one UK punter backed him to become the 45th President of the United States at odds of 150/1. By election night in November 2016, Trump had crossed 270 electoral votes and won.

The maths on that bet: £100 at 150/1 returned £15,000.

Trump wasn’t quite that price on election night itself – the market had updated as he won the Republican nomination and the polls tightened. By the final weeks, he was trading between 3/1 and 5/1 on most platforms, with Betfair Exchange showing Hillary Clinton at around 80% implied probability. Still – those who backed Trump at any point during his campaign at long odds collected heavily.

The miss followed the same pattern as Brexit. The professional betting market – dominated by educated, urban, globally connected participants – was systematically underestimating a political movement it didn’t fully understand. The actual voters who were going to put Trump in the White House were not the same people pricing the markets. That gap between who bets and who votes is the structural inefficiency that punters who called both Brexit and Trump were exploiting.

Corbyn at 100/1: The Shot That Changed British Politics

Jeremy Corbyn entered the 2015 Labour leadership contest as what the gambling industry calls a “token candidate” – someone who enters to represent a wing of the party in a contest they have no realistic chance of winning. Multiple bookmakers opened him at 100/1.

He won.

His victory at those prices would have paid £100 into £10,000. By August 2015 – weeks before the result – the odds had crashed as polling of Labour members showed him leading. The punters who spotted the membership enthusiasm before the market priced it in cleaned up.

The Corbyn upset introduced a lesson that’s directly relevant to today’s markets. Labour leadership contests are decided by the membership and trade unions – not by MPs, not by media opinion, and not by the kind of voters who participate in political betting markets. The gap between what Westminster insiders think will happen and what Labour members actually want can be enormous. That gap created 100/1 Corbyn. Understanding that same gap today is part of why Angela Rayner’s market movement – from 7/1 to 9/4 in a matter of weeks – looks fast but not necessarily wrong.

The 2017 Snap Election: Labour From 100/1 to 11/4

Theresa May called a snap general election in April 2017, expecting to increase her majority and settle Brexit negotiations from a position of strength. She was the heavy favourite. Labour under Jeremy Corbyn were trading at around 100/1 to win most seats when the election was called.

By election night, Labour had gained 30 seats. The Conservatives lost their majority. Labour’s odds had crashed from 100/1 to 11/4 within weeks of campaigning starting as polling showed an extraordinary surge.

The punters who saw it coming – who understood the enthusiasm gap between Labour’s grassroots campaign and Conservative complacency – made extraordinary returns on a bet that looked suicidal the week it was placed.

The Uxbridge By-Election: Labour’s Safe Seat That Wasn’t

More recently, in July 2023, Labour’s Danny Beales was expected to walk into Boris Johnson’s old Uxbridge and South Ruislip seat after Johnson resigned in disgrace. The Conservatives had won every election in the constituency since 1970. Labour held a big national lead in the polls. Beales was a strong favourite.

The Conservatives held the seat – primarily because Mayor Sadiq Khan had recently extended the Ultra Low Emission Zone into the constituency, making car ownership more expensive for local residents. An issue that had nothing to do with national politics flipped a by-election that the market hadn’t priced.

It’s the clearest recent example of local factors overriding national polling in a specific contest – and it’s the kind of thing that the May 7 local elections could deliver at scale across 134 councils simultaneously.

What These Upsets Have in Common

Every major political betting shock shares one structural feature: the market was pricing the safe, comfortable, establishment assumption while a different reality was building in the electorate.

Brexit punters who won weren’t necessarily better at statistics than those who lost. They were closer to – or more willing to believe – the ground-level sentiment in Leave-voting communities that national polls and urban betting markets were systematically undercounting.

The lesson for 2026 is direct. Reform UK is currently priced at 15/8 for the general election and 3/10 for the local elections. Those prices reflect the dominant market consensus. The question any serious bettor asks is: what’s the market assuming that might be wrong? Is it overestimating tactical voting coordination against Reform? Is it underestimating Restore Britain’s vote-split effect on Labour? Is it anchoring too heavily on the last 18 months of polling while missing a trend that’s already shifted?

The biggest political betting upsets weren’t predicted by people who had more information. They were called by people who questioned the consensus a few weeks earlier than everyone else.