Democrats are favored by bets in relevant individual Senate races in both platforms. For example, in Texas, James Talarico has the upper hand on both platforms 74% to 76% odds of winning the seat over Republican Ken Paxton. Troy Jackson also enjoys a lead of 73% against Maine Republican Susan Collins on both platforms, while in Michigan, Abdul El-Sayed is favored by 78% on both Polymarket and Kalshi at the moment of writing.
This time, traditional pollsters are on the side of prediction markets, with CNN’s data analyst Harry Enten describing the potential outcome as a “polling earthquake.”
“Just within the last few hours, we had the equivalent of a polling earthquake, And to keep up with the weather metaphors… it tends to point to a blue wave,” said Enten on Sunday.
Prediction markets entered the political sphere with strength after accurately predicting President Trump’s victory in 2024 when some traditional pollsters gave an edge to Democrat candidate Kamala Harris. The midterm election is the biggest relevant test for these platforms since then, with Kalshi CEO Tarek Mansour calling these “the first prediction-market election,” with over $1.4 billion traded in midterm-linked bets.
Nonetheless, since 2024, the accuracy of these platforms has been spotty at best, with prediction markets failing to track the upset victory of President Luiz Inácio Lula da Silva after giving Senator Flavio Bolsonaro the lead before the runoff election.
The influence of prediction markets on voting events is under scrutiny for the first time, with institutions like the Center for Democracy and Technology stressing that they can distort candidate decisions, impact voter behavior, and undermine confidence in election processes and outcomes.