Average will be taken from RCP 2-way average.
Trump took the lead on September 11th, and has held it since.
This market resolves to the month in which RCP is updated to show Biden in the lead, or else "Not before Election Day". Resolves based on Pacific Time if there is ambiguity. This is the date that RCP is updated, not the date the polling took place.
You can find more questions like this on the polling dashboard.
Related questions
Looks like April is a no barring anything wild happening. https://www.cnn.com/2024/04/28/politics/cnn-poll-trump-biden-matchup/index.html?Date=20240428&Profile=CNNPolitics&utm_content=1714306087&utm_medium=social&utm_source=twitter
Quinnipiac added (tie), back to Trump 0.3.
https://www.realclearpolling.com/polls/president/general/2024/trump-vs-biden
Morning Consult poll added, average now Trump +0.2.
https://www.realclearpolling.com/polls/president/general/2024/trump-vs-biden
@TimothyJohnson5c16 Not sure, seems like RCP wants to keep Trump ahead. They might drop off the last 4 at a time.
@FoxKHTML Can't get much closer than Florida in 2000. Imagine the nightmare if it actually does come down to that few votes.
@PlasmaBallin I mean I think it will be very close in several states. GA, AZ, and MI most likely
@FoxKHTML I agree that all the states will be close, just not as close as Florida in 2000. Even if your expectation value in some states is one candidate winning by a few hundred votes, the confidence interval on that should be so large that the chance of it actually being that close is very small.