Research
Essential question: when is the best time to exit a position - lock in the winnings already on the table, or hold it to close? Every strategy on this site so far is entry-only: buy YES, hold to resolution, win or lose the full unit. Tap Out asks the other half of the question - once a position is up, is there a point in the day where taking the guaranteed partial win beats gambling the rest of it on the close?
Buy YES in the morning. If the bucket is winning, its YES price rises through the day - which means the NO price on that same bucket falls by roughly the same amount. Instead of selling the YES position directly, buy the matching NO contract later in the afternoon at that lower price. Holding both YES and NO on the same bucket locks in a fixed payout at resolution regardless of outcome, at a total cost of less than $1 - the gap between the two entry prices, minus fees on both trades, is the profit that gets locked in right then instead of being left to ride on the close.
Two different rules for "when" are on the table, and it's not yet clear which one the data actually supports:
Built from every Polymarket snapshot captured since 2026-05-22, through 2026-11-06.
Research
Essential question: how frequently, and at what time of day, does the top-priced bucket actually change hands? With several buckets in play on a given day, the leader isn't just one favorite losing to one challenger - it can trade among two or three contenders throughout the day. Tracked continuously at hourly checkpoints (7am-11pm local): a checkpoint requires one bucket to be strictly the highest-priced, no ties - a tie means no clear leader that hour, so it's skipped rather than guessed at, and the next valid hour is compared against the last one before the gap. Y-axis is a rate (changes at that hour / city-days with a valid reading that hour), not a raw count, so an hour with denser data coverage doesn't look more volatile just from having more chances to catch a flip. A second chart below breaks down what those changes actually were: how many were the bucket that led at 6am reclaiming the top spot, versus a bucket that hadn't led before taking over for the first time.
Same settled city-days as elsewhere on this page, but using the actual settle indicator: for each settled city-day, the hour where the bucket that goes on to win is first the unique top-priced bucket - checked from midnight, not just 7am on, so an overnight appearance isn't missed. Everything before 6am is pooled into one bar rather than shown hour by hour, since overnight checkpoints are sparse on their own; 6am keeps its own bar since it's already the baseline hour used elsewhere on this page. Frequency, not a rate - a straight count of city-days. The number on top of each bar is the average price the winner was actually trading at the moment it first took the lead. Bars below the 30-count minimum used elsewhere on this site are shown lighter. Built from every Polymarket snapshot captured since 2026-05-22, through 2026-11-06.
Not how often the lead changes hands (the chart below), but how many DIFFERENT buckets ever took a turn at the top - a day where the lead flips back and forth between the same two buckets four times still counts as 2 here, not 4. Same 7am-11pm window and same city-days as the chart below; no settlement requirement, since this is about leader dynamics generally, not tied to who eventually won.
Bars (left axis) are how often the lead changes at all - changes at that hour / city-days with a valid reading that hour. The line (right axis) is a different question with a different denominator: of the changes that happen at that hour, what share are the 6am leader reclaiming the top spot versus a genuinely new bucket taking over. Bars below the 30-observation minimum are shown lighter; the line has no point at 7am, since it's the first checkpoint and there's nothing to take a share of yet.