Tracking Prediction Market Momentum
A market sitting at 40% tells you where the crowd is. A market that was at 22% last week tells you something is happening.
Level tells you consensus, change tells you news
Any single probability reading is a snapshot of consensus. It answers 'what does the crowd currently believe?' but says nothing about whether that belief is settled or actively being revised.
The change over a period — commonly a week, long enough to filter daily noise and short enough to stay current — answers a different and often more actionable question: what has the crowd changed its mind about? A large move means new information has arrived and been absorbed into real money positions.
Why a seven-day window is a reasonable default
Single-day changes in event contracts are dominated by noise, especially in markets with modest volume. One sizeable order can move a thin market several points without any new information existing.
Windows much longer than a week begin to blur distinct catalysts together. A month-long change on a Fed market may contain a CPI print, a jobs report, and two speeches, which makes attribution impossible. Seven days usually isolates a recognisable narrative.
Reading momentum together with level
The combination is more informative than either alone. A market at 80% that has been at 80% for a month is a settled expectation, and the interesting question is what would break it. A market at 80% that was at 55% last week is a live repricing, and the crowd may not be finished.
Contested markets that are also moving quickly — near 50% with a large weekly shift — are where genuine disagreement and new information overlap. Those are usually the most information-rich contracts on a board.
The noise filter: always check volume alongside the move
A 20-point weekly move on a market with a few thousand dollars of volume is not a signal, it is a liquidity artefact. The same move on a market with millions in volume reflects a lot of capital changing its mind.
This is why momentum should never be read in isolation. Weight every move by the money behind it, and treat large shifts on thin markets with scepticism until volume confirms them.
Momentum as a macro read
Aggregated across related contracts, weekly odds shifts function as a real-time macro sentiment gauge. If rate-cut odds, recession odds, and credit-stress contracts are all moving in a consistent direction within the same week, that coherence is itself evidence — separate markets, priced by different participants, agreeing on a change in the macro picture.
The Radar plots every live market by probability against its 7-day momentum, with bubble size showing the money behind each move.
See where the crowd stands now →Quick answers
Why use a 7-day change rather than a daily change?
Daily changes in event contracts are dominated by noise, particularly in lower-volume markets where a single order can move the price several points. A week filters most of that while staying current.
Does a big move in odds mean the market is now correct?
No. It means new information has been absorbed into prices. Markets can and do overreact, particularly on thinly traded contracts where a small amount of capital moves the price a long way.
How do I tell real repricing from noise?
Check the volume and liquidity behind the move. A large shift on a market with meaningful volume reflects real capital changing its mind; the same shift on a thin market is usually an artefact.