Sentiment vs Market
For each tracked market we gather recent discussion across Reddit, news and the web, score every item for what it implies about the market's specific YES/NO resolution (not its tone), and aggregate that into our own implied probability. We then compare it to Polymarket. A gap is a potential edge — but we only call it a signal when the divergence is large and the sample is big enough. Informational only, not financial advice.
Federal Reserve interest rate decision
Low confidenceWill the Fed cut interest rates at its next FOMC meeting?
US inflation cooling
Low confidenceWill US year-over-year CPI inflation come in below 3% this year?
US recession in 2026
Low confidenceWill the US enter a recession in 2026?
How the number is built
Collect mentions in the trailing 30 days → score each for relevance, stance (toward YES or NO) and confidence → weight by relevance × confidence × recency × reach → map the weighted index to a probability. The mapping is an honest placeholder until we have enough resolved outcomes to calibrate it against. Small samples are always flagged low-confidence and never surfaced as a strong signal.
See the live odds these compare against on the Fed decision dashboard and the recession tracker, or size a position on the decision tools.