Signals

Updated just now

How to read the numbers

Implied / market-implied YES

The probability of YES implied by current traded prices (mid or last). It is what participants are paying for, not a claim about real-world odds.

Model estimate

A rule-based heuristic from the signal engine when a rule sets one, not a black-box forecast. Some signals only describe liquidity or spreads and may show no model estimate.

Edge / gap

The difference between the model estimate and market-implied, in percentage points (model minus market for YES). Filters may use “largest gap.” This is informational only-not trading advice or guaranteed advantage.

Stance (above / below / near estimate)

Compares market-implied to the model estimate when both exist. Labels are not buy or sell recommendations.

Confidence

A simple UI clarity label for signals (not a prediction). It summarizes the signal’s own magnitude/quality metrics into one of: Low, Mid-low, Mid, Mid-high, or High.

Volume

Reported trading activity for the market, for context on size and liquidity.

Change & sparklines

Movement in market-implied YES over the window labeled on the card-often 24h where data allows.

Signals

Rule-based flags from ingested public data. They are not trade recommendations.

More detail in Methodology.

Rule-based flags from public prediction market data across Polymarket and Kalshi-not trade recommendations. Market-implied is live probability; model estimate is the rule’s heuristic when present. Edge is the gap between them (not another name for market-implied). Few cards usually means few active rule hits in your project-try Include wide-spread if books are gated out.

View
Src
Books
Cat
Stance
Type
Gap
Opp
Conf
Edge
Liq
Vol
Spr

Preset views sort or bias the list. Wide-spread includes markets with wider books that are hidden from the default feed.

Why only 0 card(s)? The worker creates signals only when rules fire and rows stay active. Very wide order books are omitted from the default feed-use Wide-spread on (disable to tighten).

No signals in this view. Try including wide-spread or loosen filters.