01
Resolve the market
The supplied Polymarket slug or URL is normalized and matched against live Polymarket Gamma data. Unknown markets fail before payment settlement.
Methodology
DisputeTracker separates observed market and oracle facts from inferred lifecycle stages and timing estimates. The API does not use an opaque model to predict an outcome.
01
The supplied Polymarket slug or URL is normalized and matched against live Polymarket Gamma data. Unknown markets fail before payment settlement.
02
CLOB metrics, Polygon RPC events, UMA oracle requests, disputes, and voting rounds are joined when available. Every missing dependency remains visible.
03
Ordered rules map observed lifecycle evidence to a named stage, progress state, next event, confidence, completeness, and freshness label.
Market identity, question, rules, lifecycle flags, update time, volume, and liquidity.
Tradable outcome prices and spread when current token books are available.
UMA requests, proposals, disputes, chain timestamps, and linked oracle activity.
Voting round phase and final outcome evidence when the public data source exposes it.
Stage confidence describes how strongly observed facts support the selected lifecycle stage. Data completeness describes how much of the expected record is present. Missing fields are returned as machine-readable identifiers instead of being silently filled.
Exact on-chain timestamps and Gamma-derived estimates remain separately labelled. An estimated duration is never represented as an exact deadline.
The score is deterministic and clamped to 0–100:
round((1 - stageConfidence) × 40)
+ round((1 - completeness) × 35)
+ min(15, missingFields × 5)
+ min(10, disputeEvents × 3)0–33 is low, 34–66 is medium, and 67–100 is high. It is a data-quality and process-risk signal, not an outcome probability or financial advice.
Operation schemas, prices, request examples, and receipt headers are published in OpenAPI. The production case study shows the same pipeline with public Tempo settlement evidence.