Catalog coverageLIVE
—
— Kalshi · — Polymarket
01 Prediction-market research system
Find edge.
Prove it. Then risk capital.
A disciplined workspace for finding structural mispricing across Kalshi and Polymarket—and testing directional hypotheses without fooling ourselves.
Data freshnessHOURLY
—
Collected at :23 each hour
Pipeline healthCHECKING
—
Reading collection history
Trading statusLOCKED
OFF
Research gates come first
02 Edge pipeline
Strategies, ranked by proof.
Structural edges first. Directional conviction only earns capital after calibration and executable-price testing.
A
Framework ready
Cross-venue
complete set
Buy complementary outcomes on resolution-equivalent contracts when the all-in cost is below the guaranteed payout.
YESA+NOB+costs<1.00
B
Signal contract
Directional
model edge
Compare a frozen, calibrated forecast with executable asks after uncertainty, fees, and slippage.
ImplementationDefined
C
Research queue
Logical
constraints
Detect prices that violate mutual exclusion, exhaustiveness, or implication across related markets.
ImplementationNext
03 Operating discipline
Every edge passes the same gates.
A good story is not a strategy. Each candidate moves from raw observation to paper execution with its assumptions exposed.
- 01ObserveImmutable source data
- 02MatchResolution semantics
- 03PriceDepth, fees, latency
- 04GateConservative risk
- 05ReplayOut-of-sample proof
CATALOG MIX
Inferred from titlesObserved topics
COLLECTOR
Recent runs
Loading run history…
CATALOG Research universe
Market catalog
Search the collected market metadata. Prices and order-book depth are intentionally not represented here.
MarketTopicClosesVenue
Loading market catalog…
—
LAB Assumption-first tooling
Strategy lab
Fast sanity checks for candidate edges. These are research calculators, not execution signals.
01
STRUCTURAL
Implemented
Complete-set arbitrage
Test the locked payout after both entry prices and a conservative all-in cost buffer.
Conservative profit$2.50
Return on capital2.56%
Edge / contract2.5¢
payout − leg A − leg B − costsPositive before execution risk02
DIRECTIONAL
Contract ready
Model-edge hurdle
Haircut the model probability before comparing it with the executable ask and transaction costs.
Conservative fair value57.0¢
Net edge / contract3.5¢
Hurdle resultPASS
(model − uncertainty) − ask − costsSignal clears the cost hurdleGATES Before a candidate becomes a trade
Edge is necessary.
It is not sufficient.
The engine rejects opportunities when any critical input is stale, unverified, or economically too thin.
✓
Resolution equivalenceHuman-reviewed rules, dates, sources, and exceptional outcomes
✓
Executable depthActual asks walked through the book—not midpoint or display probability
✓
Verified feesCurrent market-specific fee schedule with correct rounding
✓
Fresh synchronized inputsBounded snapshot skew and pessimistic legging latency
✓
Risk capacityEvent, venue, correlated exposure, and capital-lockup limits
DATA Collection observability
Data & runs
Health and lineage for the hourly read-only catalog collector running on this server.
LINEAGE
HourlyCollection path
01Venue APIsPublic market metadata
→
02NormalizeShared market schema
→
03PersistImmutable gzip artifact
→
04SurfaceThis research UI
!
Catalog only This job does not yet capture synchronized order books. Arb detection must use a separate, time-aligned snapshot process.
VENUES
— totalLatest coverage
K
—KalshiRegulated exchange
P
—PolymarketEvent markets
RUN LOG
Recent collections
StatusObservedKalshiPolymarketArtifact
Loading collection history…