—
Opportunities · 24h
LIVE FEED
Today’s barometercross-venue conditions
—
best edge per set · 24h
Phase 8 experiment · as of — AWAITING SCAN
Live spread radar
Watching 8 cross-listed contract pairs across Kalshi and Polymarket (ForecastEx not active — normalizer is a stub), plus ~30 Kalshi multigame parlays for synthetic mispricings. The engine logs every observed spread; the learner updates match confidence and Kelly fraction nightly.
No capital at risk. This engine only logs observed spreads in Phase 8.0 — the “LIVE FEED” ticker shows watched markets, not executed trades. Numbers stay at zero until pairs clear the risk gate.
—
Last scan
01 State of the system paper bankroll — no capital, Phase 8.0PAPER-V0
Bankroll · PAPER
—
cap rises only after Phase 8.3 PROCEED
Kelly fraction
—
recalibrated on rolling 30d Sharpe
Match patterns watched
—
≥20 obs required before live capital
Sharpe (30d, paper)
—
target > 1.5 for shadow→paper
02 Scan funnel is “0 pairs” the engine working, or a silent format drift?
fetched
—
normalized
—
certified
—
paired
—
tradeable
—
Diagnosis
NO DATA
| venue | fetched | normalized | certified | paired | tradeable | funnel |
|---|---|---|---|---|---|---|
No scan funnel recorded yet
Each scan records fetched→normalized→certified→paired→tradeable per venue. This distinguishes “everything filtered” (engine working) from “nothing matched” (silent format drift).
| ||||||
03 Recent opportunities
| ts | contract | buy | sell | spread | edge_adj | qty | decision |
|---|---|---|---|---|---|---|---|
No opportunities logged yet
Engine in Phase 8.0 — Polymarket and Kalshi books will start producing pairs as the 2026 NBA Finals approach.
| |||||||
04 Match-pattern confidence
| pattern | confidence | obs | status | last updated |
|---|---|---|---|---|
Awaiting first paper trade
Confidence updates as paper trades close. Asymmetric: agreement +5%, divergence −30%.
| ||||
05 Settlement risk per venue pair
| venue pair | risk | obs | status |
|---|---|---|---|
No closed paper trades yet
Risk score = dispute_rate × 0.50 + late_payout_rate × 0.02 over rolling 30d.
| |||
06 Parlay arbitrage observation-only
Parlay obs · 24h
—
Kalshi multigame markets sampled
Actionable mispricings · 24h
—
parlay price ≠ ∏ leg prices > $0.02
Biggest mispricing · 24h
—
edge per contract (Phase 8.0 logs, no trades)
| parlay | legs | bid/ask | fair lo/hi | direction | edge |
|---|---|---|---|---|---|
No parlay scans yet
First scheduled
cross_market_scan.bat run populates this. Math: parlay_yes_price vs ∏ leg_side_prices on the same Kalshi venue. Single-venue, multi-leg arb — synthetic legs not yet hedged in Phase 8.0. | |||||
07 Basket Dutch-book intra-venue · no settlement-divergence risk
Events scanned · 24h
—
Polymarket negRisk events (World Cup, weather)
Actionable baskets · 24h
—
ΣYES_ask < $1 or complete buy-all-NO < (N−1)
Best edge per set · 24h
—
realistic edge at executable depth (post-fee)
| event | side | legs | ΣYES ask | ΣNO ask | edge/set | max sets | note |
|---|---|---|---|---|---|---|---|
No basket scans yet
First scheduled basket scan populates this. Math: buy one YES of every candidate for Σask, collect exactly $1 at resolution. Edge/set = 1 − Σask − fees. Intra-venue, so no cross-venue settlement risk.
| |||||||
08 Today’s miss reasons
| reason | count | turn this knob |
|---|---|---|
No misses to classify yet
Engine logs every rejected opportunity. Distribution of reasons drives the next operator action.
| ||
09 Phase gate
8.0 · Build
no capital
Adapters live · matcher ≥95% manual-spot accuracy
8.1 · Shadow
no capital · 14 d
≥30 sim trades · Sharpe > 1.5 · DD < 8%
8.2 · Paper
no capital · 30 d
Same metrics · realistic fill modeling
8.3 · Live small
$500 · 30 d
Sharpe > 1.0 · zero settlement disputes
8.4 · Live scaled
up to bankroll cap
Continued Sharpe > 1.0 · DD < 10%