apps2 · betting — prediction-market pulse

Combined 24-hour betting volume — the leading-indicator pulse

…

(a) Total volume over time — Kalshi vs Polymarket vs combinedUse Show to pick Individual (Kalshi vs Polymarket, fills the panel) or Total (combined only) — not both, so the small lines never get squashed. Y-axis auto-fits tightly to whatever is shown (no empty band). SRAD / GENI / DKNG ride a separate normalized scale (each its own line) — compare each stock's shape (did it follow the flow?), not its level.

Show Window View (bars)400 cumulative (lifetime) Compare Scale bold basket

(b) Open interest over time — capital committed (notional basis)Use Show to pick Individual (Polymarket vs Kalshi OI, fills the panel) or Total (combined only) — not both. Y-axis auto-fits tightly to whatever is shown (no empty band). Dashed grey = order-book depth (≠ OI). SRAD / GENI / DKNG ride a separate normalized scale — each its own individually-selectable line (chips below).

Show Window View (bars)400 Compare Scale bold basket

Open interest compared on a USD-notional basis. Polymarket OI is true dollar notional; Kalshi OI is contracts × $1 max payout (a notional ceiling, not mark-to-market). Order-book 'liquidity' is book depth, shown separately — not open interest.

(c) 24h-volume momentum — accelerating or fading flowy-axis = Δ combined 24h volume vs the prior reading (USD). Above zero = speeding up; below = slowing down.

Window View (bars)400 Compare Scale bold basket

(d) Sports vs non-sports mix — where the betting dollars sit'share' view: y = sports % of combined 24h volume (0–100%). 'stacked' view: y = USD, sports stacked under non-sports.

Window View (bars)400 View

(d2) Kalshi category heat — WHICH category is driving the flowrows = Kalshi categories (sorted by latest 24h volume) · cell = percentile of that reading vs the category's OWN banked history (ONE shared ramp, endpoints labeled on the chart) · ramping — recorder live since ~2026-06-30, percentiles firm up as history deepens. Click → synced dateline; hover any cell for the exact reading.

Window View (bars)400 Metric

Top markets — heaviest action right now (live snapshot)

Same 10 markets, second view (the do-both rule): x = 24h volume, y = open interest (both log-$, ONE shared scale so the dashed vol=OI diagonal is the x=y line) · bubble SIZE = vol+OI footprint · FILL = venue · OUTLINE = category (same hue as the category-heatmap rows). Above the diagonal = conviction parked; below = day-trade churn. Live cross-section only — per-market history is only now being banked (time views arrive as it accrues).

(g) Lead-lag — does rising exchange flow LEAD each stock (SRAD / GENI / DKNG)?Per-venue flow (Kalshi/Polymarket, vol/OI) vs EACH equity separately — +lag = flow leads the stock by that many trading days · −lag = the stock leads the flow. Flow is bucketed to one value per day and stepped onto each stock's trading-day calendar, then A2.leadLag correlates the daily changes at every lag. ramping — a statistically firm read needs ~63 overlapping trading days of banked flow (≈ 2026-09-29); until then bars sit inside the wide noise band (honest, not broken).

Flow (leader) Lookback Max lag

Sportradar (SRAD)

Genius Sports (GENI)

DraftKings (DKNG)

Read-through — the actionable signal for SRAD / GENI / DKNG

Vol + OI lead the flowPrediction-market volume + open interest lead the flow for the listed enablers.

Rising = tailwindRising exchange vol/OI = a tailwind for SRAD & GENI (sports data feeds, B2B revenue scales with handle) and DKNG (sportsbook take).

Each enabler its own lineOn charts (a)/(b)/(c) each enabler is overlaid as its own individually-selectable line — toggle SRAD, GENI and DKNG separately to see whether each equity actually followed the betting flow.

Divergence = the signalPer-stock divergence (vol up, but one name flat) is the actionable signal: the market hasn't yet priced a surge in betting activity into that enabler — and you can now see which one is leading vs lagging instead of an averaged blur.

Lead-lag (panel g)The lead-lag profiles quantify it: for a chosen per-venue flow, panel (g) computes how many DAYS that flow leads each stock, at what correlation, and over how many observations — so “flow leads the equity” is a measured number per name, not a hunch. Ramping toward a firm read as banked history deepens.