Activity Density
Activity density describes how much market activity occurs per unit of time—for example, the number of trades, order-book changes, or price/probability updates per second (or per minute) in a prediction market.
It’s a useful concept because “nothing is happening” can mean either (a) the market is stable, or (b) the market is inactive and the displayed probability is becoming stale.
What counts as “activity”
Activity density depends on what your feed considers an event. Common choices include:
- Trades (executions)
- Order book updates (adds, cancels, partial fills)
- Best bid/offer changes (quote updates)
- Price / implied probability updates
Always specify which event types you’re counting. Two systems can report very different densities for the same market if one counts every order-book mutation and the other only counts last-trade or probability ticks.
Why it matters
Activity density impacts both how you interpret a market and how you engineer your data pipeline:
- Freshness / staleness risk: low density increases the odds that the latest probability is old information.
- Signal reliability: thin markets can look “stable” simply because updates are rare.
- Burst handling: high density often arrives in spikes around news; consumers need buffering and backpressure.
- Capacity planning: sustained high density informs WebSocket sizing, queue throughput, and database write rates.
How to measure activity density
You’ll usually compute activity density over a rolling window.
Common metrics:
- Events per second (EPS):
count(events) / window_seconds - Trades per minute (TPM):
count(trades) / window_minutes - Updates per market:
count(events) / (window_seconds × number_of_markets)(useful for dashboards)
Operationally, many teams track p95 / p99 density (not just the average) to plan for bursts.
Example
If a market produces 240 probability updates in the last 60 seconds, its update-based activity density is:
240 / 60 = 4updates per second
If those 240 updates included 12 trades, then trade-based activity density is:
12 / 60 = 0.2trades per second (12 trades per minute)
Both numbers can be “true” at the same time—they answer different questions.
Activity density and FinFeedAPI
When you consume prediction market data programmatically, activity density helps you decide:
- whether a market’s probability stream is updating frequently enough for alerts and monitoring
- whether you should aggregate (e.g., per-second bars) to reduce noise and write volume
- how to size consumers to avoid lag during high-activity bursts
FinFeedAPI’s Prediction Markets API is designed for programmatic access to live and historical prediction-market probabilities and updates—so you can compute activity density consistently across markets and time windows.