The math behind 500,000 binary markets per day.
Five hundred thousand sounds like a marketing number. It is not. It is arithmetic, with two factors and a small footnote. One hundred thousand commercial flights cross the global network every day. Each one, structured as a settleable contract, produces five binary outcomes. The total is a multiplication.
The multiplication
Global commercial flight volume runs around 102,000 daily passenger flights according to OAG, and Flightradar24 puts the figure closer to 130,000 once cargo and private aviation are included. We use 100,000 as a clean baseline. It is conservative.
Every flight has the same five settleable questions: was it delayed by more than 15 minutes? More than 30? More than 60? Was it cancelled? Was it diverted? Each answer is a yes/no contract — a binary market. Five questions × 100,000 flights = 500,000 binary contracts per day.
That is roughly 18,000 markets per hour, or 5 markets per second, every second, forever. The category never sleeps.
Why exactly five tiers
The temptation is to pick a clean number — three tiers, ten tiers — and let aesthetics decide. We did not. Five is what the underlying industry already uses.
15 minutes is the FAA's on-time threshold. A flight that arrives within fifteen minutes of schedule is reported as on-time. This is the most important threshold in aviation operations and the one most travelers care about, even if they do not know the number.
30 minutes is the threshold most travel insurance policies use to trigger delay coverage. It is also the operational threshold most airline disruption-management systems treat as the line between a routine late arrival and an event that requires customer accommodation.
60 minutes is the threshold that typically triggers refund and rebooking policies, and is the line at which delay enters the territory of severe disruption.
Cancelled and diverted are tail-risk outcomes that price differently from delays. A cancelled flight has a fundamentally different downstream economic impact (rebooking, hotel, cascading delays for crew) than a delayed flight, even a severely delayed one. A diverted flight is rarer still — typically weather, medical, or mechanical, and almost always priced at a discount because diversions resolve before passengers reach their destination.
Five tiers is what the airline, insurance, and traveler ecosystems already think in. Aviax did not invent these thresholds; we encoded them.
Why aviation, not sports or elections
The category we want has three properties: continuous flow, observable settlement, and a settlement criterion that is unambiguous and machine-readable.
Sports has continuous flow during seasons and observable settlement (referee decisions, scoreboard, ESPN). But the observation surface is brittle in subtle ways — overtime, replays, fouls that swing outcomes, post-game corrections — and the volume is fixed by the league schedule. The NFL produces about 270 regular-season games per year. The NBA produces around 1,200. Combined major leagues globally, low tens of thousands of events per year.
Elections have observable settlement and unambiguous resolution. They do not have continuous flow. A US presidential election year produces a few dozen contracts of meaningful liquidity. Off years produce far fewer. The category is fundamentally episodic.
Aviation has all three properties. The flow is continuous (no off-season, no weekends). The settlement is observable (aircraft signal is the source of truth, independent of any single party). The criterion is unambiguous (did the flight cross a specific threshold or not). One hundred thousand events per day for as long as commercial aviation exists.
Volume implications at platform liquidity
The headline number tells you the supply. The interesting math is what happens when you multiply supply by per-contract liquidity to get total notional book depth.
At low per-contract liquidity ($100 per market) — the kind a sportsbook would price a niche event category — 500,000 markets implies $50M per day in notional book depth. That is comparable to a mid-sized US sportsbook's handle on a single category.
At Polymarket-class current per-contract liquidity (around $1,000), 500,000 markets implies $500M per day. That is in the territory of a large day on Polymarket today, but instead of one good day every two weeks, it is every day.
At mature per-contract liquidity (around $10,000 — the level Polymarket reaches on its biggest contracts), 500,000 markets implies $5B per day. At that scale, the category is a meaningful fraction of total prediction-market volume globally.
The bottleneck is not demand at any of these levels. The bottleneck is the oracle. Without a settlement-grade feed of flight outcomes, none of this volume is achievable, even if every other piece of the stack is in place.
Why this number is achievable now
Aircraft signal infrastructure reached settlement-grade reliability in roughly the last 18 months. Coverage of commercial airspace via independent receiver networks is dense enough that signal observations are continuous and verifiable from multiple sources. The historical archive is deep enough — eight years — to train detection models on every realistic delay pattern.
On the platform side, Polymarket's matching engine and contract factory have matured to the point where minting half a million contracts per day is not a stretch. Kalshi's CFTC-regulated structure is in production and has been processing event contracts at scale.
The arithmetic was always there. The infrastructure was not. Both sides are now ready.