π« On-time odds Β· delay risk, hour by hour
Which flight is actually going to show up on time?
Delay risk isn't random β it follows time of day, the month, and how congested each end of the route is. Pick a route and see when to fly, who to fly with, and how risky it really is.
On-time
β
Arrivals within 15 min of schedule, annual average.
Avg delay
β
Mean delay across all departures on this route.
When delayed
β
Typical delay, on the flights that are late.
Route baseline
β
Average on-time across every airline on this route β the route-level figure, separate from the selected airline's score.
Best time of day
05:0011:0017:0023:00
Best airline for this route
How the model works β Illustrative, grounded in real delay behaviour. Base on-time ~92% for an early-morning departure off an uncongested airport; every extra tier of hub congestion (slot-constrained airports like AMS, CDG, FRA, LHR, MAD) shaves points, time of day degrades from ~88% in the morning to the low-60s by late evening (delays cascade through the day's aircraft rotation), and the month layer reflects summer thunderstorms & ATC peaks (JulβAug worst) and winter de-icing/fog. Airlines carry their real-world on-time reputation. Route service is a bounded, illustrative fixture: each selectable pair names only the carriers that publish nonstop service on it (airline route maps / schedules, Sep 2025), with a per-route provenance note shown on the airline panel. Only those curated pairs are offered in the selector β this is not an exhaustive market dataset and does not claim to cover every airport pair; the fixture is the seam where a Eurocontrol or OAG feed would later slot in. Airport codes & coordinates from OurAirports; on-time structure mirrors Eurocontrol network performance and OAG/Cirium punctuality league tables.