Ferry capacity: what sold seats actually tell us

Seats reported as sold are not a count of people who actually boarded. The sales fields in this archive are final values visible when the API response was retrieved, not a regular set of booking snapshots taken 24 hours before departure.

Can a cancelled sailing still have bookings?

Yes. Of 21,772 cancelled records, 5,497 (25.2%) have a positive passenger-sales value. Zero in the remaining rows does not prove cancellation because of low demand. Stopped sales, refunds or later record changes could also affect the final value; this source cannot distinguish their timing.

When is a load ratio meaningful?

A basic reported sales ratio divides sold passenger capacity by total passenger capacity. Its denominator must be positive, matching capacity types must be compared and anomalies must be flagged separately. Averaging individual sailing ratios differs from dividing all sales by all capacity. Quietly labelling either measure “route occupancy” can mislead readers.

Measurement limits in this archive

Completed records include 56 passenger-capacity exceedances and 1,456 vehicle-capacity exceedances. We do not present those inconsistencies as proof of actual overloading. The median passenger-sales value among completed records is 272; it is not a median count of boarded passengers. Adding vehicle spaces and passenger seats also does not create a meaningful occupancy measure.

What would a popularity ranking need?

It would need a comparable period, understood timetable coverage, known sales-snapshot timing and actual passengers carried. More archive rows do not necessarily mean a more popular route. Our route pages therefore show observed sailings and cancellations by direction, without ranking demand or popularity.

Sources and data

İDO: official sailing and ticket information · MGM: Marmara marine forecast

Data methodology and scope