The rankings only count flights the network watched happen. A flight earns credit because our hub saw it fly in real time — logged in from your phpVMS instance, or not, has no bearing on the score. Reporting flight counts to us through the sync channel doesn't move your ranking either; only flights the hub observed live do.
Every pilot's counted hours are capped within any rolling 48-hour window, the same way flight-time-limitation rules cap real duty periods. Hours flown past the cap still complete normally — nobody's flight gets interrupted or rejected — they just stop adding to the score. This is realism bookkeeping, not a penalty: the same pilot flying around the clock for weeks straight would look implausible in a real airline's records too.
Ranking credit is anchored to wall-clock time, not simulator time. A flight flown at 4x sim rate is measured against the wall clock it actually took, so it earns roughly a quarter of the credit a real-time flight would. This isn't a punishment for using time acceleration — it just means credit tracks the real time a flight took to fly, the same way it would for any other activity.
Flights with physically implausible data, such as groundspeeds beyond what the aircraft can do, get flagged as suspect and excluded from the next ranking computation while they're reviewed. Nothing punitive happens automatically. A human looks at flagged activity before any decision is made about it.
An airline that's found to be deliberately manipulating its flight data can be removed from the rankings entirely. That call is made by an operator reviewing the evidence, never by an automated system. It's reserved for deliberate manipulation, not for the ordinary noise of real flying.