Hacker News Reader

Comment by dagurp | Hacker News Reader

> Overall, airplanes are a remarkably safe means of travel. Even on carriers registered in highly mountainous and desperately poor Nepal, which truly is one of the most dangerous countries in the world to fly in, just one out of every 60,000 passengers has died in aviation accidents since 2000.

Wouldn't it make more sense to measure in number of crashes per flights instead of number of passengers?

View on HN

That would make small countries look unfairly safer because of their shorter flights. Maybe deaths per passenger-kilometer? Unless the duration of the flight doesn't really affect the risk.

The duration of the flight can be ignored. Almost all accidents happen during take-off or landing which is why statistics usually refer to the number of passengers.

Edit: Data to back it up: http://www.1001crash.com/index-page-statistique-lg-2-numpage...

Makes sense. Takeoff and landing are the most critical phases of flight (much less time to correct mistakes).

On the other hand, the risk of driving from New York to Boston is drastically different from the disk of driving New York to LA. So if you want to make a safety comparison between different transit systems, looking at deaths per passenger mile makes sense for the big picture.

But if all you're looking at is the risk per flight like this, it's like you said, takeoff and landing are where crashes happen regardless of distance.

If I have to move my piano 120 feet down the street, cars are considerably safer than planes.

It's very difficult to compare miles traveled. Even when comparing a cross country trip, how safe am I driving if I come anywhere close to US airline's safety standards: very well maintained car, never drive tired, avoid bad road conditions, etc...?

Yes, thank you for saying this. You'll often see the relative safety of different modes of travel compared using passenger miles. Passenger miles are an important consideration for some modes of travel, but not air travel.

This wouldn't work for Nepal, where the flights are short from a 2D perspective, but are absolutely necessary from a 3D one.

Like flying into Kai Tak, but significantly scarier?

I propose that the measurement be in crashes per aircraft type.

Not if you want to compare safety between different transport methods: train, car, bike, plane, elevator, motorcycle, walking, etc.

The "just" before the "one out of every 60,000 passengers has died" is rather problematic IMHO.

Reply on news.ycombinator.com