My thanks to Airbubba for his encomium. I hope I deserve it!
Originally Posted by brooksjg
'Your' course? or a course you were at?
This one.
Concerning my crude PRA, both brooksjg and infrequentflyer789 point out that identifying revenue with net income isn't realistic. I know that, indeed I thought I had pointed it out.
Do they have a better one which can be justified using publically-available figures?
My trivial calculation yields a low estimate for the risk; in other words, if you think net income is less than this revenue estimate, which it almost certainly is, then my crude estimate of risk is an underestimate. It still seems to me pretty high, rather higher indeed than Pace's estimate of zero risk. That's why I said it yields insight, because many discussants seem not yet to have seriously attempted to estimate the risk.
Infrequentflyer789 apparently wants to add another term:
Originally Posted by infrequentflyer789
The main cost of not flying is the cost of getting your pax back.
Is it?
On the first day of the ban, I was in Delft, in the Netherlands, with people who had travelled from Vienna, Suffolk and Liverpool, all busy trying to rebook themselves by rail. None of those people have been permanently lost to their carrier of choice that day. As far as I know, Eurostar has not had a noticeable permanent slump in bookings since the debacle in mid-December 2009. And I doubt all those transatlantic passengers are going to start travelling by ship. (I do know of one set of colleagues who travelled back to Brunswick from Beijing on the Trans-Siberian railway! But they didn't all do it; my colleagues who took a chance on waiting for the next available flight managed to get back a few days earlier.)
So I am not sure about a claim that it is the
main cost. But if you have a way of including some estimate of it in a risk analysis, please go ahead.
You also want to say that the severity of losing all engines is not necessarily catastrophic. Sure, as anyone who read about the Air Transat "save" of August 2001 knows, but it is usually taken to be catastrophic as far as certification and regulation goes. Some - indeed some would argue most - engineering definitions of risk classify an event category by worst-case outcome.
Originally Posted by infrequentflyer789
Feed those numbers (150k) into your risk equation
I am not sure what equation you mean – I didn't propose one.
I don't see off-hand how «feed[ing] those numbers» in to anything I wrote will give me something which I can interpret as a risk estimate, but as I said if you can indeed produce a modified risk analysis with it according to the De Moivre formula, and you can justify the approximations, more power to you. I'd like to see it.
PBL