Unusual number of crashes out of JFK

[Otterman]

Is it me, or are other people noticing that flights departing JFK are coming down in far greater numbers then statistically they "should" be. Not referring to any of the WTC flights, since none of them took of from there. But Swissair, Egyptair, TWA 800, AA 587, are just a few in the last few years, more the further back you go. JFK does not represent an extremely large volume by American standards, but it is something that crossed my mind. Any ideas, or just a statistical anomaly (probably). I am flying to JFK today, as I have done for at least 6 times per year for the last 12 years. No problems at the airport in my estimation, kind of strange though that going in you fly a non-precision approach about 50% of the time (Canasie approach 13L/R or VOR on 22L). And the departure off 31L followed by the Canarsie/Bridge climb is another interesting feature. Not many other international airports have this kind of setup (can’t think of any off-hand right now).

[stagger]

It's not just you, I'm sure many people have noticed what they think is a statistical anomaly and have erroneously assumed something strange is going on. People’s intuitions about these sorts of things are notoriously unreliable.

JFK has about 350,000 aircraft movements per year. Obviously many different aircraft types are included in this number but taking 1 in a million as the fatal event rate (not totally unreasonable) then you would expect to see one fatal event every 2.9 years - or 0.35 fatal events per year.

If you start counting in 1996 then including the four major disasters you’ve listed the event rate would be 4 in 6 years – or 0.67 events per year.

Is this strange? Well, there are at least 30 airports that are as busy or busier than JFK in terms of aircraft movements. Given this, a sample size of thirty, is it strange to find that one of these airports has an accident rate substantially higher than the expected mean? Not necessarily. Some airport has to be in the right hand tail of the distribution and if it wasn’t JFK it would be another airport.

Given a mean fatal event rate of 0.35 per million the probability of getting 4 in 6 years would be 0.095 (or about 1 in 10). So just by chance you might expect 1 in 10 airports with 350,000 movements per year to have 4 fatal events in 6 years.

[pete sahut]

My opinion is that this high rate of incidents/crashes from JFK is only pure coincidence.
As far as I know, the common (and fun-to-fly) usage of rwy 13 L/R with CRI arrival, and departure by using 31L w.turn soon after t/o, is mainly due to the three traffic areas around the three airports (JFK,Newark and La Guardia) situated tight together in the NY area. They all have to keep their traffic from each other, and other procedures then mentioned above will slow down (-I believe,as a pilot) the arr/dep rate possible at the airports.

[Roadtrip]

All of the incidents you mention had nothing to do with physically with JFK. Swissair, TWA800, and AA587 (probably) were mechanicals. Egyptair was another Arab murder/suicide.

As far as the Canarsie approach, I don't like it much either, but like any approach, safe if properly flown.

[ 17 November 2001: Message edited by: Roadtrip ]