Assigning equal probabilities won't always work. In the case of the Hawker
collision, the jet was descending, so diving would not have increased
separation as fast as climbing. In that particular region, jets are more
likely to be descending than climbing, so maybe the choice should be
influenced by what one knows about likely traffic in the area. South of
Reno, climb. North of Reno, dive. 'Course, that depends on which way ATC is
directing traffic... which is why Minden's glider educational materials tell
us to listen to ATC once in a while.

But that aside, on average it seems to me that (in a glider) diving would be
better because one will accelerate over time, increasing the distance from
where one started out. Zooming would not be as effective because speed will
decay, so the separation will initially grow but then will not grow so much.
Then one will either have to stay up there or descend back down through the
original altitude.

Since both pilots are nearly equally likely to choose climbing or diving,
leaving a 50% chance of ending up near each others' altitude, it seems to me
that diverting horizontally 90 degrees (perpendicular to the approaching
aircraft's heading) will help just as much as as changing altitude. Again,
both pilots could choose to turn the same way, so a 50% chance of still
being on a collision course. So turn AND climb/dive, and you end up with
more like a 25% chance of being in each others' way.

Roger (who's only had close encounters of the feathered kind)

"Martin Gregorie" wrote in message
...
Mike Schumann wrote:
If you are on a collision course, how are you going to know whether
going
down is the right solution? He might be 50 ft below you, and you are
going
to dive right into him. Or he might also decide to descend.

If you need to get out from in front while thermalling, you'll need some
more airspeed to do so and that means lowering the nose to accelerate
while, hopefully, turning to aim off to one side of him.

The opposition has first to see you in order to decide to descend. If we
assign equal probabilities to him climbing, turning or diving then
anything you do to move away from his current course has a 75% chance of
being right.

Now lets be generous and give him a 50% chance of seeing you.
Probabilities multiply when they are combined, so that gives you an 87%
chance of being right.

This brings to mind an interesting suggestion that I was taught when I
got
my power license about avoiding collisions with birds: Always climb, as
the
birds will tend to dive.

That's only useful if you have an engine or airspeed. If you're
thermalling and try that you'll end up descending about 2 seconds later
and, if the opposition saw you pull up, he'll be down there waiting for
you. Besides, if he read the same book you're quoting he'll be more
likely to pull up than to dive.

In any case, you didn't say what you'd do, so I'll ask again. What would
you do that's better than diving out of the way?


--
martin@ | Martin Gregorie
gregorie. | Essex, UK
org |