NavList:

   

Reply

Is Peter Fogg really claiming that he has a method which can reduce the
error resulting from random scatter to less than simple averaging will do?
Yes or no? If so, I can always produce sets of simulated data, which are
affected only by computer-generated random scatter, on which he can try his
magic, to substantiate that claim.

I understood that his reason for declining such trials, when last offered,
was that that his procedures could not be expected to improve on such
Gaussian scatter, but could only improve on non-Gaussian outliers. If I'm
wrong about that, the offer remains open.

George.

contact George Huxtable, at  george{at}hux.me.uk
or at +44 1865 820222 (from UK, 01865 820222)
or at 1 Sandy Lane, Southmoor, Abingdon, Oxon OX13 5HX, UK.
----- Original Message -----
From: "Peter Fogg" 
To: 
Sent: Wednesday, December 29, 2010 11:50 PM
Subject: [NavList] Re: errors in plotting and a possible/partial fix
thereof, as menti...


| George Huxtable wrote;
|
| > Initially quoting Gary -
| >
| > I am attaching a revised MOO table for your ship, 18 knots and a 5
minute
| > period that you can use for advancing your LOPs and for figuring the
slope
| > method for eliminating random error.
| >
| > Then George opined:
|
| > Were those words intentional? If he has evidence for any method of
| > eliminating, or even reducing, random error, that does more than simple
| > averaging of multiple observations, we would all be very interested.
| >
|
| Nah.  For quite a long time I was somewhat bemused at how difficult
| otherwise skilled and interested navigators seemed to find this fairly
| simple idea and, for anyone willing to try it, the self-evident virtues
of
| using the slope technique to analyse a series of sights in order to
reduce
| random error.
|
| Let's remember that at worst (ie; in cases where the method has little to
| offer since the pattern of sights shows little scatter) it tends to give
a
| result similar to averaging, while at best (ie; where there are
significant
| outliers) creating good data from bad, which averaging can't do, while in
| any case providing a picture of those sights to ponder, together with the
| actual slope of the apparent rise or fall of the body observed, to
compare
| those sights against.
|
| Nevertheless, over time it seems to me that this situation has changed,
and
| I get the impression that many of our regular posters understand and have
| tried the technique, use it at least occasionally and now accept its
| virtues.
|
| In other words George, many of those "very interested", if not all, have
| moved past you.
|
| However, I'm not so sure that the implications of having reduced random
| error as best can be practically done before plotting the triangle have
| fully sunk in.  The recent interest in the symmedian point is an example.
| Apparently it is a better centre of the triangle, assuming the position
| lines sides of the triangle don't meet at a point due to random error.
|
| That's fine but could be rather irrelevant if the aim, once random error
has
| been dealt with, then becomes to deal with non-random error.  Dividing
the
| angles formed by intersecting position lines, then following those
bisecting
| lines to where they meet at a point, this point being the reduction of
that
| triangle back to a single point free of non-random error, does this.  If
the
| azimuths have a greater spread than 180 degrees this point will lie at
the
| centre of the triangle.  If less than 180d the meeting point will lie
| outside.
|
| Always; no 25% here!  The whole 25% tale is a somewhat irrelevant furphy
for
| anyone using these two simple tools, since the basis of the 25% chance of
| the fix lying within the triangle relies on an assumption that only
random
| error is present, and I am still waiting for anyone, yourself included
| George, to support that assumption.
|
| Armed with these two simple tools the navigator is ready to deal with
both
| random and non-random error in a practical manner which results in a fix,
| not free of both - that would be unrealistic; some error will probably
| remain - but most likely with a reduced extent of both.  A better fix.
|

   

Reply