NavList:
A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
From: John Karl
Date: 2013 Mar 21, 19:01 -0700
Frank,
The estimated position is not my idea, it’s been around a long time, for example it’s even in Dutton’s “Nautical Navigation.” I simply pointed out that the traditional running fix doesn’t follow the simple logic of estimation theory: use all available information, make no unjustifiable assumptions, and allow no contradictions. Given the NavList discussion on statistics and mention of Bayesian theory, I thought some readers might find my post (#23001) interesting. Do we really need simulated scenarios to point out the obvious, that the traditional running fix completely, and arbitrarily, honors one independent component of the DR track while completely ignoring the other – all with absolutely no justification?? Can the simple observation that this assumption is completely unrelated to the DR uncertainty (and is thus false, unjustifiable, and contradictory) be just hand waving and hot air?? Well, I’d be happy to respond to any specific inquiries on that point.
And of course in tradition, particularly in a topic as old as marine navigation, it is difficult for many people to even lean outside-that-box, even for an simple discussion. The likes of E. T. Jaynes, John Burg, and Enders Robinson found this true decades ago when they started applying Bayesian concepts to inference problems. And my observation is trivial compared to their subtle (apparently to many) change in viewpoint from that of the “frequentists” statisticians to that of Bayesians.
To all, a hand wave & a warm breeze,
JK
----------------------------------------------------------------
NavList message boards and member settings: www.fer3.com/NavList
Members may optionally receive posts by email.
To cancel email delivery, send a message to NoMail[at]fer3.com
----------------------------------------------------------------