Quote:
Originally Posted by kapuhy
Celestial fix will give you much better result. Dead reckoning error is based on distance between fixes, so if you didn't ask navigator for a fix for some time, his error will grow (by default, up to 15% of distance travelled since last fix - you can change this in OFEV). Unfortunately he lacks common sense to realize that with all that sea around he can't possibly be in Warsaw.
|
Interesting . . . I must say, he doesn't seem very good at his job. I spent some time lurking in daylight about 15 or 18 km away from the northern promontory in Danzig Bay. I had come to a halt sometime after my watch crew identified the coastal battery and I did some observations on the bearings to the right-hand and left-hand edges of the visible land. I then (with the boat at full stop and motionless) asked the navigator to get a fix with celestial. His estimate was just simply BAD, and literally impossible given the measures I was observing.
I've noticed him doing this type of thing at least a half-dozen times.
In contrast, after departing Memel, his initial fix (in daylight at stop after Harbor Pilot dropped us off) for the boat wasn't too bad. Just eyeballing it, it seemed a bit off (he seems to want a Mediterranean vacation cause he does seem to error in the southward direction
), but only a minute or less (~<500m). Plotting a course toward Danzig at like 235 or 245 and then sticking to that course at a slow or 1/3 speed . . . his estimates also seemed pretty good. But once I got into the Bay, and started alternating between various courses, depth settings, forward motion, backward motion, etc. his estimates seemed to get worse.
It is not a really big deal in terms of gameplay. If the boat is close enough to land to get a fix on two reasonable locations then one can always do at least as well if not better than the navigator algorithm, so him being in error during harbor lurking operations isn't a big deal.
Mainly I was just curious how it works (the navigator location-fixing algorithm) and if the somewhat variable degree of error is common. Also, I assume it is best to correct the estimates when it is obvious?