View Single Post
Old 06-16-09, 08:53 PM   #34
LGN1
Ace of the Deep
 
Join Date: Mar 2006
Posts: 1,138
Downloads: 147
Uploads: 12
Default

Quote:
Originally Posted by Pisces View Post
If it was only that simple. Percentage errors add up. If you have a mastheight error of 5 percent in the recognition manual, and an observation error of 4 percent, the endresult is anywhere within 9%. The error in the end result is not as big as the worst of it's parts. Except when the biggest cause of errors is a magnitude bigger than the rest. Then you can ignore the lesser ones to make it easy on yourself.
Yes, sure, I agree with you, but using makman94's data is probably still more accurate than the uncorrected GWX values (at least for some ships). I don't know for sure, but I guess that also in real life they didn't know the mast height more accurately than +-1m (ship type, loaded/not loaded ship,....?). And I think 10% is still acceptable.
LGN1 is offline   Reply With Quote