I firmly believe that NOAA weather stations do not have windows.
I also feel that, before computers took over, humans did a better job of predicting the weather. The human mind has a wonderful facility for recognizing patterns. The number crunching weather computer programs, while improving, still lack the pattern recognition which a trained weather forecaster has (or at least, used to have.) I feel that Chris Parker does a much better job than all the fancy NOAA gadgets.
If one is going to venture to sea, he needs to rely on himself for the weather forecast. Listen to the weather conditions
over the course of several days, see where the fronts are and how fast and in what direction they are moving, take weather courses so you know how two fronts are likely to interact, watch the weather for several days before you venture out and make your own predictions. See how they agree with what actually happens where you are. The more you do this, the more accurate you will become. Nobody ever gets it right 100% of the time, but you'll probably do as well or better for your immediate vicinity than those who are attempting to predict on a much larger area.
It would be interesting if someone started a database which would compare predictions of various prognosticators with what actually happens. Whether or not.