Professional weather forecasting is an applied science. We apply our knowledge of atmospheric physics with our experience working with numerical weather predictive models. Because these weather models are paid for with tax dollars, they are in the public domain, which means the model output is freely available at a wide variety of websites where a lot of unexperienced, overzealous non-meteorologists (and a few broadcast meteorologists, unfortunately) seem to think this "raw model output" is as good as an actual forecast.
Passing off raw model output as a forecast is not what experienced weather forecasters do. We analyze. We consider past similar situations. We evaluate recent model performance and consider known model shortcomings. Our resulting forecasts are never perfect and, once in a while, quite wrong. However, we are almost always close enough to be useful while the array of unfiltered model output often leaves people confused and badly misled.