The news that a snow system was a possibility this weekend began to show up in the various weather sources a week ago, with varying detail. At that time, the storm was a low-pressure area 6,000 miles west of here. No meteorologist in the world could forecast where that storm was headed without the use of computer models.
The fault in these so-called "medium-range models" is that in order to calculate what the weather will be in seven to 10 days, they must lose some spatial resolution. They lack details. Subtle features such as plumes of dry air or warm air are often missed, resulting in model forecasts of rain or snow that are too high.
Sometimes, inexperienced forecasters will buy into these too-high model forecasts of very heavy snow. Don't be duped. It is usually not realistic to attempt to forecast snow until about two days before the storm hits.