Accuracy of long range forecasts?

[snowpromod]

I hope this question is allowed - when reading/watching weather predictions I am always amazed at the difference in long range forecasts. My question for Mr. Dee is, how far out can the weather be forecasted, and at what time frame does it become more of an educated guess?

As an example, here are two different forecasts for the same location on Thursday March, 4th:
Accuweather: H 41 - L 20
Weather Channel: H 34 - L 22

 

[jd]

There is computer guidance that goes out to day 16. Most of the reliability starts to really run out after day 7, at least with the details. There is some reliability out to day 9 or 10 for things like a big cold snap or big warm up and even some times a big storm. Basically the bigger the event, the better the models tend to do with it happening or not.

I cannot explain the difference in the two forecasts you provided, although would venture to guess that one or both are computer guidance numbers that have been "tweaked" by a human.