It was, in fact, wrong. The chance was 100%. Their models were inaccurate, due to data they couldn't have, I imagine. This was predictable this morning with perfect knowledge. Knowing the weather 6 hours in advance to that level is just not unknowable in that way. So if someone hands you a die and says there is a ~17% probability that you roll a 4, and you proceed to do just that, is that person wrong or right? From a guy who regularly states the % likelihood of things happening, I expect a firmer understanding of probability. In other news, this baseball game needs to happen. Neither. The earlier posters confused a late hs/early college math or chemistry class type probability with a model-based probability, which is entirely different. The model-based probability is only correct to the degree that the model is correct, and no meteorologist will tell you he has a perfect model. Edit: but if you were to pause the camera as the die was thrown, then use a perfect model to determine how it would land, yes, that's wrong. Further edit: or you could have a simple 1/6 model which doesn't pretend to use complex factors to decide, in which case it's undetermined.