I’m with Raindance on this. The typical strong Nino response in New England is warm and wet, and well….. it was both a historically warm and historically wet winter here. The overfitting analogy is a great one, it’s easy to think deeper analysis = better, but it doesn’t necessarily work that way. When you are training machine learning models, your accuracy actually begins decreasing after a certain point because the model begins to capture the noise rather than capture the overall trend line. It’s the same idea with this, sometimes the best way to look at it is to zoom out and look at the large scale temp and precip profiles rather than over analyze every 500mb nuance. Just to be clear, I am as guilty as anyone of overanalyzing the 500mb, so im not saying this to take a dig at others.