Your measurements back my intuition. Also it seems that positive NAO during negative NAO cycles are more likely than negative NAO during positive NAO cycles.
There's another factor that is tilted even more strongly towards less snow, which is that it isn't enough to have a negative NAO, you also need to have a west-based NAO. That probably lowers a favorable NAO pattern for snow to even less than 36%!
We do have rare winters like 1960-61, 1993-94 and 2002-03 where you can get a lot of snow without the cooperation of a negative NAO though. The Pacific needs to be great, which it isn't right this winter.
As per your sea ice data, is this the kind of extreme winter we're going to need to see normal sea ice coverage up north? In other words, how do you get a pattern where it's very cold up there AND very cold down here also? And when was the last time that happened, 1978-79?
Also how does the sunspot cycle factor into your calculations. I am sure you are aware that we are going to reach the peak of the cycle next year, as we have extremely hot and dry summers every 11 years or so, going back to 1933, 1944, 1955, 1966, 1977, 1988, 1999, 2010, so the next one would be in 2021. Do you think the SE Ridge flexing its muscles so strongly this winter is an early sign of the upcoming summer and especially Summer 2021 being extremely hot and dry (we get those 11 year peak heat summers with triple digit highs because the summers are dry as well as hot, a lot of rain would mute the highs.)