For the numetrically inclined, one of the most interesting reads in hockey over the past few years has been Alan Ryder's website at hockeyanalytics.com, which contains a number of Alan's detailed research papers on hockey data and statistics. Alan's the hockey equivalent of baseball's sabermetricians - made famous in Moneyball. For example, his study on first-shot quality makes a strong case for Luongo being the best goaltender in the League - and I would hazard a guess that Nonis has read it.
One of the more positive developments in hockey journalism this season, is that Alan has been given his own column in the Globe & Mail to bring his analysis to the masses. And it's available on the 'net.
His first substantive column identifies the teams most likely to improve upon their performance from last year (Pittsburg, Ottawa, Minnesota and Boston) and the teams most likely to fall off this season (Carolina, Los Angeles, Dallas and Buffalo) based upon the "Pythagorean Formula" for projected winning percentage based on known GF and GA totals. When applied to the historical GF and GA of NHL teams since WWII, its predictions are about 94% correlated with actual winning percentages.
It's a good read. If you like stats, this is the column for you!
Oh, Ryder's take on Vancouver? It turns out the 'Nucks neither underperformed or overperformed last season. They performed pretty much exactly as predicted by the model based on their GF and GA.