Some early season projections are starting to leak out from the people that use stat based models. Here is one of the latest ones that puts Pittsburgh atop the NHL
Raffi Torres has been given a professional tryout from the Carolina Hurricanes. This is sure to make a lot of people upset because he is one of the dirtiest players in the league. However, he is a lot like one of the former dirtier players in the league, Matt Cooke. They both exhibited the ability to play good hockey in-between taking cheap shots at their fellow players. Because of that there is usually going to be at least one team willing to take a chance on them. His three most recent NHL seasons saw him put up 1.68, 1.78 and 2.08 5v5 points per 60 which is very good. Although, due to suspensions and injuries Raffi Torres has only played 11 professional games since 2012-13 (5 NHL 6 AHL). He was productive but is now 34 years old. Can he still function? Matt Cooke did up until his career ended
Like it or lump it teams around the NHL will continue to take chances with dirty players if they can show the ability to play.
It does make you wonder why a guy like this can't find NHL work right now though before somebody like Torres
Travis Yost wrote about using 5v5 points/60 and relative CorsiFor/60 to predict future offense. Those two measurements have the highest r2 values in trying to predict future offensive success. He combined the two.
The uniqueness of these two measurements (avoiding collinearity) allows us to consider the possibility of using both variables to forecast future scoring, as opposed to simply one measurement versus the other. And by using a combination of the two variables, our model can forecast with a higher level of accuracy.
Here are some Pittsburgh centric figures of note for the 2016-17 season:
Thanks for reading!
Pittsburgh Penguins Blogger • 
Join the Discussion:
»