Thank you! Well said and well researched by both of you. I doubt Nonis put as much thought into stats as you've done. Funny how some on here seem to put little consideration on stats. Stats measure performance. Anyone care to disagree?
They measure performance based on a certain sample size yes. There are guys playing in the NHL right now who have had a hot half-season who will never have another one again (one-hit wonders). There are guys who are slumping this season. I would never give out a rich long-term contract to a player who hasn't proven himself over a large sample size. It doesn't make sense to me.
Btw here is 2010-13 (3 year) data for the top 72 NHL defenseman (3000+ 5v5 mins) in 5v5 TOI. Unfortunately Hockey Analysis doesn't include this season for multi-year data (there's no 2010-14 or 2011-14 option available on the site). So that 3 year sample doesn't include this season.
http://stats.hockeyanalysis.com/rat...000&teamid=0&type=goals&sort=PCT&sortdir=DESC
Dion Phaneuf is ranked 50th/72 in goal differential. He ranks 30th/72 in Defensive Zone Start % (DZFO%) and 63rd/72 in Offensive Zone Start % (OZFO%). You could make the argument that Dion Phaneuf is asked to take on a lot of defensive assignments. But sure let's give 50m+/7yrs to a guy whom we're going to assume is a Top 10 defenseman in the league (after all we're gonna pay him like one). Without a shred of evidence that he's actually capable of being that guy over a long-term sample...
Let him walk. I don't care if he's a UFA and if he can get a better deal with some **** team. I don't care if it's going to take some time for us to groom or acquire an adequate replacement. Dion Phaneuf is not a part of a Stanley Cup winning puzzle. So what if we make the playoffs? Half of the conference makes the playoffs. Big deal. In the playoffs, a team can always get lucky and win a Stanley Cup. 16-28 games is a small sample size compared to 82. But we want to maximize the team's chances of winning a cup.