View Single Post
Old
01-30-2012, 05:12 PM
  #72
sousuffer
Registered User
 
Join Date: May 2007
Posts: 255
vCash: 500
Quote:
Originally Posted by NHRangerfan View Post
As I stated in an earlier response, stats in hockey or any other sport not named baseball need to be related back to team performance...so I would agree with your statement that because a formula hasn't been developed doesn't mean one can't be. However there is one caveat, any new stat must need to be regression tested, simply stated if you applied it to a TGO or Messier or Mary-O or Howe, etc what would the stat tell you about them?
I say that it needs to be regression tested because of the way the NFL developed that god-awful QB rating, Johnny Unitas is tied for 67th All-Time QB rating, he shares the same rating with Otto Graham, Jim McMahon and Bert Jones.


I agree completely...in baseball, they've tried to do this with the WAR stat, but I feel like it undervalues power hitters like Adam Dunn and overvalues fast players that can steal bases and field well like Brett Gardner. The problem with stats in hockey is the vast differences in eras, equipment, goaltending advances, defenseman blocking more shots, and many other fcators. Either way, statistics has ways to adjust for this in a way similar to park factor (Citifield is where homers go to die, where Citizens bank park is a bandbox). The key to any stat or metric is to identify the variables that contribute the most significant aspects to team success (as you said). This is similar to the "Wins" concept in baseball...you don't look to reproduce 100 RBI or 30 homers...you factor in TOTAL contribution and decide how to replace the wins that a player contributes. The key to assigning the value of intangibles (in addition to goals and assists) to wins is by looking at correlations between things like "hustle" plays and team success. Also interesting is assigning value to secondary assists, which in many cases, contribute nothing.

For instance, does a team that block a lot of shots correlate with team success? Or does it mean they spend too much time in their own zone and therefore NEEDS to block a lot of shots? I remember when the Rangers were terrible before the lockout, they always seemed to lead the "hits" category...all that seemed to mean is that they never had the puck and needed to hit people to try and get it back (their record reflected this).

In baseball, it was always thought that stolen bases and sacrifice bunts were critical to team success...after statistical analysis, it was shown that giving away outs (even to advance a player a base) almost NEVER benefits the team and is NOT good baseball (despite what every "baseball mind" would tell you). In addition, they found that your stolen base success rate needed to be a certain percentage (I believe it's 80%) before the team positively benefits from it. Just because "experts" from a sport tell you that specific actions are "playing the right way" doesn't mean they know what they're talking about. Again...it's anecdotal evidence...people remember a sac bunt that led to a walk-off win in the playoffs a lot more than unproductive sac-bunts that lead to nothing in a weekday regular season loss.

On a different note, "hockey experts" always give the Rangers credit for blocking shots. It would also be interesting to see whether or not Tortorella's "block all shots at all costs" mentality leads to increases in injuries (which obviously reduces overall future team success).

sousuffer is offline   Reply With Quote