HockeyAnalysis.com Statistics
Welcome to the HockeyAnalysis.com Statistics website. Here you will find all sorts of statistics that I have calculate and I think you will find useful. Statistics are available for both the 2006-07 and the current 2007-08 season. You can access these statistics from the menu on the right.
Power Rankings
The Power Rankings are algorithm based and takes into account a teams record (while considering shootouts as ties) and the strength of the opponents that they have played. These rankings may vary considerably from the actual league standings (and other people rankings), particularly early in the season, because of the heavily unbalanced schedule and relative differences in strengths of the divisions. These rankings are a basis for the algorithm calculated predictions I make at PredictHockey.com. I have also included tables showing each teams record againt the top 10, middle 10 and bottom 10 rated teams so you can get a better idea of how tough an opponents schedule has been and how well they have done against various difficulty opponents.
Player Rankings
The Player rankings are calculated by considering how many goals are scored or given up when a player is on the ice compared to when that player is not on the ice. Also considered is how well players who play with him perform compared to how well they perform when they are not playing with him. The player performance with and against other players stats (see below) is a basis for the computation of these stats. Ratings are given for defense, offense and overall performance. The ratings are ice time independent. Also shown is a Total Contribution rating which is loosely defined as the players rating multiplied by the players ice time. So, a player with a good overall rating and who plays a lot will get the highest total contribution scores.
Takeaways/Giveaways/Hits
The NHL keeps track of what they call RTSS stats which include hits, giveaways, takeaways, blocked shots, etc. While these are nice stats to take a look at they are relatively useless as presented by the NHL. The problem is that game monitors in different arenas have different ideas of what is a hit or a giveaway or a takeaway, etc. Some cities are really stingy in giving players credits for these stats while others are quite generous in crediting players. So what I have done is to attempt to remove these biases by adjusting the stats to reduce the number of stats awarded in 'generous' cities and boosting the number of stats awarded in stingy cities. So far I have done this for hits, giveaways and takeaways for both teams and players. I have also taken a look at how many giveaways have led to shots on goal and goals against but in doing so instead of using adjusted stats I only looked at road stats which would eliminate a significant portion of the bias in the RTSS stats. It's not perfect but it is better than nothing
Stolen and Blown Games
This is a very crude way of evaluating a teams goaltending by looking at how many times a goalie has blown a game or stolen a game. To do this I look at goals scored by the team as well as the number of shots the opponent took (and thus saves the goalie made). Please look at the the specific pages for formal definitions of stolen and blown games.
Player Performance with Teammates and Against Opponents
Finally I have a whole lot of tables (for pretty much every player) of how each player performs when they are on the ice compared to when they are not on the ice as well as when they play with teammates vs when they don't play with them and when they play against certain opponents vs when they aren't playing against certain opponents. These stats are the basis for computing my player rankings. Players who consistently make the players around them better while making opponents become less productive than normal will get the best player rating.
