Welcome to PuckOn.Net

Puck on Net is a site dedicated to exploring and developing hockey statistics. Because this site is interested in novel ways to look at hockey statistics, you may see some ideas here that you've never seen before. Below is a short description of some of these ideas, however this list is far from comprehensive. This site does assume you have an understanding of some of the more common terms statistical terms like Corsi or Fenwick.

Score Adjustment

Score Adjustment is a correction made to Fenwick or Corsi to account for changes in possession that occur due to the current score of the game. When teams are ahead, they generally fall into a small defensive shell to protect their lead and end up shooting less. Likewise when teams are behind they tend to take more risks and shoot more to try to catch up. This effect becomes more pronounced when the score difference increases. Because of this, events that happen when teams are trailing or leading by different amounts aren't entirely equitable to each other since the frequency of those shot attempts change. Score Adjustment corrects for this disparity by more heavily weighting shots for players on a team when they're ahead and diminishing the weight of shots for players on a team when they're behind. You can find more information on Score Adjustment here.

Venue Adjustment

The phrase "home team advantage" has a very real statistical counterpart - home teams shoot more than away teams do. The reason for this isn't in the scope of this site - it could be due to player familiarity with their own venue, rules benefitting the home team, or the roar of the crowd cheering you on - but it's apparent that it's slightly easier for teams to generate possession when they're at home versus away. Thus, a home performance isn't directly comparable to an away performance. Venue Adjustment corrects for this by slightly increasing away numbers and slightly decreasing home numbers. The article introducing Venue Adjustment to this site can be found here.

Event Adjustment

Categorically speaking, it's fairly obvious that goals, shots on goal, blocked shots, and missed shots are all different, so it should be no surprise that treating them all the same may introduce some unintended statistical noise. Unfortunately unadjusted Corsi and Fenwick do just this and regard each different shot attempt type as equitable to the next. It turns out we can remove some of this noise by weighting each event type by its rarity. The more rare events, such as goals, tend to have stronger weights while the more frequent events such as shots on goal tend to have weaker weights. The article introducing Event Adjustment to this site can be found here.