Skip to Main Content
The Official Site of the Tampa Bay Lightning

Hockey Analytics 101: Understanding advanced stats and how they're measured

by Andrew DeWitt / Tampa Bay Lightning

If you follow hockey in any form, you’ve no doubt heard advanced statistics and analytics since several NHL teams  hired analytics teams to assist their hockey operations staff over the summer. Some of the more common terms are starting to make their way into broadcasts and media covering hockey.

As hockey analytics make their way into the mainstream, it’s important to understand some basic terms and how they’re used to measure team performance and individual performance.

Here’s a basic rundown of the most common advanced stats:


Corsi is the most common advanced stat you’ll hear about in traditional media coverage.

If you hear someone refer to Corsi Close, the stat refers to “close game” situations and is only measured when the score is within a goal in the first two periods and tied in the third period. This reduces score effects of games that aren’t close when the team leading goes into a defensive shell and the team trailing attempts many more shots, especially in the third period.

The Formula: All team shot attempts for minus allowed while a given player is on the ice in 5 on 5 situations. Shot attempts are: shots on net + missed shots + blocked shots. Corsi for players and teams is measured as a percentage or a differential. Anything more than 50 percent shows that the team or player being measured records more shots than being given up approximating puck possession and more time in the offensive zone. If your team is in the offensive zone, it can’t give up goals.

What does it mean: Corsi is a more accurate measurement of team puck possession than individual performance because Corsi performance depends so much on teammate performance.

The Kings finished with the highest Corsi percentage in the 2013-14 regular season and the Chicago Blackhawks finished second. Of the top 10 Corsi teams last season, eight made the playoffs.


It’s calculated in the same way as Corsi except it removes blocked shots from the equation. Named after Battle of Alberta writer Matt Fenwick because he pointed out that its possible that blocking shots is a skill and should be excluded from Corsi.


This metric is best for measuring how “lucky” a team or player has been and can indicate if the team or player will likely regress back to mean as the sample size grows. It accounts for factors that are not incorporated in Corsi/Fenwick such as shooting percentage.

What it measures: on-ice shooting percentage + save percentage … a PDO of more than 1.000 is considered “lucky” while a PDO of less than 1.000 is considered bad “luck”. In the long run, for most teams, these numbers trend towards 1.000 over the course of an 82-game season.

Points per 60 minutes of Ice time (Pts/60)

Calculated by: Points/time on ice x 60

What it measures: This normalizes all players by looking at their scoring rates, regardless of their total time on ice. Last season, Steven Stamkos averaged 3.20 points per 60 minutes, which was fifth best in the NHL. In the last five seasons, Stamkos has never finished lower than 15th in the NHL in points per 60 minutes.

Zone Starts

Calculated by: Offensive zone starts/(Offensive zone +defensive zone)

What it measures: The percentage of times a player starts his shift in the offensive zone vs. the defensive zone in 5 on 5 situations only. It shows how coaches use players whether they’re a defensive zone face-off specialist or they’re expected to score goals. Players who start more shifts in the offensive zone are likely to score more points and have better Corsi and Fenwick ratings.

View More