Hello again. Below is the passing data from the first two games of the season. I'm a huge soccer fan and those of you familiar with EPL Index and Opta Sports Data will see where I was coming from with doing this.
In soccer, the field is divided into three sections: defending third, midfield third, and attacking third. It made sense to use that a as model with hockey since we have three zones even more readily defined than soccer: defensive zone, neutral zone, and offensive zone. Then, I tracked each pass attempted and completed in each zone for each player during 5-on-5 situations.
To do that, I first had to decide "what is a pass" really. I settled on the simple definition of, "a reasonable and deliberate attempt to get the puck to a teammate." This means I'm not tracking chips out of the zone, dump ins, or anything a player does with a puck that doesn't appear to be a pass to a teammate. In grey areas, I let common sense decide (i.e. three players cycling the puck in the corner to each other).
By tracking pass data, I'm hoping it will eventually reveal several things about players: 1) how accurate are they in each zone? 2) How often do they have the puck? 3) How much of the on-ice 5-on-5 group's shot generation goes through them?
The last point about shot-generation is also a big reason why I was interested in doing this. Corsi and Fenwick take into account shots attempted, but doesn't specify how the team arrived at that shot attempt or how it was generated. Was it a breakaway? A tic-tac-toe pass play? A slow, slow soft from the point with no screen? Not every shot is created equal, nor are its origins.
So, I decided to track each pass that resulted in a shot to see which players are passing Corsi machines. I say Corsi because even if the shot is then blocked, I'm still tracking the shot attempt generated (or SAG as you'll see on the chart below).
As a further way of see who was making the most of their passes, I divided a player's total attempted passes by their total shot attempts generated (SAG) to arrive at a type of SAG efficiency rating: passes per SAG. That tells me how many passes they are attempting before a shot is generated.
via i.imgur.comFor the above Devils-Penguins Game 1 chart, the think that stood out to me right away was Andrei Loktionov generated more shots than anyone else by far. Granted, he was playing with Damien Brunner for several shifts, who probably would shoot if I passed him the puck, but still. Not only was he the 2nd best passer that game, behind only Ryan Clowe, but he generated a shot every 2.13 passes.
via i.imgur.comFor the Devils-Isles Game 2 chart, the outliers were Adam Henrique with 7 SAG and a pass per SAG rate of 3.43, good enough for second on the team that night behind only Stephen Gionta's rate of 3 passes per SAG (he had two).
I plan on doing cumulative data every 10 games or so to arrive at a SAG rate per 60 minutes, overall passing rates, but in each game you can see who bossed the puck or who struggled. Two games is too small of a sample size to determine much of anything in the larger sense.
Questions? Comments? Suggestions? Let me know what you think about this project. Like the zone exit data, I think how I approach it now is different than how I'll view it at the end of the season. Thanks for reading!