Corsi begets scoring chances, scoring chances beget goals - Vic Ferrari
If I asked you to give me a simple one sentence all-encompassing description of what the point of hockey is, what would you say? My guess is that you would probably say something similar to "in hockey you're trying to score more goals than your opponent". You'd be right, yet from that one sentence a hockey mind can come up with infinite different strategies for reaching that goal.
You could have a team oriented towards offense or towards defense. You could have a puck possession team or a dump and chase team. You could have a team that shortens the game by not allowing anything to happen in the first period if you want (Mr. Tippett). Ultimately though, you have to score more goals than your opponent. To do that consistently you must generate more net scoring chances than your opponent.
Understanding scoring chances are key to this stage of the evolution of statistical evaluation in hockey. More scoring chances have been proven to correlate strongly with more goals scored, and that shouldn't be a shock. It's a very intuitive idea which is part of what makes the tracking of scoring chances necessary. It makes sense, and it has math to back it up. The correlation between Corsi, Scoring Chances, and Goals has been proven.
The data, however, is limited and sparse for reasons too numerous to appropriately discuss in this space. Thus, The Scoring Chance Project was born to give hockey researchers a larger sample of data to study. I will be taking part in the project to track the Stars and give DefendingBigD some exciting and relatively exclusive content. Follow the jump to get all of the details.
We should probably get two statistical definitions out of the way immediately.
Corsi - This is the number of shots directed towards the net when a given player or team is on the ice. Negative Corsi means you give up more attempted shots than shot attempts you generate. Attempted Shots Against (negative) plus Attempted Shots For (positive) equals Net Shots Attempted, which is your Corsi Number.
Scoring Chance - We're all familiar with Scoring Chances, but the concept is relatively open-ended. For this project there will be a more concrete definition. A Scoring Chance will be awarded when the puck is directed towards the net from a dangerous scoring area. Blocked shots will not be included, but missed shots will be. The nebulous-sounding "dangerous scoring area" is loosely defined as the area inside "home plate" in the image below:
*If you look at this image and think "hey, that looks surprisingly like my definition of the slot" then you're on the right track. Image originally appeared on Copper & Blue.
These pieces of data are tracked by a good number of NHL teams and guarded as a trade secret. Derek Zona of Copper & Blue put together this timeline of NHL Scoring Chance tracking to drive home the point:
As Vic Ferarri has written, Neilson was streamlining his reporting long before anyone else. We also know that Harry Sinden was tracking the equivalent of Corsi during the 1972 Summit Series. If Sinden was doing this in the Summit Series, we can also reasonably assume the Boston Bruins to be doing the same in the late 1960s.
All of this is significant because we know that NHL teams are teams are tracking advanced stats (AKA microstats, underlying stats) to some degree. The Sabres have been doing it since the days of Roger Neilson, same with the Rangers. The Sabres have been using Corsi since Jim Corsi has been there. Smith has been tracking scoring chances since 1980 and has worked for the Canucks, the Devils, the Rangers, and Hurricanes, so we know that those four teams are also tracking, at the very least, scoring chances.
The fact that this data is followed by so many teams shows how valuable it truly is, but the fact that they guard the information so closely limits the ceiling of what Hockeymetrics can become. So, for a few years the community has taken it upon itself to track individual games. This year the goal is to track the entire season to get a full cupboard of data to analyze.
Now, what will I be doing? I will be DVR'ing each Stars game, and, using the aforementioned rules, recording Scoring Chances for each Stars game. For each Scoring Chance I will record the number of every player on the ice for both teams, the situation (5v5, 5v4, 4v5 etc), the time, the period, and any notes. These will be tracked through www.timeonice.com, and each tracker will have access to unique data you can't easily find anywhere else. I'm not 100% sure what reports will be readily available to me, but some sample reports that will be spit out (thus easy to work with) can be found at In Lou We Trust.
The numbers in a vacuum aren't the exciting aspect of this project. What is exciting is the potential applications of this data. New methods of evaluating a player's Quality of Competition and Quality of Teammates will become available. New efforts to grade special teams units will be possible. Would you like to know which goalies face the easiest shots, and which face the most challenging shots on a nightly basis? No problem once this data comes pouring in.
We at Defending Big D are excited to be associated with the project, and personally I am very excited to be working first hand at gathering the data. Over the course of the season you can feel free to email me at any time with any questions you have about the project, about the Stars numbers, or about Hockeymetrics in general. If I don't know the answer there is a pretty good chance I know someone who does, and I'll make sure to get you the most accurate answer possible.