More on Hockey Stats: What's a Corsi?
I want to be up front in saying that I have never taken a statistics course - calculus yes, statistics no. And, I am not an expert in the analysis and generation of advanced or enhanced hockey statistics. But, I have been keenly interested in what stats coaches have been finding useful at the university, Junior and NHL level over the past few months and so here is a little synopsis of what I have learned.
Two things that are important in the stats area. First, looking at data and generating statistics can pretty much tell any story you need. As the old adage goes, “you can make statistics say whatever you want”. Second, for many teams, generating advanced statistics is simply an impossibility when you don’t have a staff member or two in the stands collecting data and a video coach taping and coding the game. It’s much easier to chart all your team’s shots when you are watching for a second time on video and the file is all marked with all the shots on net.
Before talking a little about the trends in statistics, I do want to add to my blog from last week when I spoke about simple statistics (goals, assists, shots, faceoffs). Two other stats that are common are +/- and Time on Ice (TOI). I am not a big fan of +/-. I believe it is too relative to who you are playing against. In other words, if you are a third line center who is continually matched against the top line on the opposing team then your +/- is probably not going to be as good as the first line center who gets to play against the other team’s second or third line.
With respect to TOI, shift length has always been a key element of my bench management. There are some apps that are easy to use to chart TOI and someone can easily sit in the stands with an iPad and record all the player changes. As a coach, TOI stats can tell you average shift length, how much time each player is on the ice and who is playing most on the power play and penalty kill. Like goals and assists, coaches will undoubtedly see some things that are out of whack with what they are watching from the bench and make some adjustments. For instance, TOI stats might show that your third line is actually getting considerably more ice time because they are shadowing the other team’s top line (who is typically playing more). If your third line is on more that means your top two lines aren’t. With that kind of data, a coach can then decide whether it is crucial to have the third line on all the time against their number one line or, would it be better to get their first and second units on more even if it means playing against the opposition’s top line.
When it comes to tracking stats in general, what coaches want to see is something that correlates to winning hockey games. If simply the number of shots taken directly correlates to winning games then coaches would be telling their players to shoot all of the time (actually, not always a bad idea!). However, many shots taken might be from an area that creates no scoring chance.
What stats are coaches looking at then to correlate to winning? Many look for a new statistic called Corsi. Here is the definition from sportingcharts.com:
(Corsi is) an advanced statistic in hockey that measures the shot attempt differential of a player while on the ice including shots on goal, missed shots and blocked shots for and against. Corsi captures the differential of shots directed towards the net as an indication of the player’s ability to generate chances vs. giving up chances.
For example, if while a player is on the ice he gets a shot on goal, his teammates shot is blocked and an opponent shoots but misses the net, the player in question will see his Corsi Number rise by +2 (his shot on goal + teammates blocked shot - opponents missed shot).
So, coaches will use this statistic for measuring all of team play, line play and individual play. From hockeywriters.com:
The theory is that better teams possess the puck more in the offensive zone, offensive zone possessions lead to more shot attempts, more shot attempts lead to more goals, and more goals leads to winning. Plus, when your team has the puck in the offensive zone, by definition, the other team doesn’t. It makes sense.
Although I believe that Corsi is a good indicator, charting where shots are taken from, how hard they are, if there was a rebound, if there was a screen on the shot and whether there was a scoring chance would all be good or better indicators. Maybe there needs to be a “Shot Rating” score that would take all of these into account - something like a Passing Rating in football. (If you want to see a tough one to calculate look that up.)
More to come on statistics next week with a further look at puck possession measurements and the application to player and team evaluation.