ON CORSI & CONTEXT
As you can see, the red line indicates team Corsi value and the blue line indicates team ZoneStart, or (Defensive Faceoffs-Offensive Faceoffs). Why is the above graph important? Because it brings context to these two advanced statistics -- they are inversely related. This holds at the team level as well as the individual level and has a large impact on individual performance as well. A line that gets buried in their own end impacts the next group of gentlemen coming on to the ice.
Corsi is far and away the best stat we currently have to measure territoriality. There is no questioning this. However, a general lack of understanding about what Corsi means, and a lack of context for game situations has led Corsi to be used incorrectly in a number of quarters. It provides minimal value without context.
It’s a raw list of people that spent the most time in one end or the other. It doesn’t account for where these players started on the ice.
A player can be put on the ice in good situations against weak competition like George Parros in Anaheim, or he can be put out in the defensive zone against the best competition like Shawn Horcoff. If they happen to have a similar Corsi values, it’s an error in grouping, not because they are similar players. Context matters when it comes to comparables.
CORSI AND CONTEXT
(March 2013)
The starting point for me, when I’m looking at Corsi data, is always remembering that there are different expectations depending on where you are in the lineup. No hockey team is made up entirely of Datsyuks, puck possession wizards who crush the opposition in terms of shots when they’re on the ice. Generally speaking, as you go further down the lineup, the players get weaker in terms of their ability to gain and keep possession of the puck.
I wanted to illustrate this because I think that the ranges of performance are something that doesn’t get talked about enough. People say “Oh, he’s in the black, he’s good” or “Well, he’s in the red, he stinks” and it misses the nuance that the “that’s ok” line is at different spots depending on where you are in the lineup. There are absolutely guys in the NHL who would have terrible possession numbers if thrust into a top line role who are valuable players lower down the lineup
Note the Corsi% difference between the average guy in bucket 1 and guy in bucket 4: 52.1% to 47.1%. If you’re using Corsi data and not taking the role that a player plays into consideration, you’re missing out on a huge chunk of information. Another point, in relation to coaching: if you add up the SF/60 and SA/60 lines for the first and fourth liners, you see a pretty huge gap. First liners see 60.8 S/60 when they’re on the ice; fourth liners see just 53.9 S/60. Coaching and defensive play kills entertainment
Once you’ve sorted by ice time, Corsi looks like a pretty solid indicator of how good a first, second, third or fourth line has performed, relative to other first, second, third or fourth lines. It’s not like there are groups with poor Corsi relative to the class that they’re in who are able to make it up with their finishing skills.
A word about fourth lines – I mentioned the old Vic Ferrari dictum about people getting way too excited about fourth lines when it’s their top lines that kill the team. This data is supportive of that, I think. Look at the spread between good and bad first/second/third lines and then look at the spread between good and bad fourth lines. I suspect that what I call the “win value” of the goals that the fourth line tends to be involved with are lower too. Fourth lines tend not to play in the last ten minutes when the game’s on the Line. I would think that a higher percentage of the goals in which they (4th line) are on the ice are irrelevant to the outcome of a hockey game than with guys who play higher up.
Of course, that doesn’t mean that you should not care about your fourth line or put a terrible one together, unless it’s a deliberate decision to permit you to focus more resources elsewhere in your lineup.
OVEREMPHASIZING CONTEXT A MISTAKE TOO
In fact, what people who have looked into these things have frequently found is that certain parts of context simply don’t make as much of a difference as you would think.
The impact of quality of competition seems to be very low over the course of a season since the difference between players’ competition faced for the most part evens out over the long run. Nearly all players face an average opponent somewhere between +1 Corsi/60 or -1 Corsi/60 (and the most extreme are around 2). That’s a 2 shot difference – basically nothing.
The impact of zone-starts is similarly a lot less than previously believed – around .31 Fenwick per o zone start (roughly .4 Corsi) as seen in this Eric T study, which is about half of what was previously believed.
This isn’t to say that all of context isn’t important of course – all findings have found that quality of teammates can make a big difference in shot differentials (since unlike competition, a team can control the level of teammates each player is on the ice with to near perfection).
But the point is that in an effort to make sure we DO control for context, we too often go overboard and over emphasize context to the point where we fail to recognize what is right between our noses. Sometimes players with bad numbers are bad. Sometimes players with good numbers are good.
PREDICTIVE MODELLING GOAL DIFFERENTIAL
Corsi is part of the larger puzzle in trying to gain greater understanding of the game and how a player can affect their team’s chance to win. Like all statistics though, it needs appropriate sample size and context, and will never tell you everything. Teammates, opponents, luck, system, strategy and what moments a coach deploys a player will always effect results.
While Corsi does tend to need less context than many other hockey statistics, there are some things that need to be kept in mind in how two players with the same Corsi% are not always created equally.
1) The relationship between Corsi% and goal differential and
2) That a “good or bad” Corsi value is relative dependent on the context of which line a player is (ie: a good Corsi% for a 3rd liner is not quite the same as a good Corsi% for a 1st line player).
Thanks to the poor sustainability of on-ice shooting percentages, the difference in goaltender talent, and the low-occurrence of goals, goal differentials are highly volatile. Corsi though is not affected by goaltending or shooting percentages and stabilizes much earlier.
You can look at a lines Corsi% and get a good idea at how much they are helping/hurting a team, and even make that relative to league average for the same lines on different teams.
Another thing to note is that that while the average for each line is different they predominately overlap in Corsi%; however, the expected goal differential per minute for different lines is quite different given the same Corsi%.
In the future we may be able to separate Corsi into its individual parts (like Stephen Burtch has been doing with dCorsi), add in context/usage variables (like Quality of Teammate or Zone Start %) and maybe even offensive production (like %TSh or P/60) to create an improved model over the one we see here.
CORSI SH% AND COACHING CHANGES
we know that teams can manipulate statistical patterns by changing their playing style based on the score of the game (score effects are a well researched and fully accepted concept in hockey analytics) so it isn’t difficult to envision that various other statistical patterns could be altered by organizational or coaching philosophies.
EFFECT OF COACHING/PLAY STYLE CHANGE ON CORSI
https://web.archive.org/web/20160411153230/http://hockeyanalysis.com/2015/02/22/bozak-corsi-dilemma/
Over the past few weeks, I have looked at the Leafs performance this season under Randy Carlyle and under Peter Horachek.
The conclusion from these posts is that a significant portion of the Leafs’ improved Corsi statistics is driven by the Leafs top line, and that outside of the top line not a lot has changed with respect to their Corsi statistics
Under Carlyle, the trio of Bozak, Kessel and JVR were pretty close to a league-worst Corsi line, with Bozak being the worst of the three. Under Horachek, they are well above the break even 50.0% line and have put up pretty good Corsi percentages. As far as Corsi is concerned, this trio went from downright awful to well above average. All it took was, I presume, a playing style change demanded by a new coach.
For several years it has been believed that Corsi is an important tool in evaluating players. It was a major component of what has driven the analytics community to conclude that Bozak is a poor hockey player. The evidence above suggests that a simple playing style change can drive Corsi from downright terrible to pretty good. This leads to a bit of a dilemma within hockey analytics, which I will call the Bozak-Corsi dilemma, with two serious questions that need to be answered:
· Is Bozak now a pretty good player?
· More importantly, if a player (or a forward line) can dramatically alter their Corsi overnight seemingly solely through changing playing style (driven by a coaching change), it must be concluded that Corsi is not primarily driven by individual player talent.
The answer to the first question is no (all Bozak’s other stats remained the same)
The second point is critically important, though, because it basically implies that Corsi has significantly less value (maybe little or no value) in individual player evaluation than previously thought, which should send ripples throughout the hockey analytics community. If Corsi is largely driven by playing style, one must conclude it isn’t an individual skill?
Comments