More on Gustavsson—what do “stats” really tell us?

Over the past fifty plus years, I’ve seen the growth in the use of statistics in all sports, including of course in the National Hockey League.  In fact, I have a specific memory of reading about the first time I heard of a team utilizing what we now commonly call the “plus/minus” stat to help them understand who was on the ice when goals for and against were happening.

I could be wrong, but my recollection is that it was the New York Rangers, in and around 1965 or so, who introduced this method of keeping track of who was on the ice for even-strength goals for—and against.  If memory serves, it was then General Manager Emile "The Cat" Francis who first revealed publicly that he was using that particular stat, though I have no idea if he “invented” the concept.

In any event, we all know that things have changed radically since then.  Team executives (and leagues) have always kept basic stats, like batting average and RBI’s in baseball and of course goals and assists in hockey.  But when I was a kid watching football, no formal statistics were kept on “sacks”, for example, and now that is everyday information—at the fingertips of even the most casual fan.

In hockey, we have progressed well beyond “plus/minus”, of course. There are so many levels of stats and so much data that I don’t honestly keep track.  I realize there is value in these numbers and every team depends on them to a certain extent—from face-off percentage to save percentage and matters much more nuanced than that.

I will leave it to those who have developed a legitimate expertise in analyzing such details to do precisely that.  While I try to be “aware” of details and key trends, I still tend to rely on my own observations and visual ‘antenna’—and experience—in determining who I think is a “player” or why a team is playing well or not.

Today, though, I will just make a broad observation that will hold no statistical weight or value whatsoever.  However, I would be interested to know if there are some other hockey observers who share my perspective. (I recognize that most will have a more modern, data-supported approach.)  My comments have primarily to do with save percentage.

Years ago, the gold standard for goalies was typically their goals-against average. In fact, so important was that “stat”, that a major end-of-season NHL award was instituted to honor the goaltender with the best GAA over the course of the then 70-game schedule (in the 1950s and ‘60s).  The trophy was named after Georges Vezina, the old-time Montreal goaltending great.  (These days it’s the Jennings Trophy, while the Vezina goes to the goalie voted as the “best” over the course of the regular-season.)  Vezina was a true legend in the game's early days.  I've included a wonderful old photo my Dad gave me probably fifty years ago (at left).

Nowadays, at least in recent times, goalies are more judged by their “save percentage”, a stat that is trotted out as near-gospel in assessing a netminder’s performance.

Bringing this discussion closer to home, our own Jonas Gustavsson has been the poster-boy for such discussion.  His “save percentage” (lifetime .898) does not fly with many other keepers, and that alone seems sufficient to deal a death blow to some fans’ interest in seeing Gustavsson stay in Maple Leaf blue and white.  Part of me sees that, not because of his save percentage, but because he has not exactly taken the goaltender job in Toronto and run with it over the past three years.  But I have opined on that subject recently, because I believe there are some legitimate reasons—not just excuses—for his lack of performing to his potential.

But here’s the thing I wanted to raise, and it is this:  I will grant that, odds are, a goaltender who has played say, 40 games in a season and has a save percentage of .940 might well be a decidedly better goalie (or at least he is playing better during that particular span of time) than a goalie whose save percentage is  .870.

But I sometimes wonder if there is necessarily much difference between a goalie with a percentage of maybe .916 and one with a percentage of .898?

I know one number sounds a lot better than the other, but then we have to (as stats people can nowadays) get into even more detail.  What about not just the number of shots, but quality shots?  Clearly, in any given game, when a goaltender faces anywhere from 25 to 45 shots, a large number will be, at best, pedestrian saves that any goalie could make.  A goalie may have a run of “easy” saves in a given night, and naturally  his save percentage will be high.  Conversely, he may only face 18 shots, but if the three shots that elude him were labeled and no goalie would make those saves, he still has a horrible save percentage that night.

I guess my simple point is that—as with the rest of life when we use certain types of “data”— stats can be manipulated a bit to help us make our point.  (I see that all the time nowadays with plus/minus.  If someone, myself included, wants to make a case for a guy playing poorly, we can trot out a negative plus/minus stat.  But someone defending that player will say, “hey, a lot of those goals weren’t his man",  "his line mates made mistakes", "the goals weren’t his fault", or "he was hopping onto the ice when the other team scored", etc.).

I’m not discounting save percentage, or goals-against average (or time of possession or any other modern-day hockey stat, for that matter).  It’s just that, like when I was a kid in the early ‘60s, I knew what I saw.  Example?  Gump Worsley would have a much worse GAA every year than Jacques Plante in Montreal, but both were amazing goalies.  Plante won the Vezina year after year. Worsley (right) played for the Rangers for a decade and routinely faced 40 plus shots a night.  Montreal had a better defense, and Plante benefitted from that, to be sure, facing far less shots and therefore managing a better GAA.  And it wasn’t just the number of shots, but quality shots, that Gump faced in New York, along and his team’s overall inability to clear rebounds and get the puck out of the zone. (Interestingly, when Worsley was traded to Montreal, his numbers improved and he won 4 Cups in like 7 seasons.)

To be clear, I’m not related to Gustavsson and won’t be devastated if he is dealt or leaves at some point. All I am saying is (and have been saying for a long time here), when we tend to want to jump on him for bad goals or poor play,  that I think the guy can play—regardless of what the "stats" tell me.  (Maybe his recent run of good games has "proven" that, maybe it hasn't.)  Sure, there are no doubt “mechanical” things that elude him, but for me, the central thing the guy has lacked is the sense that those around him truly believe in him.  Yes, he has to play well enough to earn that trust and support, but he has not had it from management here in Toronto, in my view, as I’ve posted here before.

Hopefully he can find that support and confidence here—or somewhere else.