I have a lot of readers who revel in telling me just how bad Jake Gardiner is. They will not, under any circumstances be swayed by the evidence that he is one of the NHL's very best defensemen.
But, as anyone in the legal profession or anyone who has taken an intro to Psych course can tell you, eye-witness accounts are extremely problematic given to the inherent human tendency to fill in gaps of knowledge with our own "information" and then construe that as a memory. This is to say nothing of the unavoidable tendency to remember things that confirm your theory and disregard things that contradict it. It is called confirmation bias, and every single one of us is extremely influenced by it.
The above is a fact and can be easily confirmed by the curious reader. From an objective scientific viewpoint, watching something and then trying to recall what you saw is a very bad way to analyze anything, let alone a group of 700 players who play 1230 separate games in a year.
What that means is that even if you could watch every player and every game, your "eye-test" would still be a bad way to make decisions on players. What this means to the scouting profession and the value of the average fans and/or sports writer's opinion is that it almost invalidates it completely - hence the deep psychological urge by the general pubic to invalidate the use of statistical analysis in sports.
You see this constantly - whether on sports radio as they inflect their voice to denote sarcasm when they say "Corsi," or in the aggressive nature and the cries of "troll" from Tweets/Facebookings/comments that follow the publication of any stats based analysis that makes a seemingly shocking pronouncement (i.e Gardiner is an elite player / Shea Weber has declined drastically.)
Now, no interpretation of data is free from bias. Two intelligent and well informed people could make drastically different conclusions from the same numbers. They would then be free to engage in an interesting and stimulating argument. So please don't mistake what I am saying as "advanced stats are always right."
Nothing is usually ever "right." If you think that the above chart is skewed because Gardiner doesn't have the same quality of competition as Doughty, that would be your prerogative. I would counter that the fact that he doesn't play with as good of other players more than balances it out, but neither of us could say we are "right" in the objective sense.
I think that there is ample evidence out there that suggests that QOC and zone-starts do not factor in as much as most people seem to think, and that there is also a ton of evidence that shows that players who put up good numbers in "easier" minutes also put up good numbers when they get a chance to play on a top line in tougher situations, so I clearly win this argument, but at least my opponent is thinking about things and offering reasonable and intelligent objections.
What I am talking about here are not reasonable objections. I am taking about the outright dismissal of statistical based analysis as some kind of Voodoo. For instance if you feel the need to write "Putting Jake Gardiner and Drew Doughty just shows you know nothing about hockey" then you are the reason I wrote this article. Congratulations.
It is OK if that is you think Doughty is way better and are not interested in hearing otherwise. But I think we allow too many fallacy based straw-man arguments to go unchecked. Never, ever, say that the reason something is wrong is because thinking it proves you don't know about the subject. I know I am getting off on a tangent here, but that is like the literal worst type of argument you can make according to the literature on logic, which I am not sure but I think might predate just about all other knowledge we have.
The condescending tone and the aggressive nature of these dismissals are as bizarre as they are ironic. You cannot dismiss the value of collecting data and then finding strong and provable correlations between certain statistics and winning. It is as wrong as saying 2+2 is 145. It is not up for debate! That is an incorrect approach to the point where it devalues what other opinions you hold.
I would also like to say that this isn't an either/or proposition. You don't have to buy into stats and become a nerd. You don't have think that your opinion on what you see holds no value - it has some. What you see can and should be talked about. You should just realize you can't possibly have seen enough games and players to make a judgement unless you supplement what you see with statistical evidence. No one could.
And you don't even have to buy into it. Just stop saying that the people who are looking at data to make decisions are the crazy ones. In the history of the entire world, whenever data driven analysis starts to contradict established "experts" there is a backlash and there are people who write them off as charlatans. But in exactly 100% of the previous instances of this, the people using the data turned out to be right.
Plays from the highlights package are the ones you're going to remember, but they only account for about 2% of a game and that can negatively effect your ability to give a knowledgeable opinion on a player, for example "Jake Gardiner gives the puck away too much."
Even if he does, what you don't see in the highlight package is that his speed and passing skill allow him to get the puck out of his own end all night long. So, on a given night, Gardiner will exit his zone just a tiny bit quicker than everyone else and, over time, this adds an incredible amount of value to his ability to suppress shots by the other team, which it turns out is the most important things, bar none, that a defenseman can do.
Gardiner doesn't score a ton of points (30 is good, but Karlsson will have 80 this year) and since he always has the puck, he is bound to give it up in embarrassing fashion occasionally. People remember this stuff and think that he isn't any good. But what you don't see is all the little things Gardiner does that have the effect of limiting the other team's shots. So that is why one of the best players in the world is not considered to be nearly as good as he actually is.
Jake Gardiner is the textbook example of why stats people and non-stats people can't get along. The thing is though, that a reasonable review of the evidence shows that when a player's stats supersede his reputation, the stats people are the ones you should trust.
My opinion is not that Jake Gardiner is a better player than Drew Doughty. However, it is true that this year, this season, one player is suppressing more shots than the other while playing on a worse team.
So all things being considered, I think that Jake Gardiner has had a better 2015-2016 up to this point than the player who I suspect would be awarded the Norris Trophy if it were voted on today.
Blogger • 
Join the Discussion:
»