As people who are obsessed over hockey in a way that I imagine nerds are obsessed over Star Wars, myself, other people who write hockey articles, the people who read them and the people who comment on them love nothing more than arguing over who is the best and who is the top five or ten at any position.
While often pointless, these arguments - when they remain respectful and diplomatic (and even occasionally when they don't) - are greatly enjoyable. Kind of a Who Shot First for a subsection of the population who enjoys their heroes in shoulder pads rather than tights and costumes.
Is Frank Millers steroid fueled beefcake Batman a better Dark Knight than Morrison's insecure, abstract Batman? Clooney or Kilmer? If you catch my drift.
In hockey, the internet has allowed us to have these conversations on a daily basis - and of course, there is and can be no conclusion, since the discussion is fraught with problems from the get-go.
Problem the First: No can decide what criteria makes a good defenseman. Should we favor offense as the Norris Trophy Voters (a dubious subset of people if ever there was one) clearly do?
How exactly do you measure defense? In today's day and age, we have virtually unlimited statistics with which to make our case and - as the rest of the world agrees, with the notable exception of Global Warming - collecting and quantifying data is the main way to understand something.
But this leads us to
Problem the Second: The NHL and those who follow it has seemingly been divided into two groups: people who welcome statistical analysis as the basis for evaluating players and people who feel it's better to trust your eyes.
Now, I won't play objective here: I belong to the group who believes measuring data is a far more accurate way to evaluate a player. Both have problems, but even if both problems were equal, I'll take the one that can at least proffer a better argument than "because I said so."
Here's why: I acknowledge that there are things I simply cannot get from a stat sheet. I understand that it is problematic to compare even the same stat across different teams and players, due to there being a massive amount of variables from team to team, whether in terms of deployment or competition or whatever.
I also understand that no two people put the same weight on the same statistics making "proof" using data, at best, extremely dubious.
I get that.
But here's the alternative: Watching the games and deciding with your own eyes who is the most effective.
Here's the problem with that: Every human is subject to massive degrees of what is called Confirmation Bias. This means that even the best off us ultimately tends to draw the conclusions we want from our observation.
A real world example is this: Jake Gardiner has a reputation for being bad in his own end. When he makes a mistake, you notice it. Yet, you do not notice when a player you aren't specifically watching makes a similar one, nor when Gardiner himself makes a solid defensive play. At best, if you conclude Gardiner is a bad defenseman - from what you think you see and what you do hear - you're conclusion is ludicrous.
There are enough studies done in real world science that make this in an absolute fact. (Which is why when you observe something, you use the scientific method to find out if what you see is in fact what is really happening).
There are also enough studies done in real world science that show that humans are terrible at observation, that they cannot be trusted to provided real memories to events they just witnessed and that generally, if you knew how truly deceptive human vision actually was, you'd probably go insane because you wouldn't trust a damn thing you saw.
Beyond Confirmation Bias - which, I stress, cannot be avoided, by anyone - there is the obvious fact that no one - not even a scout, a GM or a Hall-of-Fame coach being paid thousands of dollars per year and making it their occupation - can watch enough games by enough teams to use their eyes to say who is the best.
So, even if we could observe things properly and accurately, we would be forced to rely on scouting reports and word of mouth i.e reputation as a large portion of our analysis.
There are many studies that show scouting reports favor older, established players and tend to fail constantly to identify the best statistical players.
Then there is the very real problem that "say it until it's true" is a very real phenomenon. There is excellent evidence that suggests that the Collective Conscience is nothing more than an echo chamber for B.S.
Which brings us to
Problem the Third:We can't trust our eyes and statistics are difficult to analyze.
So, big deal. There is no way we can get a consensus on this subject anyways. All we should be trying to do is have conversation, question, think and maybe learn something. It's all in good fun. I mean, no one is going to reference Tanner's Defensive Rankings when they do anything - and since it's not even printed on real paper, you can't even use it to wipe anything or realistically suggest someone put it somewhere inappropriate.
I think it's important we come to this conversation with the idea that we can't come to a conclusion. That it is literally impossible.
But we can dissect reputation and we can recognize guys who maybe aren't getting the recognition they deserve. For instance, is Weber really the best? Just how far has Chara's game fallen? Is Muzzin legitimately Doughty's equal? I'll bet you have answers off the top of your head for each of those questions. But ask yourself how do you know that?
Just assuming the Eye Test is a valid way to make these conclusions, how many of each of these teams games did you watch? Did you take notes? How closely did you pay attention and what are your criteria for deciding what makes someone good?
At least if you made these declarations with statistics you could show how you got your conclusion and leave it to the person you're telling to decide just how poorly you applied what your data, or how poorly you weighted each criteria.
We should acknowledge that an exclusionary stance against either statistics or observation is pretty ridiculous. People say there are gaps in statistics and that they cannot show the true worth of Player X or Y. Sure, that might be true, but the gaps in your ability to watch a game and see - between trips to the fridge and arguments with your buddies, and likely a rapid decrease in your sobriety - are just as bad.
In short, both ways are problematic and its stupid to make them mutually exclusive. I have never met anyone who doesn't watch hockey but enjoys hockey stats. It would be like reading music instead of turning on the radio.
Conclusion:
Ultimately, where I come in on this divide is that I watch a lot of hockey highlights (which favor offense) and a lot of games that feature Toronto and Arizona (which feature a lot of losing) and I am subject to the same collective conscience as everyone else that tells us that the best defenseman is Shea Weber and the best player Sidney Crosby.
The data tells us something different and as a fan, a nerd, a writer, an analyst, it is my prerogative (and job) to try to deduce something out of the fact that the data and the concrete majority consensus differ.
All I can do is take what I see, and then look at as many different stats as I can. Then I take those stats and I try to place relative importance on each of them so that I can conclude who is the best. I won't convince anyone, and I won't be right. There can be no right answer, because the fact is there will always be two groups: one who thinks they know better than the general public, and one who knows that fifty million Elvis fans can't be wrong.
But knowing that doesn't make everyone equally right, and here is why I put more stock in statistics than some guy's opinion, even if the stats aren't perfect:
1. No one says stats are perfect. I think most people who are professional data analyzers would tell you that no one stat can measure everything or determine what is "best" in any field or instance of quantification.
To look for one in hockey is misguided. Measuring anything to make a subjective judgement is fraught with caveats and always will be.
You can measure the temperature, but if you say there is an optimal temperature, all your doing is making an informed guess on what most people will be the most comfortable at.
2. Experience tells me the point of most data is to get an idea of something. You can get accurate and scientific data, but as soon as you use it to make a comparative conclusion, it can't be a "fact".
3. The eye-test is the worst way to make determinations. That doesn't make your observations meaningless.
4. No one can agree on what qualities make up the best defenseman, so there is no TRUE answer.
5. The data is imperfect but still better than just outright guessing or going along with consensus.
6. In general, "collective conscience" or generally considered "common sense" are - AT BEST - usually extremely simplified, and often out-right wrong.
Look at the most popular catch phrases, the common buzzwords, the most popular movies, musicians, political ideas and you'll realize these aren't people you know or even want to.
A Global Warming denying, Maroon Five fan who watches the Voice, Real Houswives of Coldwater Ontario, and goes to a lot of Nickleback concerts, has a Billy the Bass hat, and says things like 'Derp' and 'Oh Snap' who is always taking "selfies" and posting pictures of food that they eat online?
You know in your heart that's wrong; that you hate that person to the core of your very sole, so what I am saying, is that in general you should be distrustful of "everyone knows that" opinions.
So here's my question: Can Drew Doughty and Shea Weber make the top five of NHL defenseman, this year based on their statistical performances?
Here is my top five D, based on nothing buy my own observation and ideas, not just based on play this year, but also on how I perceive their talent level and skill-set.
1. Erik karlsson
2. Drew Doughty
3. PK Subban
4. Shea Weber
5. Duncan Keith / Kris Letang
If I had to vote a Norris ballot just off the top of my head, right now and today, that's the order I would put them in. I might actually get sentimental and put Letang first because he had a stroke and I like his story.
Tomorrow, I will use statistics to make a list of the top five defenseman in the NHL and we'll compare and contrast and see how they look.
Blogger • 
Join the Discussion:
»