I've previously discussed this topic in articles before.
This is called "latent variables".
Cool.
I spent a while looking at some of the articles you've written on the subject. It's interesting and well presented. In the interests of a fun debate, I'm going to do my best to poke some holes in it.
But first, what I agree with:
I have to say though that I am extremely skeptical on your fight argument, both in injury protect and booster. There is a lot of counter evidence.
Ok. I'm going to buy this. This wasn't the main thrust of my argument, so let's agree for now that fighting isn't helpful (other than perhaps for entertainment value).
The issue with the immeasurable is that you can't measure it.
Is the player actually helping out there?
Does it actually eventually lead into making up for where they hurting?
For sure. We're relying on gut feeling, rather than hard data, to qualify the impact of these latent variables ("intangibles").
In
one of your articles, you address this, from the standpoint of coaches and other team decision makers, as follows.
They could be wrong on how much value to place into intangibles, they could be wrong which intangibles impact more, they could be wrong in how much intangible an individual has.
They could also be right.
I agree with this.
You also
say this about intangibles:
It is a completely understandable and logical assumption to make by a rational and win-interested human to make.
Still, there is the caveat that the coach or manager may be assuming wrongfully. We know that latent variables must have an impact on the game. It makes sense that they do and there is empirical evidence in workplace sociology studies.
I agree with this, too.
I'm glad you brought up the workplace sociology studies. This perhaps means that latent variables are not truly "immeasurable". They're just... "fuzzy". We know they're there. We just lack the proper techniques and data to perfectly quantify their impact.
However, the issue with this is that just because something seems intuitive, it doesn't necessarily work out in the order of importance that one would assume.
For example, we see this in face offs. Face offs matter, but not nearly to the degree that many would, and used to, think.
I buy this. Conclusions such as this are certainly a victory for advanced stats. They refute an incorrect assumption, an incorrect gut feeling that was (is?) prevalent in today's NHL.
But, now on to the
main point I want to challenge.
But, there is the issue with no one truly knowing the value of Y, or intangibles.
The safest way to look at intangibles then is as a tie breaker. On-ice contributions should still stay the primary focus with evaluating a player.
I don't see how you can justify this statement.
Since we don't know the value of Y (although our gut tells us it is important, and there are those sociology studies), I would counter it is equally dangerous to ignore it.
It's like if I gave you a mystery box. There might be poisonous spiders inside. Or there might be a bazillion dollars. Or a cure for cancer. You don't know. Is it more dangerous to open it? Or leave it closed?
In the absence of any evidence, it's hard to argue either way. But, we do have some evidence - namely that gut feeling, and those sociology studies.
If in fact the mystery box (the intangibles) are beneficial, the coaches who use them have an advantage over those who don't. So, ignoring the box is dangerous. Ignoring the box makes you a worse team.
The faceoff thing does show how our gut feelings can be misguided. And I've also appreciated arguments you've made elsewhere in how our brains tend to key in to single events (like a big hit, or bad turn over) rather than seeing the big picture.
But there are also cases where the human mind is very effective at processing information and making decisions on "immeasurable" perceptions. There are many examples of this in computing. Image recognition is one. Humans are extremely good at looking at a picture, and telling you what it is. Computers (which are akin to analytics here) are very bad at it. I remember an entertaining story about this from a Computer Science class. The US military invested in computer technology to identify pictures of tanks, say from images taken from satellite imagery. They thought it worked. But...
Eventually someone noticed that in the original set of 200 photos, all the images with tanks had been taken on a cloudy day while all the images without tanks had been taken on a sunny day. The neural network had been asked to separate the two groups of photos and it had chosen the most obvious way to do it - not by looking for a camouflaged tank hiding behind a tree, but merely by looking at the colour of the sky. The military was now the proud owner of a multi-million dollar mainframe computer that could tell you if it was sunny or not.
From
Neural Network Follies
The point is, in the absence of proper analytics, or proper computers, the safer bet may be to rely on "gut feeling". It's not infallible. But it's better than nothing, if it's all you got.
So if a coach's gut feeling tells us that Peluso's intangibles outweigh Petan's skill in certain match-ups, it's hard to argue. We can't point to analytics to do it, because we don't know the value of the intangibles. Any counter-argument we make is equally an attempt to measure the "immeasurable". That puts us on equally shaky ground that we're accusing TNSE of doing. The only difference is that Maurice is a lifelong expert, and most of us are not. And if we have to trust gut feeling one way or the other, I'd go with the lifelong expert.
).
