Foyston had success at every forward position.
That's kind of a vague statement. Like saying that Frank Nighbor or Joe Malone had "success" at LW. Doesn't quantify at all what level of experience they actually have at the position.
A few games at one position in real life, regardless of success, doesn't necessarily mean a player can or should play there in the ATD.
Did I just imagine the post where I explained that averages don't really apply to the subset of people selected for professional sports that well?
No, but I don't see why it's important.
Like if you adjusted a hypothetical multi sport athlete from 1900 to the NHL today and the same theoretical person to the NBA today they would be different sizes.
but who cares? We're doing a hockey ATD, not a basketball one. We should be specifically interested in size trends among this particular subset.
A player adjusted to the late 90s would be slightly larger than now etc.
That sounds like a problem for a person who expects the concept of adjusted size to be an exact science or thinks it already is.
It doesn't really matter what standard is used as long as they are accurate relative to eachother. We could "adjust to the late 90s" as you alluded to, or to today (which is nearly the exact same thing), or to 1950, that is inconsequential. Modern sizes "feel" better though, because that's what we know and see every day in the NHL.
(the late 90s size peak is actually what I use for adjustment. Heights are only down 0.2". If league size continues to drop beyond the five pounds that it has from the 2006 peak then as time goes on we will have a generation of smaller players which would buck the past trend of sizes always increasing, but for the most part the players from the generation that are really starting to sway the league average downwards are so young that we're not drafting them yet, so this is still a consideration for a couple of years down the road)
I have to say that I was surprised to see "adjusted size" rear its ugly head again. Not just because it's an eyesore in bios.
The size of humans absolutely has increased over time, mostly due to nutrition. But it hasn't been linear and it hasn't been evenly distributed.
Hod and Bruce Stuart would huge for their era, and would be like the largest players in this league by the "adjusted size" formula.
Contemporaries of Hod and Bruce talked about how they were so big because they were "corn fed" farm boys. Yes, there was a high degree of variability in nutrition back then, which there really isn't anymore, at least not in the industrialized world, where calories are plentiful.
The increase in average human size has not been linear, it is determined mainly by access to better nutrition than the previous generation. (People born during the Great Depression were not really bigger than their parents; size has basically plateaued for those born since the post-WW2 industrialization of agriculture and proliferation of petroleum-based fertilizers made calories cheap and plentiful).
I don't see why any of this is that important. There's no reason why one can be very interested in the average size of top level hockey players over the years and yet not interested in all in the average size of humans in general and the reasons for the differences from one generation to the next. It's enough to be aware that those differences exist in hockey and act accordingly.
by the way, at least in the NHL the size differences over the years have moved in as close to a linear fashion as can be expected:
https://hockey-graphs.com/2015/02/19/nhl-player-size-from-1917-18-to-2014-15-a-brief-look/
(I know this isn't key to what you were saying, but it's inaccurate to say that the Stuarts would be "like the largest players in this league" by adjusted size. Hod would only come out to 6'5", 230. Regina has three defensemen who adjust to that size, Orillia has one, and there's a handful more out there I'm sure. Bruce was obscenely tall and would come out to a lanky 6'7" and 220, and why shouldn't he be when he was a full 2" taller than anyone significant that he played against?)
I don't understand the need to create formulas for everything. Just say big players then would be big now and leave it at that.
I don't understand this thinking at all. Couldn't one just as easily say "I dont understand the need to create formulas for everything. Just say a good scorer then would be a good scorer now and leave it at that."? Why is that not the same thing? (actually, years ago we had guys who said things very similar to that... they went by the name of God Bless Canada, among others)
Here in the ATD community we like to dig, we like to analyze and we like to quantify. To us, it's not enough to just say that Mogilny had 1054 points and Mosienko had 698, or that Mosienko was a top-10 scorer 5 times and Mogilny only was twice. Neither compared them in a way that was fair or accurate. We said there had to be a better way to quantify this in a way that was fair. To us, it would never be enough to say "well, they were both good scorers and leave it at that."
And we have succeeded. VsX is far from perfect but now we all use it exactly in the way we should use adjusted size - As a guide, with the caveat that there is a certain margin of error attached to it. Just like VsX is only a starting shorthand from which we should then ask, "what about their linemates? what about playoffs? what about outside their best 7 years? what about per-game scoring? what about goals vs. assists bias?", Adjusted size should only be a starting shorthand from which we should ask, "how did they use their size? were they physical or soft? Were they known as being strong? were they durable? what strengths or weaknesses did this size or lack of, give them according to accounts of their play?"
One point of contention - just because one can explain the size differences, doesn't mean the bigger players didn't have a certain amount of competitive advantage against their peers. But I don't think that's what you meant.
agree.
Their are a few sore spots with adjusted size for me.
First, the level of crudeness in the metrics we have. If we're being honest, we all admit that the metrics aren't based in hard data. But I can't help but feel that's a cop out. Either the metric is good or it isn't. Continuing to bring it up, while tacking on "but don't take it too seriously" isn't the answer.
Second, the idea of an "adjusted" quantity used to stand in for a real one smacks of half of a lie. One of the many advantages VsX and it's variants has over H-R's Adjusted Points is that they more or less do what they say they do. They explicitly state that the number will be a percentage of another, real, number. The research aspect of the ATD should promote better information, not "alternative" information, you catch my drift? It's a stone cold fact that Wayne Gretzky scored 8.3% more points than Sergei Fedorov in 93/94. It is not a fact of any kind that Gretzky scored 119 adjusted points that year, regardless of how you arrived at that conclusion.
If we want to use a metric at all (and I approve of more work being done) the best one might be a percentage over/under average, with the benchmarks set at 6 and 200 (those being satisfying round numbers, much in the same way 100 is a satisfying round number for points/82 games).
I don't think our positions are that far apart, really.
When someone says "hod Stuart's height adjusts to 6'5" they're really saying "Hod Stuart was 3-4" taller than the average player in his time"; they're just also packaging in with it the obvious fact that someone who is 3-4" taller than the average NHLer today would be about 6'5".
I agree that using percentages would be more accurate than raw numbers; however, I don't see the need for such granularity. For example, a player 4" taller than the average in 1906 is 5.9% taller than average, and a player who meets the same standard in 2006 is 5.5% taller than average. I don't see the need for such distinction - do you? With the two averages being so close together, 4" means about the same thing in both cases. I'm not sure creating the sense that we're acting like it's more accurate than it is really helps its credibility at all.
(even in weights, a player 40 pounds heavier than average in 1906 and 2006 is 19.5-23.8% above average, a small enough difference that we're probably better off just leaving it alone and saying 40 is 40)
Certainly it could be made more accurate by tweaking the cutoffs for years used and adjustments used, but then we're getting into fractions of inches, single pounds, odd cutoffs that are not easy to remember offhand, and all for accuracy gains that are minimal. My rough system, for example, was tested using the years in which players born right in between cutoffs would be 26. That's just a catch-all age that's meant to work for everyone as best as possible, but in reality some players were just starting at that age (meaning the league average was slightly higher during their career than we are assuming for them) and some others were nearly finished by that age (meaning the opposite). If we are going to start attempting to make something that
perfectly explains how much bigger or smaller than average a player was in his career, it'll almost surely have to be done on a player-by-player basis.
Back to what I said at the start, I have no problem with someone saying "Hod Stuart was 3-4" taller than the average player in his time", and I'll treat him as being taller than a player like Vadnais, who was only 1-2" taller than the average player in his time, which is similar to what you all will do as well. If I take that an extra step further and say that Stuart is 2" taller than Vadnais on an adjusted basis, that really shouldn't be a problem.
