HyugaRicdeau
Baller/Shot-caller
NK:
Sorry to single you out specifically NK, what Scamp has said is correct: you're just a victim of circumstance here, no bashing of skillz intended.
Simna:
I take issue with your claim that "This list is made by results of tournaments and matches on an erratic timeline and sometimes not by results at all." I don't really know what this sentence is supposed to mean, but it insinuates things that I don't like.
Delphiki, and Simna:
As far as having "standards," I think all of the panelists agree on what is relevant information about a player, and we have made as many facts available to one another as possible. In this way, we in fact already have a set of standards/guidelines. If you want further explanation, refer to my 2 posts on this page of the WC forums: http://infinityfx.net/teamwc/viewtopic.php?t=9&postdays=0&postorder=asc&start=45
I think it is unreasonable to force everyone to weight all factors in the same way however, as this presumes an objective measure of skill, the existance of which I argue against, the same way that I argue that essentially any straight formula for rankings out of tourney placings is void. This is from my post in the WC Forums:
I disagree with the idea of it being "like a math problem." [paraphrase of Simna's earlier post]
In my first posts in this topic a page back, I gave various factors that influenced my decisions in the rankings. Now, in theory, you could assign a weight to each factor, get a quantitative evaluation of success in that factor, and add them up and rank each person in descending order. In practice this is not possible, I think you have to take too many SUBJECTIVE criteria and try to force OBJECTIVITY on them. This is rarely a worthwhile practice.
Well doesn't MLG use a points system to generate rankings? Sure, but that doesn't mean they have any theoretical basis; they are essentially completely arbitrary (even ignoring the fact that they measure only how someone does at ONLY MLG tourneys): there is no theoretical reason that X place is worth exactly Y points more than Z place.
If we try to invent a mathematical basis for our rankings, I guarantee it won't be much more accurate than MLG rankings.
You may ask about the points system we use (of giving 26-X points to Xth place, summing them, and ordering based on that), and the answer to that is that is the same as taking the average place that each person got on panelists lists. There is no subjectivity forced into objectivity there.
Let's just take the case of considering one person's win versus another. In order to have any idea of what this means, it is necessary to know what the skill gap is between the two players (by skill here I mean overall ability to win in any and all circumstances, essentially the thing that determines ranking). How can we hope to mathematically evaluate such a thing? Since no two people (let alone all panelists) will ever agree precisely how good someone is? How can you possibly convince someone else that player X is 10% better than player Y, or that beating player X is 20% harder than beating player Z, for example? "10% better" means nothing. Yet coming up with such a number is central to the idea of a mathematical formula for skill: If we are to compare the wins of 2 players mathematically, obviously we need each win to be assigned a numerical value.
To summarize this into one point, a math equation for rankings requires OBJECTIVE MEASURES (i.e. stats, like in baseball for example) that can be ordered from least to greatest. All we have in Smash that is objective to go on is a list of who beat whom (which in themselves have a HIGH degree of extenuating circumstances), but I maintain that it is impossible to come up with a meaningful way of extracting actual numbers out of them. And like in baseball, I think a poll is the best way of ranking players. Despite the enormous amount of stats that are collected in basball, in the end the MVP awards come down to a simple vote. I think we should follow this.
Sorry to single you out specifically NK, what Scamp has said is correct: you're just a victim of circumstance here, no bashing of skillz intended.
Simna:
I take issue with your claim that "This list is made by results of tournaments and matches on an erratic timeline and sometimes not by results at all." I don't really know what this sentence is supposed to mean, but it insinuates things that I don't like.
Delphiki, and Simna:
As far as having "standards," I think all of the panelists agree on what is relevant information about a player, and we have made as many facts available to one another as possible. In this way, we in fact already have a set of standards/guidelines. If you want further explanation, refer to my 2 posts on this page of the WC forums: http://infinityfx.net/teamwc/viewtopic.php?t=9&postdays=0&postorder=asc&start=45
I think it is unreasonable to force everyone to weight all factors in the same way however, as this presumes an objective measure of skill, the existance of which I argue against, the same way that I argue that essentially any straight formula for rankings out of tourney placings is void. This is from my post in the WC Forums:
I disagree with the idea of it being "like a math problem." [paraphrase of Simna's earlier post]
In my first posts in this topic a page back, I gave various factors that influenced my decisions in the rankings. Now, in theory, you could assign a weight to each factor, get a quantitative evaluation of success in that factor, and add them up and rank each person in descending order. In practice this is not possible, I think you have to take too many SUBJECTIVE criteria and try to force OBJECTIVITY on them. This is rarely a worthwhile practice.
Well doesn't MLG use a points system to generate rankings? Sure, but that doesn't mean they have any theoretical basis; they are essentially completely arbitrary (even ignoring the fact that they measure only how someone does at ONLY MLG tourneys): there is no theoretical reason that X place is worth exactly Y points more than Z place.
If we try to invent a mathematical basis for our rankings, I guarantee it won't be much more accurate than MLG rankings.
You may ask about the points system we use (of giving 26-X points to Xth place, summing them, and ordering based on that), and the answer to that is that is the same as taking the average place that each person got on panelists lists. There is no subjectivity forced into objectivity there.
Let's just take the case of considering one person's win versus another. In order to have any idea of what this means, it is necessary to know what the skill gap is between the two players (by skill here I mean overall ability to win in any and all circumstances, essentially the thing that determines ranking). How can we hope to mathematically evaluate such a thing? Since no two people (let alone all panelists) will ever agree precisely how good someone is? How can you possibly convince someone else that player X is 10% better than player Y, or that beating player X is 20% harder than beating player Z, for example? "10% better" means nothing. Yet coming up with such a number is central to the idea of a mathematical formula for skill: If we are to compare the wins of 2 players mathematically, obviously we need each win to be assigned a numerical value.
To summarize this into one point, a math equation for rankings requires OBJECTIVE MEASURES (i.e. stats, like in baseball for example) that can be ordered from least to greatest. All we have in Smash that is objective to go on is a list of who beat whom (which in themselves have a HIGH degree of extenuating circumstances), but I maintain that it is impossible to come up with a meaningful way of extracting actual numbers out of them. And like in baseball, I think a poll is the best way of ranking players. Despite the enormous amount of stats that are collected in basball, in the end the MVP awards come down to a simple vote. I think we should follow this.
.
