Why do Some Argue Against Skill Based MM?
Дата: 30.07.2021 01:39:29
Mojo_Riesing, on Jul 29 2021 - 16:17, said: Look, i don't how much plainer i can be. It's very clear
this "debate" over skills, in some form, being used for
match-making is in dispute. Some, like you perhaps, use data
collected from the existing system of accepted practice to validate
you views. You feel that mostly things work well. Probably for you
personally as well.
Some are using the evidence of what they "see" in the moment repeatedly along with value judgements about what "fair" means (basically). The feel things arent' working well, and for them that has a personal impact as well.
"Your" side rejects value judgements and the "other side" isn't convinced by the mathematical projections they are given. So, stalemate, and it goes ever on. That's the key point here. Neither side is EVER going to convince the other. Just how it is when in essence they are like on different planes of existence never to meet.
Mojo_Riesing, on Jul 29 2021 - 16:17, said: Regarding scientific method, we have remarkably little of that in evidence. Just using math doesn't make it science or imply use of method. Things like testing (and re-testing), independent peer review, absence of personal benefit/belief are key parts of scholarly level science. I don't see a lot of that here, slapping a label on doesn't cut it. Whatever. I've said all along Neato has the numbers at his command and i'm impressed. He taints some of his work with his comments. Credibility suffers when you do that and people see that. Just how it works.
Read ANY scientific journal and you'll see much of what i mention. Review, testing, transparency, statements that address personal bias issues are what you'll find. None of that here and it's no wonder, it's just a game here.
I have seen enough that i think implementation of SBMM is just not worth all the fuss despite a suspicion somewhere in all this there is value in the concept. Being practical, pragmatic, i'm perfectly happy to push it off the table IF other elements of match-making could then be addressed. I've been clear about that too.
Some are using the evidence of what they "see" in the moment repeatedly along with value judgements about what "fair" means (basically). The feel things arent' working well, and for them that has a personal impact as well.
"Your" side rejects value judgements and the "other side" isn't convinced by the mathematical projections they are given. So, stalemate, and it goes ever on. That's the key point here. Neither side is EVER going to convince the other. Just how it is when in essence they are like on different planes of existence never to meet.
DeviouslyCursed: Ok, this is much better. I don't use data to determine
whether Random MM is "fair" or not, that's used mostly to go
against the "it's rigged" nutters. Fair is a subjective thing.
As I have pointed out multiple times, both Random MM and SBMM are
each fair in the own way, and also unfair in their own way. Someone
stating one is fair is an idiot if he doesn't expect the other side
to come back with the opposing statement. You won't get anywhere
because it is subjective. So because both are fair and also unfair,
what should be discussed is the positives and negatives each
side comes with, and whether it is worth changing the MM. This
is where I have an issue with the SBMM fans. Most of them think it
will fix steamroll/blowout/landslide battles. It will not. This is
where the data comes in to play. This is where science and
statistics matter, and "feels" should get thrown into the garbage.
I don't care how hard one thinks steamrolls are caused by skill
disparity, I don't care if they "feelz it in their bones" that it
is true. It simply is not true. No changes to MM will fix this
issue. Period. The other thing that is brought up is the skill
mismatch games. Here again the SBMM argue in a dishonest
fashion, calling them "certain wins/losses" when all data shows
that they are not. The underdogs win some of the fights. Quite a
few of them, actually. If you argue you want a challenge, nothing
is more challenging than getting a team that's supposed to lose a
victory. The SBMM fans then say that the matches are not fun.
This is Neato's specialty as he has the data, and he can correct me
if I'm wrong, but his data indicates the games basically play out
the same, regardless of the skill levels of the teams, whether
similar or mismatched. This is why he says the only way someone
knows if they are having fun or not is if xvm tells them it is okay
to have fun in this battle. You are basically telling us that
data isn't going to fix the issues these players have presented.
The problem is, SBMM won't fix it either. You are asking us to
address a situation that simply is not real. Reality doesn't care
if they think it's true. Reality doesn't care if they "know" it's
true. Reality cares about what actually is true. The Random MM is
not causing most of the issues they are complaining
about. The one valid argument against Random MM, is Win
Rate differences. Good players win more under Random MM. This is
true. But again, it is only a matter of opinion on whether there is
anything wrong with that. Once it is pointed out that good players
are not GIVEN and advantage, but create their own advantage by
playing well, the SBMM proponent then inevitably goes back to one
of the above arguments because they know stating a player doesn't
deserve to win if they are playing well is kind of a prick thing to
say. Also, they argue this despite the relatively tight win rate
spread this game has. SBMM will cause issues. As I have
pointed out, players will be more likely to play their best tanks,
or as that one purple player pointed out, that means an increase in
the most OP tanks in the game. SBMM will cause win rates to
tighten or go all the way to 50% (probably 49% if draws still count
as losses) SBMM ill cause WG to have to reevaluate incentives
for winning, because SBMM will allow you to literally do
nothing and still get approx 50% win rate. SBMM will
cause you to lose your ability to affect your win rate yourself. It
will be a meaningless stat.
Mojo_Riesing, on Jul 29 2021 - 16:17, said: Regarding scientific method, we have remarkably little of that in evidence. Just using math doesn't make it science or imply use of method. Things like testing (and re-testing), independent peer review, absence of personal benefit/belief are key parts of scholarly level science. I don't see a lot of that here, slapping a label on doesn't cut it. Whatever. I've said all along Neato has the numbers at his command and i'm impressed. He taints some of his work with his comments. Credibility suffers when you do that and people see that. Just how it works.
Read ANY scientific journal and you'll see much of what i mention. Review, testing, transparency, statements that address personal bias issues are what you'll find. None of that here and it's no wonder, it's just a game here.
I have seen enough that i think implementation of SBMM is just not worth all the fuss despite a suspicion somewhere in all this there is value in the concept. Being practical, pragmatic, i'm perfectly happy to push it off the table IF other elements of match-making could then be addressed. I've been clear about that too.
DeviouslyCursed: Still working on this, have to go back to
work....
Why do Some Argue Against Skill Based MM?