It does not require evidence to be able to identify that other ban criteria can be acceptable.
What happens if you shift the required matchup ratios by 5 against the character to be banned (So they'll get banned for 85:15 rather than 90:10)? Right, pretty much nothing. No new characters become banned, no banned characters become unbanned. But it's a different ban criteria. So is it not acceptable? Could it even be better, in a game that's not released or not been taken competitive yet? Edit: How can it not be acceptable, when the current results are exactly the same with either criteria?
The values are arbitrary, there simply aren't that many games available to apply them to with enough borderline characters to see where the line for them needs to be drawn. That of course assumes you can accurately enough judge the matchups to even have those numbers have relevance, but if you can't do that you can't justify a ban using that criteria anyway.
Actually, a great deal, because this criteria is applicable to tactics as well. While the number of characters might be quite few, the number of TACTICS is certainly enough. Furthermore, the difference is easily discernable, given a large enough number of games.
Again, the clear distinction here is that banning is a last resort, whereas your suggestion moves it considerably closer to being a first option.
Especially because it deals with cast size inconsistently, for example, you'd be hard pressed to find a game with a cast of 12 that doesn't fit your new criteria.
And actually, it's quite easy for something to be unacceptable when the results are the same. What if, for example, if I was asked to compute 2+2, and I use the mechanism for 2*2? I still get 4, don't I? But isn't my methodology wrong?
Oh, you could always say that if I use different values the issue will be revealed, but that just brings up the core point here, the problem of induction. Basically, you can never be sure you're not correct about something if you are basing your understanding on "it usually works", if you don't have some sort of deductive background to derive your standard from, you can never be sure that you won't come across a case where you have that one case.
Which is why we have a mechanism known as "falsifying the null hypothesis", where we take the null and subject it to continuous testing, if the testing disproves it, we discard it for the new null.
Ultimately, in Sirlin's philosophy there are both elements of placing the null and deduction. The actual percentage is changeable (and is a null hypothesis for our use), but it is drawing a line based on deduction.
So, again, ultimately, like all null hypothesizes, the burden is on the person intending to make the change.
Exactly. It's in the dictionary, but that doesn't mean it's current.
Have you never heard the phrase "They may be with me, but they aren't my friends," used as a common response to that definition of friend? The "friendly" definition of friend trumps "associate" when you refer to a group of people as "your friends," as you did with salaboB.
Actually, quite the opposite, your view is outdated.
The weakening of the word "friend" has been occurring for quite some time now, and is a known psycholinguistic fact. Part of it has to do with social networking (facebook especially plays a role, along with other sites that refer to all contacts as "friends"), but there's a general move towards expanding friend to mean well beyond what "friend" used to mean, and this fact is expounded in the dictionary.
However, I will note that Merriam-Webster is updated on a consistent basis to reflect changes in language, which is why you'll find "google" and other similar terms there.
Generally speaking, a group of people who you have close personal relationships with are "close friends", whereas "friends" can include a wide variety of relationships beyond that scope.
So yes, your views on what the word "friend" means are a bit old-fashioned.