I never said it should be based solely on tournament evidence; I said tournament evidence is required, or else it's pure speculation. If you can't demonstrate that what you're saying in theory can actually happen--and has been shown to happen--then your theory is bunk and has no place in a practical tournament setting.
That is, until it's been shown to happen.
And I'm not disagreeing with that.
What I disagree with is the assertion that it's pure speculation, what theoretical modeling does is it takes the technical data of all known moves and techniques and models them based on how a match develops at the top of the metagame.
It's easy to show that the technical data is correct, and all the moves are demonstratable. The only question left is whether the theoretical model is correct, and that's where the discussion comes in, and it's all grounded firmly in fact.
So, while it's not pure speculation, it is necessary, though not sufficent.
Everybody accepts that the roster is not 100% balanced--nobody's arguing that it's not (at least I'm not). In cases where both players are of equal skill and are both playing at the top of the metagame (a thoroughly impossible situation), character choice will decide the match. That's just the way it is.
Because of this, bans are sometimes necessary, but only in extreme circumstances; I.E., overcentralization.
If you can prove, via tournament data, that MK objectively overcentralizes the game to the point where he warrants a ban, then the pro-ban camp has a valid argument. Up until then, they have yet to come up with any sort of substance to support their side.
Theory is not everything; there has to be practical data involved.
Well, understand that using "the top of the metagame" as the standard for match-ups allows players to know the tools they have at that level even if they're not at that level, and select the tools they need to practice in order to win at the top of the metagame.
Regardless, that's why the ratio is so dependant on how much x character beats y characters at the top of the metagame, because that defines who whether the overcentralization threshold is achieved.
Yeah, the essay question was a looongggg time ago; and even before that they made you have an interview with senior debaters over AIM. You had to pick a topic, argue your view, and hope that your interviewer thought you were a good enough debater.
That was really old school
Way before my time I guess, I only came to smashboards a little while before Brawl was the major thing. Oddly enough, it was to ask a question about Sheik's upthrow, I was trying to figure out how to use it and was drawing a blank, I was mad noobish back then. The response I got was "to bait a double jump".
Maybe if you provided some sort of tangible example for this? Which is of course what I've been saying along. If you don't have real evidence, where is the practicality? You can't base a game around theory that hasn't been practiced. Hence, the importance of tournament data.
I.E., don't fix what's not broken.
I disagree, theory often tells us a great deal more and can often predict future trends. The fact is, our theory IS grounded in evidence, frame data and the like IS definately evidence.
The real practicality is that delving deeply into the theory often tells us things that we miss, like for example. As an example, a long time ago when Brawl was a lot younger, I looked at Snake's match-up thread and compared it to MK's and said, "MK is more bannable because MK beats most characters, and Snake beats MK, MK clears the field for Snake". That model is still in place for tournaments, even though MK usually wins, MK is still boosting MK, and I think it's pretty obvious which character is more bannable now right?
Same for saying that Snake didn't have the advantage over MK and neither did DK, I looked at the known technical details and made judgement calls about the characters, and tournament results have ultimately shown that my theory was correct.
Again, over-centralization of skill skews tournament results far too greatly to be a sole indicator, and they definately cannot be the primary indicator for something as precise as match-ups. For bans, they're more useful because the scale of testing is much larger, but they're still not the sole indicator.