The so-called neutral stages have never been "neutral", and the results in Brawl especially were often really lousy with small starter lists. I main Mr. Game & Watch. He's all around a very good character, but he's nearly unplayably bad on Final Destination and Smashville is just awful for him too. Both of those stages turn the otherwise winnable but hard ICs/MK/Olimar match-ups into hard counters (like G&W vs ICs is 1-9 at best on FD; you might as well give up and not play), and they make every other match-up worse too (I don't think he actually has an advantage against a single other viable character on those two stages). No matter what, I always have to strike both of these stages on any starter list which means 3 stage lists that include both are just automatically awful outcomes for G&W. Even on five, if both are included which is always true, I get my worst stage out of any remaining stages (usually mostly mediocre G&W stages among the others even if not as just awful as these two). On the other hand, G&W just loves Halberd and rather likes Delfino too; having them in the starter list lets him get stages that are actually decent for him as his opponents strike those to offset the strikes he absolutely must make to FD and SV. Sometimes you get an MK player who really likes Delfino and we play game one there, and when that happens, we both end up feeling really good about the game one stage since we both believe it gives us an advantage (IMO vs MK Delfino is a very soft G&W good stage, perfectly worth it for the MK player if he as a player is super comfortable with the stage). I just don't see at all how banning so many stages and having so few starters did any good in Brawl; it completely screwed my character, and as far as I can tell, it made the game's overall balance a lot worse in general as characters like ICs and Olimar ended up artificially just way too good in way more match-ups than vs G&W.
I believe we see the same sort of thing in smash 4 too. Final Destination is easy to pick at in this game; it's a near-perfect Little Mac stage that gives aerial characters nowhere to go but down into his super armored arms, but on "not FD", so many characters can deal with Mac's unique mix of pros and cons so much more effectively. Sonic too is really crazy on FD; Hammer Spin Dash is still not on a lot of people's radar but one of the most powerful mobility and attacking options in this game that is a million times better on FD than on "any stage with platforms". Battlefield isn't perfect either. Bowser I feel is mostly really good in this game, but BF's platform layout is perfectly wrong for him. They just get in the way of a lot of movement he wants to do, nearly completely take Bowser Bomb and dair out of his moveset, and his size makes it nearly impossible for him to use them defensively. Greninja too I feel just struggles with Battlefield specifically; the platforms are just at this perfectly wrong height and spaced just perfectly wrong relative to the open ground so it's really hard for him to use any of his ground moves from below or poke at the platforms with any aerials. In smash 3ds, just add up the lists of characters who either win or lose uniquely much from stages like FD or BF, and then compare it to a similar list for Prism Tower. I think you'll find that the Prism Tower list is just as short as the FD or BF list; stages like that are every bit as "neutral" as any others. Even more, since every stage will have significant flaws and match-ups in which it is a particularly unfair stage, including a ton of different stages and notably different types of stages (you NEED to include both static and traveling stages) will really smooth it out. Smash 3ds really did have a shortfall of quality stages which made this hard (still possible), but we won't have that problem on Wii U which is the version that really matters anyway.
On top of the fact that any short starter list will invariably produce massive winners and losers in terms of characters, starter-counterpick dichotomy just doesn't work from a meta perspective. Real players only practice so much, and they formulate different strategies to win on every stage. This is actually a lot of work, but real players know something obvious. To win a set, you need to win 2/3 games. Since your opponent's counterpick is your biggest disadvantage, you focus on winning on your own counterpick and the game 1 stage. The game one stage is always going to be from the starter list. Your best winning play is to pick a character whose natural best stages are starters and to counterpick those starters with your counterpicks. You have to learn minimal stages, and you'll get maximal results. Learning counterpick stages is a lot of work, and no matter how good you are at it, it will never win you sets as you still have to win game one or a counterpick match on a starter your opponent picked. It's just a sucker's play to bother with that at all, and of course when no one learns these stages (except in my beloved if naive Midwest which never seems to shy away from learning tons of new stuff whether it really helps us or not), the matches on these stages will be awful since these stages tend to be complex and require tons of knowledge to play well and you're inputting players with low stage knowledge. If the matches on these stages are awful and it's just annoying anyway when "that one guy" doesn't understand the strategy to win and picks the outlandish cp stages anyway, people eventually just want to ban the cp stages altogether so a small starter list will always lead to an eventual small total stage list. Why would we design this implication into our stage rules?
This is already a long post in which I've pretty much just laid out why the starter-counterpick dichotomy is bad and why a short list of allegedly "neutral" stages will always be anything but. At this point I've asserted we need a ton of starters and no counterpicks. Our problem is that we have three principles that all tend to lead to the best game but run contrary to each other:
-Stages should be selected quickly.
-The chosen stage should be fair as often as possible.
-The game should have as much variety as possible.
Sacrificing any one point makes things fairly easy. If you use a small stage list, you directly move a lot of variety by removing stages and effectively remove more by making fewer characters viable. If you don't care about the fairness of the stage you pick, you can just random a stage which is super fast and super high variety. If you don't care about time, you can full list stage strike from 21 stages to get a super fair outcome that will encapsulate a mountain of variety. Here's what I can glean from this:
Stage striking as a procedure is not practical past 13 stages. 13 is a very good number (balanced striking, the only other number that works as well is 9), and no matter what happens, this makes 13 a good floor for legal stages. We need to adamantly insist on no stage lists for smash wii u with under 13 total legal stages and for all 13 at such events to be starters. However, we're going to have more than 13 good stages, and we need to figure out how to make them all game one legal. If we get precisely 18 or 26 good stages, we have one easy bisection option of just making even numbered rounds use one starter list while odd numbered rounds use the exact opposite list. That too seems unlikely; we're betting on lucky numbers. I think all things being considered a pretty good solution would be this:
For locals, you just randomize the starter list per event from the pool of good stages. To guarantee no super lopsided results, divide the legal stages between "static" and "dynamic" and preserve the ratio. If your legal stage list contains 12 mostly static stages and 8 dynamic ones, your randomized starter list will be guaranteed to contain 8 static stages and 5 dynamic ones. Players can always counterpick any of your legal stage pool, but the stages that don't get picked for a specific event will be counterpick only that day (and yes, that means that some days Final Destination or Battlefield will not be starters; it's okay). For larger scale events, vary the stage list per round, re-randomizing every time but always including every stage that was excluded from the previous round in the current round and maybe having a similar rule against a stage being allowed to be on the list too many rounds in a row. Do this before the event so players can see in advance what they'll be facing. This might be hard to coordinate if you run your event really loosely and have a ton of different rounds playing at the same time, but that's bad TOing in a few ways so just don't do that and it will be easy to communicate the stage list per round.
That solution isn't perfect, but I can't come up with anything better. I'd love to full list strike from 21 stages or whatever, but it would really slow down tournaments and prove a huge barrier of entry to newer players. I'm just plain not willing to ban stages over that kind of procedural concern though; I think a compromise something like this is going to be our best play.
