A Reconsideration of Stage List Philosophy. A Reimagining of the Stage Selection Process.

[BLVolition]

Chaper 1: "Starters" vs. "Counter-picks"

A Conundrum Explained

There are fruitful discussions going on about tournament viable stages in Super Smash Brothers Ultimate. These conversations have led me to the conclusion that the idea of “starters” and “counter-picks” is a social construct. This ideology seems to be gaining popularity but a concern of mine is people missing what is at the core of this school of thought. This text aims to hone in on that point.

Symmetry is inherently neutral. One cannot gain or lose anything as an effect of interfacing with identical aspects on any axis of something that is symmetrical. For example, any two points on the circumference of a perfect circle are equidistant from the center thus neither is at an advantage or disadvantage for their distance from it. There are many symmetrical stages in the Super Smash Brothers (SSB) series where for a time the prior-described analogy can be recreated. That is to say, symmetrical SSB stages have the theoretical capacity to be neutral. Nevertheless...

What causes a symmetrical SSB stage to lose its neutrality is the addition of an animate-asymmetrical-dichotomy. In other words, a “matchup.” Seeing as both a player and character matchup are needed to play a game of SSB we are definitively without a way to compete in true neutrality. Note, this is no fault of the stages themselves. Stages are merely vessels in which matchups can exist. That being said…

What is a “starter?”... Typically, a “starter” is a symmetrical stage that is perceived to be “more neutral than not.” Unironically, this is more or less a sound logical proposition. But…

What is a “counter-pick?”... The term undoubtedly implies strategic disparity. Colloquially, it is the term used to refer to “polarizing stages.” A stage that is asymmetrical is not inherently neutral***.*** Given that, there is merit to the idea of classifying stages with asymmetrical characteristics as a “counter-pick.” However, given the aforementioned, we know that even symmetrical stages pose a gain or loss of advantage. The significance of this is twofold. For one, the polarity of stages as a means of classification is baseless because each and every stage is indistinguishable in the existence of poles. Two, the hypothetical validity of said classification is still weakened by the fact that the amount of polarization that stages impose is not empirically measurable and/or is largely conjecture because of the necessary addition of matchups.

The only clear threshold is symmetry versus asymmetry. Even so, when a system ensures each combatant is given equal opportunity to play on a stage that is as neutral as possible the idea of a “counter-pick” is rendered moot. This holds true even if the selection pool contains both symmetrical and asymmetrical candidates. Therefore, divisions of any kind in the stage list cannot be wholly objective and are unnecessary. In addition, a decision to diverge from an equalizing system is abstract and likely out of sync with competitive integrity.

“What do you propose?”

1. A single odd numbered list with no distinctions between “starters” and “counter-picks.”
2. Stage Election

Chapter 2: Stage Election


Twitter: BLVolition
Reddit: BLVolition
Discord: BLVolition#7448

Stage Election
A new voting based method for picking stages in competitive Super Smash Brothers.

Process:

1. Rock, Paper, Scissors.
2. Winner votes for a stage they want to play on.
3. Loser agrees or denies. If the loser denies the stage is struck. Go to step 4.
4. Loser votes for a stage they want to play on.
5. Winner agrees or denies. If the winner denies the stage is struck.
6. Repeat steps 2-5 until there is one stage left. Play on the leftover stage

Advantages Over Traditional Striking:

1. Functionality only requires the stage list to be odd numbered.
2. Speeds up stage selection with larger lists by virtue of functioning on agreements.
3. Provides structure for lists void of the “starter” vs. “counter-pick” dynamic.

 

[Necro'lic]

While I definitely love the ideas presented, I think it'd be nice if in your conclusion, you succintly put in your idea of the DBV system and how it would work in the stage picking process. I know you technically do in the middle of the post, but it seems appropriate to put it at the end too for a tl;dr of sorts.

Otherwise, I can actually see this working. I can't believe we never thought of doing the opposite of strikes and doing a vote system. I had a similar idea, but it was definitely less complicated than this, but I also thought of it off the top of my head, so...

 

[BLVolition]

While I definitely love the ideas presented, I think it'd be nice if in your conclusion, you succintly put in your idea of the DBV system and how it would work in the stage picking process. I know you technically do in the middle of the post, but it seems appropriate to put it at the end too for a tl;dr of sorts.

Otherwise, I can actually see this working. I can't believe we never thought of doing the opposite of strikes and doing a vote system. I had a similar idea, but it was definitely less complicated than this, but I also thought of it off the top of my head, so...

I changed "In Conclusion" to "To Reiterate" and put the rules again at the end. That was a good suggestion, thank you! Also, thanks for reading!

 

[Amazing Ampharos]

I suggest a major amendment to make this work a lot better. Also give players the option to simply agree to the opponent's stage even if the vote isn't the same. That avoids the following major problem scenario I'll demonstrate with Smash 4 stages.

Let's say this is the stage list ranked from my favorite to least favorite legal stages:

Halberd
Delfino Plaza
Battlefield
Lylat Cruise
Town and City
Smashville
Duck Hunt
Umbra Clock Tower
Final Destination

For game one I suggest my favorite stage, Halberd. My opponent anticipates this and suggests Delfino Plaza which I would have suggested next. Under the rules as written here, this strikes my two favorite stages and is a major defeat for me in stage procedure. The strategy to avoid this involves a lot of guessing and mindgames that frankly don't belong in stage selection procedure. If I have the right to agree to my opponent's proposed stage, then I will simply agree to Delfino Plaza after it is suggested and we'll play game one there which means my opponent will only suggest if if they really want to play there. There is a bit of an agreement game of chicken effect (maybe both of you would agree to the other's stage and whoever agrees first "loses"), but that's way better than the alternative.

Overall I like the idea for game one. I think it's probably too tedious/long-winded for game two+, but I think the general idea is a clear optimization on striking and also helps in that it works with stage lists of size 2n + 1 instead of just 4n + 1 (I see no reason this procedure wouldn't work with a list of 3, 7, 11, 15...). You do need to have a phone or something to do constant double blind picks here (since you need to announce your votes simultaneously), but that's feasible.

 

[BLVolition]

I suggest a major amendment to make this work a lot better. Also give players the option to simply agree to the opponent's stage even if the vote isn't the same. That avoids the following major problem scenario I'll demonstrate with Smash 4 stages.

Let's say this is the stage list ranked from my favorite to least favorite legal stages:

Halberd
Delfino Plaza
Battlefield
Lylat Cruise
Town and City
Smashville
Duck Hunt
Umbra Clock Tower
Final Destination

For game one I suggest my favorite stage, Halberd. My opponent anticipates this and suggests Delfino Plaza which I would have suggested next. Under the rules as written here, this strikes my two favorite stages and is a major defeat for me in stage procedure. The strategy to avoid this involves a lot of guessing and mindgames that frankly don't belong in stage selection procedure. If I have the right to agree to my opponent's proposed stage, then I will simply agree to Delfino Plaza after it is suggested and we'll play game one there which means my opponent will only suggest if if they really want to play there. There is a bit of an agreement game of chicken effect (maybe both of you would agree to the other's stage and whoever agrees first "loses"), but that's way better than the alternative.

Overall I like the idea for game one. I think it's probably too tedious/long-winded for game two+, but I think the general idea is a clear optimization on striking and also helps in that it works with stage lists of size 2n + 1 instead of just 4n + 1 (I see no reason this procedure wouldn't work with a list of 3, 7, 11, 15...). You do need to have a phone or something to do constant double blind picks here (since you need to announce your votes simultaneously), but that's feasible.

That amendment is a good one (especially logistically) but is victim to the Influence of Sequentiality. An order of who would decide to agree, disagree, or vote again would need be implemented I think. Also, requiring the presence of phones or pen and pad to vote history and bans/DSR gives each player a empirical data to look at to help them make quicker more informed decisions which kind of counteracts its self as disadvantage.

As far as Game 2+ the current system in place isn't bad by any means but it's definitely not objective. Still, a marriage of the two is certainly a better solution then the traditional "starters vs. counter-picks."

 

[Ryu_Ken]

I assumed I understood what "starters" and "counterpicks" were and how/why we use and pick them, but I've come to realize that they're not really as black & white as we think. The logic behind some stages being starter/counterpick can be inconsistent and contradictory (like how Dream Land 64 and Pokemon Stadium 1 have switched back and forth between being either starters or counterpicks throughout Smash entries, at least from tournaments I've participated in or seen). I've never understood why the "counterpick" stages had to be reserved for later matches in a set instead of being available from the get-go.
Though I haven't fully grasp every single term used, this proposed method makes more sense to me than the starters/counterpick method. Let's see if I'm understanding this right...

Spoiler: Proposal in Practice

Two players meet up for their set, which is a Best of 3. This is the legal stage list available to them:

1. Halberd
2. Delfino Plaza
3. Battlefield
4. Lylat Cruise
5. Town and City
6. Smashville
7. Duck Hunt
8. Umbra Clock Tower
9. Final Destination

Before them are two cups, each filled with paper slips with numbers on them. They both draw from their cups, not looking into them. Player 1 holds "4." (Lylat Cruise) and Player 2 holds "6." (Smashville). These stages are striked, and the players draw again. This time, both of them draw "7." (Duck Hunt), and this is the stage they'll play on for Match One.

Match One ends with P1 as the winner, and their stage list now looks like this:

1. Halberd
2. Delfino Plaza
3. Battlefield
4. Lylat Cruise
5. Town and City
6. Smashville
7. Duck Hunt
8. Umbra Clock Tower
9. Final Destination

At this point, paper "3." is taken out from both cups. P1 chooses to ban Smashville, so now paper "6." is taken out of both cups. The stage list now looks like this:

1. Halberd
2. Delfino Plaza
3. Battlefield
4. Lylat Cruise
5. Town and City
6. Smashville
7. Duck Hunt
8. Umbra Clock Tower
9. Final Destination

P1 and P2 at this point go through the same process of drawing the paper slips out and playing through each match until one of them is declared the winner of the set.

Side note: this is just an example, but I hope that if TOs decide to run a tournament with this proposal, they make this a bit more streamlined (i.e. RNG app or website instead of drawing loose material from a cup)

I feel like I'm rambling a bit, but I just wanted to bump this thread up so others can see it.

 

[Necro'lic]

I read through the reddit post you did for this, and I'm starting to think you're gonna need some more distinction between what happens each round. For example, I'm not entirely sure about your idea of having the winner ban a stage during round 2, since round 2 is usually supposed to be in the loser's favor. Also, as you say in some of those replies, the smaller the stage list, the better it gets for both players to try and game the system, making anything past game 2 a bit weird.

 

[BLVolition]

I read through the reddit post you did for this, and I'm starting to think you're gonna need some more distinction between what happens each round. For example, I'm not entirely sure about your idea of having the winner ban a stage during round 2, since round 2 is usually supposed to be in the loser's favor. Also, as you say in some of those replies, the smaller the stage list, the better it gets for both players to try and game the system, making anything past game 2 a bit weird.

These are all valid points and I'm trying to find the workarounds. Are you on Discord by any chance?

 

[Necro'lic]

These are all valid points and I'm trying to find the workarounds. Are you on Discord by any chance?

I am, but it's late where I am, so... wanna do this tomorrow night? I am hopefully available.

 

[BLVolition]

I am, but it's late where I am, so... wanna do this tomorrow night? I am hopefully available.

Np. Message me anytime, @BLVolition#7448.

 

[Amazing Ampharos]

Yeah, you need some primacy for who has to agree first probably (unless you just want a game of chicken which is probably less bad than it sounds), but I do think the amendment is necessary to avoid various exploitation of strategic voting. I pointed out the strategy of voting for stages you know your opponent likes but that you don't anticipate them voting for. There's also the strategy of naming your "middle" stage first and, if your opponent doesn't also name that stage up front, just escalating your stages toward your most favored ones which if your opponent lists their favorite stages with opposite preference to you will end with you on your best stage. So yeah, the agreement thing I think is necessary.

Game 2+ why not just simplify to something more basic? A basic loser picks 3, winner picks 1 from those is functional, fast, requires no memory of past events, and should mostly produce a desirable stage.

 

[BLVolition]

Yeah, you need some primacy for who has to agree first probably (unless you just want a game of chicken which is probably less bad than it sounds), but I do think the amendment is necessary to avoid various exploitation of strategic voting. I pointed out the strategy of voting for stages you know your opponent likes but that you don't anticipate them voting for. There's also the strategy of naming your "middle" stage first and, if your opponent doesn't also name that stage up front, just escalating your stages toward your most favored ones which if your opponent lists their favorite stages with opposite preference to you will end with you on your best stage. So yeah, the agreement thing I think is necessary.

Game 2+ why not just simplify to something more basic? A basic loser picks 3, winner picks 1 from those is functional, fast, requires no memory of past events, and should mostly produce a desirable stage.

All things I hadn't considered when I was initially writing this. On top of that, someone on Reddit seems to have proved the agreement amendment just provides for a different type of strategic voting exploitation.

All in all, I'm glad I put it out there and got some feedback but I think it might be back to the drawing board.

 

[Amazing Ampharos]

Strategic voting will always be a thing. All stage selection procedures are gamed, and striking is really no different (here's a strategy: form a list of all stages you intend to strike and strike in reverse order of your perception of a typical opponent's preference for the stages, striking Smashville early and Lylat late for instance: this maximizes your expected value from strikes versus a naive sequencing). The question is what is the meta if you have agreement?

In general I will have a list of preferences 1-N with N being the stage list size. In general all stages can be broken into three categories for me: stages I consider favorable, stages I consider unfavorable, and the middle stage. In theory my safest play is to always name stages from my favorable side (fishing for whatever I perceive my opponent is likely to agree to hoping for an agreement on an advantage for me) and if my opponent suggests anything that was not on my unfavorable list I should agree to it. This strategy guarantees I will never play game one on an unfavorable stage and gives me a pretty decent shot at playing on a stage somewhere on the favorable spectrum (I only get the middle stage if my opponent names it or if we make it to the very end of the procedure and it's the remaining stage). This is the safe strategy.

There is always the strategy of exploiting asymmetry of information. If I name the middle stage first, I run serious risk of it being agreed to prematurely (eliminating my opportunity for a favorable stage), but if my opponent rejects it, then I can begin scaling up through escalating stages on my favorable list. A disagreeable opponent will eventually find themselves pinned into my absolute favorite stages. This seems to be a problem.

However, is it really? Assume your opponent follows the same rule you do of "always agree to a stage that is middle or to your favor". If their list is fundamentally different from yours, this may cause them to reject the so-called middle stage, believing it to be in your favor, and wait for a stage you prefer more but that they also prefer more. This is actually a MORE optimal outcome as both players believe they have gotten stages biased in their favor which is actually more fair than neutral in theory; we can just let the gameplay decide who was right as is proper in competitive gaming. In fact, the strategy of "always agree to any stage not on your unfavorable list" is absolutely guaranteed to give a fair outcome to both players.

The checkmate scenarios only happen if you deviate from the strategy of "always agree to the middle stage or any stage to your favor". If you get the middle stage or a stage only slightly in your favor offered early, you may perceive that you can reject it and get a better stage by continuing the procedure which maybe happens if your list and your opponent's list are sufficiently different, but the risk is that your opponent blocks this and you get corralled into an unfavorable stage. However, this is itself a form of justice; if you stubbornly reject a stage you believe to be fair hoping for something more stacked in your favor, is it not fair that it can blow up in your face and you can be forced to something less fair? The meta really is strongly in favor of playing it safe here, and as an added bonus, the gameplay is actually emergent from the "Smashville?" meta because that can just be a first offer in this system. And yes, you can offer in a 1-2-2-2-2...-1 sequence and literally approximate striking except with the chance to cut it way short. If your list is 4n - 1 instead of 4n + 1 that sequencing will be imbalanced but won't change the fundamental guarantee of the system (which can also be rebalanced by making the sequence 2-3-2-2-2-2...-2-2-1).

What's the problem?

 

[BLVolition]

Strategic voting will always be a thing. All stage selection procedures are gamed, and striking is really no different (here's a strategy: form a list of all stages you intend to strike and strike in reverse order of your perception of a typical opponent's preference for the stages, striking Smashville early and Lylat late for instance: this maximizes your expected value from strikes versus a naive sequencing). The question is what is the meta if you have agreement?

In general I will have a list of preferences 1-N with N being the stage list size. In general all stages can be broken into three categories for me: stages I consider favorable, stages I consider unfavorable, and the middle stage. In theory my safest play is to always name stages from my favorable side (fishing for whatever I perceive my opponent is likely to agree to hoping for an agreement on an advantage for me) and if my opponent suggests anything that was not on my unfavorable list I should agree to it. This strategy guarantees I will never play game one on an unfavorable stage and gives me a pretty decent shot at playing on a stage somewhere on the favorable spectrum (I only get the middle stage if my opponent names it or if we make it to the very end of the procedure and it's the remaining stage). This is the safe strategy.

There is always the strategy of exploiting asymmetry of information. If I name the middle stage first, I run serious risk of it being agreed to prematurely (eliminating my opportunity for a favorable stage), but if my opponent rejects it, then I can begin scaling up through escalating stages on my favorable list. A disagreeable opponent will eventually find themselves pinned into my absolute favorite stages. This seems to be a problem.

However, is it really? Assume your opponent follows the same rule you do of "always agree to a stage that is middle or to your favor". If their list is fundamentally different from yours, this may cause them to reject the so-called middle stage, believing it to be in your favor, and wait for a stage you prefer more but that they also prefer more. This is actually a MORE optimal outcome as both players believe they have gotten stages biased in their favor which is actually more fair than neutral in theory; we can just let the gameplay decide who was right as is proper in competitive gaming. In fact, the strategy of "always agree to any stage not on your unfavorable list" is absolutely guaranteed to give a fair outcome to both players.

The checkmate scenarios only happen if you deviate from the strategy of "always agree to the middle stage or any stage to your favor". If you get the middle stage or a stage only slightly in your favor offered early, you may perceive that you can reject it and get a better stage by continuing the procedure which maybe happens if your list and your opponent's list are sufficiently different, but the risk is that your opponent blocks this and you get corralled into an unfavorable stage. However, this is itself a form of justice; if you stubbornly reject a stage you believe to be fair hoping for something more stacked in your favor, is it not fair that it can blow up in your face and you can be forced to something less fair? The meta really is strongly in favor of playing it safe here, and as an added bonus, the gameplay is actually emergent from the "Smashville?" meta because that can just be a first offer in this system. And yes, you can offer in a 1-2-2-2-2...-1 sequence and literally approximate striking except with the chance to cut it way short. If your list is 4n - 1 instead of 4n + 1 that sequencing will be imbalanced but won't change the fundamental guarantee of the system (which can also be rebalanced by making the sequence 2-3-2-2-2-2...-2-2-1).

What's the problem?

This is how I first perceived things playing out and if the amendment actually allows for this sequence of events then bingo. Correct me if I'm wrong but in light of the amendment there isn't even a need for the votes to be double-blind, no?

 

[Amazing Ampharos]

This is how I first perceived things playing out and if the amendment actually allows for this sequence of events then bingo. Correct me if I'm wrong but in light of the amendment there isn't even a need for the votes to be double-blind, no?

Yes, I made that point above. This is actually pretty much just inverted striking at that point. It's not that different from other ideas that have been given before involving proposing stages for agreement but has endgames built in if no one agrees and, if both lists are identical, allows either player to instantly shortcut the process by suggesting the middle stage.

 

[BLVolition]

https://twitter.com/BLVolition/status/1062886737825931264

 

[Necro'lic]

lol, I forgot about the discord thing and now I'm sleepy again. Oops XD

 

[infomon]

Logistically, how would you suggest people do "blind voting"? We're already terrible at even doing blind-picks for characters going into game 1.