Determining the procedure to pick stages in Smash 4

[popsofctown]

The rules for Dominion say the loser should get to go first, which is already a huge advantage. Player one has like a 60% winrate in competitive play (maybe it's 55, I don't remember, it's a pretty big deal though).
There was some brief popularity for a variant that presented 12 cards and let each player player ban 1. The primary purpose of that was to remove cards people find unfun or uncompetitive without them being banned specifically.

Some cards have a lot of consensus on being poorly designed and unfit for play, like Rebuild, mostly Ill-Gotten Gains, and Familiar. Others have some controversy surrounding their actual impact on skill vs. luck tradeoffs, like Possession, Tournament, and King's Court. These controversies seem to be fairly consistent amongs broad levels of skill, for the most part people don't get better and learn to deal with it and change their mind, mid level and high level players will be pretty evenly split on things. (Low level players occasionally have a Pirate Ship is OP phase but it should be a pretty short phase)

 

[popsofctown]

@unit: that moves the centralization to game 3, which is still a problem, but admittedly an improvement because as you pointed out game 3 would at least be the one most informed by players learning eachother's habits, instead of the one -least- informed.

I do like the idea of FLSS all three games.

 

[LiteralGrill]

FLSS for all three games would be amazing, but insanely unrealistic for time obviously.

The idea to allow two counterpicks then the final stage honestly isn't too bad. It gives both players a chance to learn about each others playstyle before the final match. I guess it would just possibly give a psychological advantage to the person who got their counterpick on game one, would that be an issue?

 

[popsofctown]

Hence you should re-use the stages that were almost used during FLSS, that's my position.

Psychological advantage might not have much skew, but would be way unfun

 

[ParanoidDrone]

Another idea:

Do FLSS or SSS or whatever for stage selection, but after each game in a set, the stage chosen is automatically banned for all subsequent games in the set. In effect it's modifying Dave's Stupid Rule from "no player may select a stage they have previously won on" to "no stage may be selected twice in a set."

IDK I'm throwing ideas at the wall and seeing what sticks.

 

[LiteralGrill]

Another idea:

Do FLSS or SSS or whatever for stage selection, but after each game in a set, the stage chosen is automatically banned for all subsequent games in the set. In effect it's modifying Dave's Stupid Rule from "no player may select a stage they have previously won on" to "no stage may be selected twice in a set."

FLSS would be the better model, I only made SSS to try and help with time. If we could make the community accept FLSS it'd be all good.

 

[Davis-Lightheart]

I know people think deadlocking is unfun just because it is a tie that leads to an RNG of 2 stage, but if the two stages are of equal length to each other, psychologically they are almost even stages in the players' sense of fairness.

 

[popsofctown]

Another idea:

Do FLSS or SSS or whatever for stage selection, but after each game in a set, the stage chosen is automatically banned for all subsequent games in the set. In effect it's modifying Dave's Stupid Rule from "no player may select a stage they have previously won on" to "no stage may be selected twice in a set."

IDK I'm throwing ideas at the wall and seeing what sticks.

That makes the number of remaining stages even. How should that be resolved?

 

[ParanoidDrone]

That makes the number of remaining stages even. How should that be resolved?

Have it count as one of the strikes for the winner (if you want to approximate DSR) or loser (for...some reason?) is all I can think of. Like I said, throwing ideas at the wall and seeing what sticks.

 

[Uniit]

For me, a little bit of RNG is just fine, the problem with tie in the SSS is time related.

Worst use case here : after the double blind and SSS on paper, a tie. Just for that, since you were on the stage selection screen, B to get back to to the CSS, edit rules, more rules, random stage switch, deselect all, select the two, and B B B Start then start.

Not that i'm lazy (well, maybe i'm), but that sounds boring. Hum, after typing this, just found out that can be decided this a coin flip... So it can be fast too.

I can to the conclusion that every option that we got has drawbacks :
- Current neutral/cp : not fair if you loose the first match, striking without cp can advantage charaters who are good on most neutrals.
- FLSS/cp : not fair if you loose the first match, and not time efficient.
- FLSS all : clearly not time efficient.
- FLSSfirst then play "seconds choices" : got some problem if you want to switch character, or needs to select the stage before characters of the first match, a player can play a char, FLSS thinking of another char, then play the 2nd and 3rd with stage advantage.
- Reverse FLSS/cp : feels inintuitive, can advange the first player to counterpick, can be tricky for BO5/7, need to focus on learning counterpicks(?).
- SSS : needs confirmation, but isn't time efficient when ties involved, need to know all stages matchups(?).

Personally, i'm fine with all that is pretty simple, like our current system, (R)FLSS/cp or FLSS all.

 

[Piford]

On the idea of making a program to aid in stage striking, I can probably cook up something for computers but not smart phones.

 

[Doval]

Regarding game 1 centralization: Suppose that the matchup of the first round is balanced. Then the most fair outcome is to repeat the matchup.

Suppose the matchup of the first round is unbalanced. Since the characters can be picked blindly, a more skilled player is disadvantaged through no fault of his own. Repeating the matchup is bad.

One potential solution to the first case is to allow the loser to lock the matchup so the rest of the rounds are played with the same characters and stage. This never harms the winner when the matchup is even or in his favor, but there's the possibility that the matchup favored the loser and he threw game 1 on purpose to lock the opponent into the match, then goes on to win games 2 and 3. The decision to lock the matchup must be unanimous.

For the case that the game 1 matchup isn't fair, the problem is that the disadvantaged player must be compensated on the next game, but the third round must be neutral.

I propose that the loser of game 2 be given the option to either change characters or stage, but not both. He makes his choice freely, only limited by bans and DSR. Then, for game 3, the players bargain. The new loser once again decides whether he's counterpicking characters or stage. If he opts to counterpick the stage, he gives the opponent a choice of 3 stages (subject to DSR and bans). If he opts to counterpick characters, he does so and then the opponent offers him a choice of 3 stages.

 

[Piford]

Regarding game 1 centralization: Suppose that the matchup of the first round is balanced. Then the most fair outcome is to repeat the matchup.

Suppose the matchup of the first round is unbalanced. Since the characters can be picked blindly, a more skilled player is disadvantaged through no fault of his own. Repeating the matchup is bad.

One potential solution to the first case is to allow the loser to lock the matchup so the rest of the rounds are played with the same characters and stage. This never harms the winner when the matchup is even or in his favor, but there's the possibility that the matchup favored the loser and he threw game 1 on purpose to lock the opponent into the match, then goes on to win games 2 and 3. The decision to lock the matchup must be unanimous.

For the case that the game 1 matchup isn't fair, the problem is that the disadvantaged player must be compensated on the next game, but the third round must be neutral.

I propose that the loser of game 2 be given the option to either change characters or stage, but not both. He makes his choice freely, only limited by bans and DSR. Then, for game 3, the players bargain. The new loser once again decides whether he's counterpicking characters or stage. If he opts to counterpick the stage, he gives the opponent a choice of 3 stages (subject to DSR and bans). If he opts to counterpick characters, he does so and then the opponent offers him a choice of 3 stages.

When you mean bans do you mean tournament bans or player bans? Like I can ban 3 stages from the list and then my opponent chooses three?

 

[popsofctown]

Since the characters can be picked blindly, a more skilled player is disadvantaged through no fault of his own. Repeating the matchup is bad.

Sounds like blind pick is part of the issue..

 

[Doval]

When you mean bans do you mean tournament bans or player bans? Like I can ban 3 stages from the list and then my opponent chooses three?

Yup.

Sounds like blind pick is part of the issue..

Looking back at the posts I missed, I suppose you could use @ Uniit 's idea of simply picking one player that gets counterpick privileges on round one. The third round shouldn't be double blind though in that case though.

 

[Piford]

Yup.

That could work, but you would need to make sure you aren't giving the winner so much influence when choosing a stage that its basically the winner choosing not the looser.

 

[Doval]

That could work, but you would need to make sure you aren't giving the winner so much influence when choosing a stage that its basically the winner choosing not the looser.

Right. It's pretty easy to tune that through the number of bans and choices.

 

[Pazx]

You know, there's an easy way to solve the problems of large stagelist picking in taking up time or confusion of procedures that doesn't involve taping the procedures on the TV. You give them a visual means of stage striking. In Project M, this was solved by giving striking functionality to the stage select menu. For online, players just need to open a separate window on their internet browser since they're already on the computer communicating with each other.

So what about live tournaments? Make a phone app. Because of the ubiquity of smartphones, you just need to create an app in which you can decide the stage striking procedures and the legal stages available. It also means you only need one phone between two people for this to work, so TOs would be able to easily help those who don't have smartphones or just forgot theirs.

go into random stage select

turn on all legal stages

strike by turning off stages

there is no need for a phone app or paper or playing cards

/argument

 

[popsofctown]

go into random stage select

turn on all legal stages

strike by turning off stages

there is no need for a phone app or paper or playing cards

/argument

DeLux pointed out players have trouble remembering whose turn it is while they deliberate. Passing a phone back and forth and having the app track whose turn is whose would help with that.

 

[DeLux]

If we did matchslips (for other data purposes) via phone app, we could get even more analytics for other things as well.

But until someone makes said phone app and it becomes a standard to the point people have it ready prior to the event, using the built in-game method is still the logistically best solution.


I want to address the centralization of game one issue -

From a mathematical standpoint, assuming stage striking from the legal list and blind picking of characters, players have a 50/50 chance of winning the set before we start factoring in their decisions. Then they select characters which shift the chances of winning away from 50/50 unless the play the ditto. Stage 1 of game 1 shouldn't necessarily try to make the matchup between characters 50/50, but should present the median MU ratio from across the legal stagelist, which from a preference on the competitive level is the "neutral" stage for the matchup, despite the matchup not being neutral. The point isn't to pick a stage that makes the matchup between characters even. Some characters just naturally beat other characters, regardless of the stage. Stage striking with blind picks yields a theoretically even matchup before we start factoring in players decision (such as which character to pick and which stages to strike) meaning it exists in the context of player vs. player. So if the blind pick yields a bad matchup, I'd assert it's because you made a bad character choice which is a reflection on character selection skill, so there is fault to be had there.

Once someone wins, this is how the math breaks down -

Assuming each game for 2 and 3 is 50/50, the winner of game 1 would have a 75% chance of winning the set while the loser will have a 25%. Let's assume this is the baseline

If the matchup ratios are skewed to 60/40 favoring the counterpickers for games 2 and 3 because of stages, the math leads to the winner of game 1 having a 76% chance of winning the set against the loser having a 24%. The odds on each game shifted 10% while yielding a 1% difference in set probability. This isn't much change, so there's credence to having a system where DSR exists and the winner would have to burn a strike on the stage that has already been played on, while maintaining blind picks through games two and three.

If the matchup ratios are skewed to 70/30 favoring the counterpickers for game 2 and 3 because of say stages and character, then the winner of game 1 have a 79% chance of winning against a 21% chance of the loser winning the set. As you can see, the odds shifted another 10% only this time it yielded a 3% difference in set probability. That leaves us with 4% for a total 20% game difference.

If the matchup ratios are skewed to 80/20 favoring the counterpickers, you are now left with 84% chance of winning against a 16% chance of winning after the results of game 1. This is a 5% change on the chance of winning the set from the previous 10% change of chances of winning each game. It is also 9% to 30% change in chances of winning the set and game respectively.

If the matchup is skewed to 90/10 for counterpicking, you're left with 91% against 9%, 7% change for the last 10%, overall 16% change for 40% overall.

Finally let's say logical extreme 100/0 for counterpicking, 100% against 0% chance of winning the set, 9% change for the last 10%, overall 25% change for 50% overall. The notable thing is that you've hit the point in this case where the 10% of change to the specific game odds in counterpicking was larger than the impact it had on the set (10% vs 9%).


Looking at the system outside the context of numbers, the point of counterpicking is that it allows for a greater variety of stages / characters because the fear of blind pick forcing people to pick generic "best case" options isn't there anymore for the losing party, as well as different stages favoring different characters inherently across the spectrum. As we can see, from a nominal standpoint, the biggest gains in diversity (as reflected by the games favoring the counterpicker) against the least impact on the overall set odds would occur by forcing players to play on stages closest to the struck to neutral that isn't the neutral for the counter picking phase.

However, we see going as far as altering matchups to the 70/30 ratio doesn't have quite a large impact on altering the set data. The fear of having things close to neutral but not being neutral aka near 50.000000001/49.999999999 would be that players would not have incentive to change characters thus creating less depth. That's why I advocate the tactial use of multiple bans (2 stage bans in 9 stage list, 2-4 stage bans in 13, 4-6 bans in 17 stages, etc.) because it balances the underwriting implications of the counterpicking system without centralizing game 1 as a prerequisite to winning the set. It also leaves you with five stages to play on across a five game set, again for depth purposes.

 

[popsofctown]

IfThe fear of having things close to neutral but not being neutral aka near 50.000000001/49.999999999 would be that players would not have incentive to change characters thus creating less depth.

There's a lot in here, but incentivizing players to change characters is a difficult foundation to work with when it's not a universally held value. I like the idea of a metagame where players are required to change characters every game by rule, because I think that variety is more fun, but there's tons of players that think you should never force a player to change characters and there's even players that think polarizing stages should be banned based on a -right- to never get pushed off your main.

 

[Doval]

From a mathematical standpoint, assuming stage striking from the legal list and blind picking of characters, players have a 50/50 chance of winning the set before we start factoring in their decisions. Then they select characters which shift the chances of winning away from 50/50 unless the play the ditto. Stage 1 of game 1 shouldn't necessarily try to make the matchup between characters 50/50, but should present the median MU ratio from across the legal stagelist, which from a preference on the competitive level is the "neutral" stage for the matchup, despite the matchup not being neutral. The point isn't to pick a stage that makes the matchup between characters even. Some characters just naturally beat other characters, regardless of the stage. Stage striking with blind picks yields a theoretically even matchup before we start factoring in players decision (such as which character to pick and which stages to strike) meaning it exists in the context of player vs. player. So if the blind pick yields a bad matchup, I'd assert it's because you made a bad character choice which is a reflection on character selection skill, so there is fault to be had there.

I disagree with this part of the assessment. Before the players have made any decisions, the probability of either one winning is meaningless, because there is nothing with which to play yet. Besides that, I disagree with assigning blame on players based on their choice for a double blind pick. With a roster of this size it's likely that there's no character that's safe from polarized matchups, and the point of double blind is simply to stop players from falling into an infinite loop of changing characters in response to the opponent's choice. There's no pretense that it's fair, put it's a necessary evil since players can't negotiate on characters.

The issue with counterpicks as used now is that every game is skewed. The first is skewed because the players generally don't pick even matchups, and then the counterpicks skew the remaining two games. E.g. The double blind results in a 7-3 matchup, then the loser counterpicks a 3-7 matchup and then the loser counterpicks a 7-3 matchup again. Ideally, the third game would be 5-5, which is why I suggested that if the game 2 loser counterpicks characters, the game 2 winner has a say in the stage.

 

[Amazing Ampharos]

Here's I think the easiest way to package things to solve all issues. Set up the rules to allow TOs to choose between the following two procedures:

Super Fast Procedure:

1. Players double blind pick characters
2. Players double blind pick custom moves
3. Players strike 4 stages each from the legal stage list (1-2-2-2-1 order)
4. Game one is played on a random stage from the remaining list
5. Loser of previous game picks a stage that was not struck game one*
6. Winner of previous game picks character
7. Loser of previous game picks character
8. Winner of previous game picks custom moves
9. Loser of previous game picks custom moves
10. Play another game with the previous choices
11. Repeat steps 5-10 until a winner is decided

*If the chosen stage is Final Destination, any omega form may be specified. If Battlefield is chosen, Miiverse may be specified instead. DSR applies and may limit stage selection.

This procedure will almost always produce a fair outcome while being very, very fast and very, very easy to explain. It also works with any stage list that has a sufficiently large number of stages (would get kinda dumb with under 15); odd numbers are not required.

Super Fair Procedure:

1. Players double blind pick characters
2. Players double blind pick custom moves
3. Players strike from the full list of legal stages in 1-2-2...-2-1 order, remembering each player's first strike
4. Game one is played
5. Loser of previous game picks three stages from legal stage list excluding the first struck stages*
6. Winner of previous game picks character and a particular stage from the loser's list of three
7. Loser of previous game picks character
8. Winner of previous game picks custom moves
9. Loser of previous game picks custom moves
10. Play another game with the previous choices
11. Repeat steps 5-10 until a winner is decided

*If the chosen stage is Final Destination, any omega form may be specified. If Battlefield is chosen, Miiverse may be specified instead. DSR applies and may limit stage selection.

This is considerably slower but produces a totally non-random result. Remembering the first strike is very similar to a stage ban but faster and probably an easy decision for most players (what is your absolutely least favorite stage?). Pick 3, winner picks from that simulates two stage bans in an ideal fashion, limiting surprise character counterpicks and avoiding blindsides of not anticipating odd stage selections. In theory this might sound worse than the above system which forces a counterpick to your 5th worst stage while here it's your 4th, but the extra information from playing game one will generally fairly compensate for that one stage difference and this system self-optimizes a lot for differing player valuations of stages (stages that one player thinks make a great CP while the other thinks aren't so bad for him get picked a lot, and that's just great). If we end up with an even number of decent legal stages, Miiverse would be added to the list of strikeable stages for game one but would still be considered the same stage as Battlefield for all other purposes (DSR, counterpicking), but we should really hope we just end up with an odd number by default.

TOs who want speed can pick the former while TOs who want the perfect result all of the time can pick the latter. Both I feel pretty well address the game one polarization discussed. The idea with counterpicking is that the end result of the character and stage selection should as a top priority be biased in the previous game's loser's favor but that the magnitude of this bias should be as small as possible. Allowing players to pick their best possible stages as counterpicks is mostly what causes lopsided situations, and by the 4th or 5th best stages you have stages that are mostly fair unless either your character just sucks (must suck if your character can't handle that many stages!) or you just suck (again, must suck if you have problem with that many stages!) both of which are pretty much personal problems belonging to players who don't really deserve to win tournament sets. Did I leave any significant holes with this?

 

[DeLux]

Why not offer two and pick one? Why not offer four and pick one? Why not offer five and pick one? etc.

 

[Uniit]

I really like this idea, but bans that a tranfered into other games restrain the character change, and can lead to weird interaction, like faking the first match character to change and have advantages in the two others.

Maybe i'm not that clear with that, but, for the firest iteration, after you first match, you're obviously not taking a char that is bad on previous "neutral" stages, even if your second actually counter your opponent's character.

Or, for the striking phase, you take your second main, strinking stage thinking for your main, and for game 2 and 3, all the stages favorise your real main.

Ok this is a wierd use case, but i do think that you're "bound" to your previous choice, and limiting others.

 

[Volt-Ikazuchi]

I think Uniit's idea of reversing the current counter-pick system could work very well.

Basically a random decision of who counter-picks first then other player's CP in the second game and double blind everything and pick stage with SSS would be very fair, despite the fact that some players will call us lunatics.

Speaking of SSS, can we forget forever that I've ever typed "Paper" on a post?

 

[LiteralGrill]

I have a funny feeling we're trying WAY too hard here.

All you need to do is make FLSS the social norm. People will not mind if it takes a bit more time that way. Because no suggestion we have will ever be used if lots of folks aren't using it.

 

[DeLux]

There's a lot in here, but incentivizing players to change characters is a difficult foundation to work with when it's not a universally held value. I like the idea of a metagame where players are required to change characters every game by rule, because I think that variety is more fun, but there's tons of players that think you should never force a player to change characters and there's even players that think polarizing stages should be banned based on a -right- to never get pushed off your main.

Another reason I do the 9 stages with 2 bans. The symmetry of stages banned against stages left in play in the counter pick phase (at least in Brawl) made it so that a player could reasonably expect to win with using one character, but would gain some (but not overwhelmingly huge to necessitate having to switch characters) amounts of a tactical advantage by having additional characters to pick from their pool of choices. Obviously it's hard to quantify where that line is, but assuming the point of having the CP system in the first place is to encourage diversity in option use (as opposed to interpreting the intent of counterpicking as favoring the loser), finding that balance of tactically advantageous in contrast to necessitating a switch is a worthwhile pursuit.

 

[T0MMY]

Back with the data from my last tournament.

Overview
Events:

For Fun!
For Glory!
Melee 2v2
Melee 1v1​

(We will be looking at For Glory! and Melee 1v1)

Rulesets:

FG: "Oregon Standard" Ruleset
SSBM: Revised Apex​

Game Settings (basic):

For Glory!: 2 stock, 5-min - Bo3; Finals Bo5. Double-elim
Stage List: Agreement Method (all available)

Melee 1v1: 4 stock, 8-min - Bo3; Finals Bo5. Double-elim
Stage List: Apex list (5 starters, 1 counterpick)​

And the data:
FG!:
Despite all methods offered records showed competitors used Agreement to choose a random Omega Stage or Final Destination and played. No other method was used. This took only seconds to decide and was not discussed on subsequent rounds of play.
No issues arose.

SSBM:
Despite Apex rules being used, Random was the most used method recorded, Gentleman's Agreement was used nearly as much but Grand finals was entirely Random which was a major focus of this study so it set it apart. Striking was the least used method.
No issues arose.

Analysis
Regarding why Agreement and Random was preferred over Striking was because Striking entailed the opponent got to choose the next stage, which seemed unfair.

Not only was this tournament successful in that it met my goals of running on time (we had about 30 attendees and ended by allotted time of business hours), there were no issues, and all players felt they had a fair match.

Big Questions to answer:
Why did attendees prefer Final Destination/Omega Stage over going to any other Stage when ALL STAGES were allowed? There was no disagreement when it was offered which would have lead to a Striking or Random method to be used.
Similar for Melee with Apex's Stage List offered a limited Stage List players were fine using Random stages.

I have some big answers for these big questions, but it is far more telling to show my solution, which is a well-formed Inclusive Ruleset.

 

[Piford]

Back with the data from my last tournament.

Overview
Events:

For Fun!
For Glory!
Melee 2v2
Melee 1v1​

(We will be looking at For Glory! and Melee 1v1)

Rulesets:

FG: "Oregon Standard" Ruleset
SSBM: Revised Apex​

Game Settings (basic):

For Glory!: 2 stock, 5-min - Bo3; Finals Bo5. Double-elim
Stage List: Agreement Method (all available)

Melee 1v1: 4 stock, 8-min - Bo3; Finals Bo5. Double-elim
Stage List: Apex list (5 starters, 1 counterpick)​

And the data:
FG!:
Despite all methods offered records showed competitors used Agreement to choose a random Omega Stage or Final Destination and played. No other method was used. This took only seconds to decide and was not discussed on subsequent rounds of play.
No issues arose.

SSBM:
Despite Apex rules being used, Random was the most used method recorded, Gentleman's Agreement was used nearly as much but Grand finals was entirely Random which was a major focus of this study so it set it apart. Striking was the least used method.
No issues arose.

Analysis
Regarding why Agreement and Random was preferred over Striking was because Striking entailed the opponent got to choose the next stage, which seemed unfair.

Not only was this tournament successful in that it met my goals of running on time (we had about 30 attendees and ended by allotted time of business hours), there were no issues, and all players felt they had a fair match.

Big Questions to answer:
Why did attendees prefer Final Destination/Omega Stage over going to any other Stage when ALL STAGES were allowed? There was no disagreement when it was offered which would have lead to a Striking or Random method to be used.
Similar for Melee with Apex's Stage List offered a limited Stage List players were fine using Random stages.

I have some big answers for these big questions, but it is far more telling to show my solution, which is a well-formed Inclusive Ruleset.

If people preferred going to Final Destination its either
a) the stage is great for their character
b) they like the visual and music of others stages but not the hazards
c) Its was easy to agree upon since they didn't want to risk random
d) They actually think Final Destination is the best stage probably because of For Glory or

 

[popsofctown]

Another reason I do the 9 stages with 2 bans. The symmetry of stages banned against stages left in play in the counter pick phase (at least in Brawl) made it so that a player could reasonably expect to win with using one character, but would gain some (but not overwhelmingly huge to necessitate having to switch characters) amounts of a tactical advantage by having additional characters to pick from their pool of choices. Obviously it's hard to quantify where that line is, but assuming the point of having the CP system in the first place is to encourage diversity in option use (as opposed to interpreting the intent of counterpicking as favoring the loser), finding that balance of tactically advantageous in contrast to necessitating a switch is a worthwhile pursuit.

Spending 900 hours in ye olde cellar with m'lord's findings from his explorations to create ye philosopher's stone is also a worthwhile pursuit in terms of its goal, but an equally difficult one. Brawl had incredible amounts of matchup polarization and yet was still dominated by players that stuck to one main. Even if we have an equally wonky stagelist this time around, the lack of chaingrabs and other fair-play mechanics in Smash 4 will make it incredibly difficult to offer dual maining as a legitimately viable alternative to single maining. Furthermore you have custom moves to invest time in and shore up weak matchups without learning an entirely new character. I wouldn't expect 0 bans in the banlist to be low enough to foster dual maining. Even Little Mac players would probably be better of quitting Little Mac to go pure Diddy or something.
Even what little dual maining went down in Brawl largely included people getting bored and knowingly sacrificing some winrate for some variety (and pocketing MK instead of quitting your main for MK counts among that).

 

[Amazing Ampharos]

Why not offer two and pick one? Why not offer four and pick one? Why not offer five and pick one? etc.

Glad you asked. I'm assuming a number of legal stages in the vague region of 20ish (+-3 either way) and the general character of what we've seen in Wii U stages (so relatively few extreme stages like Norfair exist). That means no reasonable character is actually going to need a huge number of bans. That being said, we still do want to eliminate the tails of the stage list with more eliminations producing better results. However, picking 5 cp stages is likely to produce an analysis paralysis situation and makes the decision legitimately hard to the point that players are significantly more likely to self-destructively make a selection error. Based on my understanding of human psychology, picking 3 versus 4 is the tipping point of when it becomes a very tedious and difficult choice versus an easy one, and I didn't feel that was quite a large enough number but close so carrying over the very first strike effectively expands it with minimal inaccuracy versus picking four.

I really like this idea, but bans that a tranfered into other games restrain the character change, and can lead to weird interaction, like faking the first match character to change and have advantages in the two others.

Maybe i'm not that clear with that, but, for the firest iteration, after you first match, you're obviously not taking a char that is bad on previous "neutral" stages, even if your second actually counter your opponent's character.

Or, for the striking phase, you take your second main, strinking stage thinking for your main, and for game 2 and 3, all the stages favorise your real main.

Ok this is a wierd use case, but i do think that you're "bound" to your previous choice, and limiting others.

Remember I'm presuming you're striking away less than half of the legal stage pool in the time optimized system so your cp characters still have a lot to work with, and as a player you're going to introduce large biases to the types of stages you want to play on that transfer well between games. Stage bans are also a slow and clumsy system; they take forever in real tournaments since deciding on bans is a very hard decision for a huge proportion of real players. The time optimized system uses this system heavily mostly just to save time and keep things simple and smooth. The fairness optimized system relies on it a lot less. In practice as well, faking your main in game one is kinda suicidal; it will make it easier to win on cp stages sure, but that advantage can never possibly exceed the advantage from winning game one (since if you lose game one, you must win on an opponent's cp to win the set). I don't think we have to worry about those mindgames.

 

[popsofctown]

I give up on this stuff. I'm gonna try to avoid this thread for a while.

 

[Davis-Lightheart]

Bored. Did a mock up what a phone app for Stage striking could look like, though I'm no UI master or much of a master of language.

Spoiler

 

[Volt-Ikazuchi]

Bored. Did a mock up what a phone app for Stage striking could look like, though I'm no UI master or much of a master of language.

Spoiler

I dig it. Simple and effective. No need to be flashy.

 

[Davis-Lightheart]

I dig it. Simple and effective. No need to be flashy.

A phone app is one of the best ways to really teach casuals the system. Almost everyone has a smartphone, and an app is something you can carry everywhere. Pack it with a tutorial of FLSS and Classic counterpicking and we'd have less problems with new players. Heck; advertise it on the sub, and I'm sure almost everyone would download it out of curiosity. And who knows; maybe we could datamine the thing for useful information.

The one problem is finding someone to actually program it.

 

[Piford]

A phone app is one of the best ways to really teach casuals the system. Almost everyone has a smartphone, and an app is something you can carry everywhere. Pack it with a tutorial of FLSS and Classic counterpicking and we'd have less problems with new players. Heck; advertise it on the sub, and I'm sure almost everyone would download it out of curiosity. And who knows; maybe we could datamine the thing for useful information.

The one problem is finding someone to actually program it.

Programming it probably isn't that hard, its finding someone willing to pay the fees in order to get it published.

 

[Davis-Lightheart]

Programming it probably isn't that hard, its finding someone willing to pay the fees in order to get it published.

Eh. Good point.

 

[Pazx]

A phone app is one of the best ways to really teach casuals the system. Almost everyone has a smartphone, and an app is something you can carry everywhere. Pack it with a tutorial of FLSS and Classic counterpicking and we'd have less problems with new players. Heck; advertise it on the sub, and I'm sure almost everyone would download it out of curiosity. And who knows; maybe we could datamine the thing for useful information.

The one problem is finding someone to actually program it.

We don't have a problem with new players; last tourney I went to had a bunch of Brawl first-timers, striking was introduced to them as "taking turns turning off the stage you don't want to play on" and banning is as simple as "you pick one stage from these that we WON'T play on and I'll choose from what's left".

I'm not entirely opposed to a phone app for how easy it will be to use, but claiming that it will be better for casuals/new players is misleading.

 

[Uniit]

I'm actually a computer science student, and i will definetly give a shot for this app.
At first you don't have to submit it to the play store. One can just download the .apk and install it on his/her phone.
And if the app is pretty sucessful, i don't mind paying the 25€ fee for the Play Store, and with it i could submit other app as well.

But don't count on me for iOS though !