All Star Mode is the Key to Balancing Project M (A Tournament Ruleset) UPDATED

[frankxthexbunny]

INTRODUCTION: The Folly of Balancing Project M as if it was Melee
​

There are 861 Matchups in Project M. Not only that, but characters in Project M are far more volatile than melee. We have a Ganon who can chaingrab to death. We have a Snake with punishes that are so brutal that he mind as well be wobbling if he gets that tranq on you. We have characters that heal themselves and characters with the aerial game of jigglypuff and yet still can play aggressively. We have Lucario. It is increasingly clear the more it is discussed on tier list threads that pushing for a game with an even matchup spread, that is to say, 50/50 matchups across the board, is not only impossible given the fact that we're unwilling to homogenize characters out of their most interesting traits, but so unreachable that in this game playing Project M competitively for money will be inconsistent for years to come. This is not just a problem for high level players that have a chance at winning first prize in a game. If high level players feel like they might lose a match because of an inevitably bad matchup, they will not waste the money of participating in a tournament. Without high level players there, low level players are less likely to attend. The fact that the diversity in character traits is causing volatile matchups for even the high tier characters makes the prospect of securely winning a tournament as the best player a matter of bracket layout and even luck. Not only can a high level player not hope to deal with all the inevitably volatile matchups that will occur with low level players, it is very likely that for years to come there will be MANY instances of hardly known players coming in with a character that hasn't been adequately explored and cleaning house with something players are not prepared to deal with.

With this in mind let us ask a few questions and see where it takes us:

What can the player do about this?

Well for one it is inevitable that everyone is going to dual main. Any character that can effectively solo through a bracket will be nerfed. Without question. Every character will counter some characters, and if they are not countered by some characters to even it out they are imbalanced and will be nerfed. Inevitably every player will have a character that they use as a main against most the matchups, a consistent character like Mario or Fox, and have an alt for all their bad matchups. This is a patch to the problem but won't fully solve it, because they do have to accurately guess who their opponent is going to play, and since their enemy will also be dual maining this will become part of the meta, and people are still going to find themselves losing at character/stage select. No matter how good you are, everyone loses at character selection eventually, and if they don't, they are playing a character that the community perceives to be overpowered. Left how it is this creates a counterpick war in the meta. If we continue on this path however we're going to have a melee top 8, and they will be decided by the characters a high level player can feel most consistent maining. Every other character will be considered too inconsistent to main reliably in a tournament setting. Even these characters will have some counters more than likely in the 41 character cast, and will lose tournaments to characters considered too volatile to main.

What can PMDT do about this?

They can try to even the matchup spread. I have mentioned that this is impossible, but let's assume they try. This creates the melee scenario I mentioned before. So let's instead stop trying to fight against the current and see if we can balance the game with this certainty in mind. Let us instead nerf the characters who have few bad matchups but many good matchups, and instead balance everyone to create a bell curve around a 50 50 matchup spread. For every 30 70 a character has we do our best to make sure they ALSO have a 70 30. This is already the case to an extent, the only difference will be PMDT keeping this in mind as they create further balance patches. If we cannot avoid a melee scenario for pushing for across the board 50 50, we balance by giving EVERYONE a few bad matchups. This technically balances the game, but doesn't solve our problem: bad matchups existing in the game still make tournament victory inconsistent and therefore not worth it. In order to solve this problem we have to realize that our goals: to have everyone be viable in tournament and to have a game with no bad matchups yet unique characters is a fools dream with our current rule set. So since this game is going to have characters with counters and characters who counter them let us turn to other games that deal with this problem.

The games I found that solve this problem are Pokemon and League of Legends. The commonality with their solution is simple: give you a team of characters that you can choose their attributes, and then have the metagame in optimizing your team. Project M has a game mode that makes this possible, and it was added as just a side gimmick for fun.

The Tournament Rules
​

Umbreon's thread (http://smashboards.com/threads/project-m-recommended-ruleset.396408/) is the base that I'll use with a few differences.

NOTE: Originally these rules stipulated that you were not allowed to repeat characters within matches. This has been argued to a satisfactory manner both in and out of the thread. I still believe that the ruleset benefits from team based fighting, but the option to play only one character should still stand with the knowledge that you can be fighting 4 different enemies in your next match. ​

• The matches are set to All Star Mode. 4 Stocks.
• Both players choose their characters blind.
• After first match they switch off choosing characters in front of each other, Winner choosing first. If you intend to play only a single character, indicate so when you choose your character.
• Note: There's an also optional pattern of switching off character choice, this is the ABBAABBA pattern, where winner chooses first, then loser twice, then winner twice, then loser twice, then winner one last time. This gives everyone an equal chance to counterpick the other. I won't make that the official rule until I've tested it yet, so for now I'll call it TO optional.

Everything else is the same. Stage counterpick rules are up to the TO but Umbreon's is a good base. You can choose the same team or a different between matches.

The Benefits of This System​

• The Rock, Paper, Scissors in matchups is mitigated GREATLY.

By having each player play 4-8 matchups per round, you'll normalize the matchup spreads present within any given game/set. it won't make any of the individual matchups more inherently balanced, but it will result in fewer matches being determined in large part by counterpick rather than in-game performance.

• Uses rules and systems already present within the game.

This is for the TO's sakes. It's simple, elegant, and easy to manage. The game already has All Star Mode, and the counterpick rules are essentially the same.

• Makes the counter pick metagame more interesting yet less decisive

The order that you choose your characters, and your predictions for their order will make some interesting mind games. if you make a mistake however, the game isn't cost as long as you have a diverse team with attributes that can handle anything the enemy team throws at you.

• Stage choice still influences the game greatly while being less likely to cost a player the game.

Another example of losing at character select, stage choice can be the difference between victory and defeat sometimes. Victory should ALWAYS be due to the player's skill, not due to bad matchups, and the stage influences matchups way less decisively when a player's characters are more diverse.

• Far easier to balance for every character being viable.

I am not saying it will still be a challenge to balance, but PMDT will not be swimming against the current anymore. All they have to do is make sure every character has as many bad matchups as it has matchups in their favor. Characters will be defined as volatile (matchups that gravitate towards extremes) and stable characters (matchups that gravitate towards 50 50) and both will be useful in a team meta.

With more normalized matchups using this method victory will be more about skill and strategy than ever, high level players will have more reason to take the game seriously, and with higher attendance from high level players low and mid level players will join as well, creating a better victory pot which encourages more tournaments and more coverage. It could be the key to making Project M as big as any smash game has ever been.

Possible Criticisms of the Rule Set​

• I do not want to have to learn multiple characters. It'll spread me too thin.

Unfortunately due to the vast number of matchups either way we decide to do the rules we're going to have to main multiple characters. King of Fighters uses the same system as this, and it has a fine competitive scene, and it's also notable to mention that project M has very very easy to learn characters compared to other fighting games. Yes to master a project M character you have to put just as much effort as any other fighting game, but I repeat, Project M is not the first game to have a system like this, and in order for it to survive adopting this system is important. It is the only way to have as many viable characters as we want.

• Project M was supposed to be like Melee. Changing its ruleset contradicts its goal to emulate a more balanced melee.

I've been over this enough but I just want to be very clear about this: goals or no goals Project M is not like melee. Melee as a competitive system ONLY works as a top 8 scenario. The complicated meta unique to melee exists because there are far fewer matchups than would be if everyone is viable. Being like Melee and having everyone viable are two mutually exclusive states. No game can do melee better than melee, and even trying is a waste of time. So why not stick to the strengths inherent in project m's unique attributes? If it has a rock paper scissors metagame, why not give it a rule set that fits with the rock paper scissors metagame in an interesting and elegant manner that does not in any way make project M less fun to watch, and in fact will improve watchability in some ways. All the best players have pocket characters. Let's just go ahead and embrace that fact and go with it.

If there are any other criticisms please voice them in the comments. I want to hear people's thoughts. I've been considering this all day and I really cannot see any problems with it if and only if the PM community is willing to let go of the past and adapt. Smash 4 has begun using custom movesets in tournament. They adapted because they realized they are their own thing and free to do things their own way. Can we prove to other smash players that we have what it takes to adapt to the unique game we've been given? Or are we just going to do things the way they've always done because we want things to always stay the same. This decision is key to the the future of project m as an inconsistent mod, or a fully viable competitive game.


I'd like to thank Zigludo, Elohemian, and ModestM00se for their assistance.

 

[mimgrim]

There are NO repeating characters within a match.

Nope nope nope.

Stopped caring after this.

This is a good way to make me stop going to/participating in P:M events and probably end up dropping Smash altogether.

 

[frankxthexbunny]

what do you propose? They can just go marth 4 times? if so why have it be all star mode in the first place?

Having the option to repeat characters presents serious balancing issues.

 

[mimgrim]

what do you propose? They can just go marth 4 times?

If that is what they want to do then they should be able to do it.

if so why have it be all star mode in the first place?

People can still do stuff like character A/character B/character A/Character B or any other set of combinations you can think of. The option is still there if players want to use 4 different characters or who want to use just 2 character, both twice each. Or want to use 1 character 3 time sand another once. Or just use the same character all 4 times.

Don't limit a person's freedom lmao.

 

[frankxthexbunny]

If that is what they want to do then they should be able to do it.



People can still do stuff like character A/character B/character A/Character B or any other set of combinations you can think of. The option is still there if players want to use 4 different characters or who want to use just 2 character, both twice each. Or want to use 1 character 3 time sand another once. Or just use the same character all 4 times.

Don't limit a person's freedom lmao.

all rules limit freedom. That's what rules are. But my rules balance the game. If you have the option to repeat characters you break the balance that all star mode creates.

 

[Chevy]

If you're going to do all star it seems you have to do 4 different characters, that's kind of the whole point. This has way too many problems to ever be the official ruleset though. Mainly character pick order, counter-pick order is going to be a mess and at least double the amount of time we spend in the menu, most likely multiply by much higher than that. There are a lot of rules you would have to test, and re-test and people are too dumb to remember all of that.

When the matches actually play out, you still have to play potentially bad MUs depending on character order. If you go in blind picking characters it's essentially luck, and any other system heavily rewards whoever won RPS or is on the counterpick.

Also how many times do I have to explain that you can't just square the number of characters in the game to count the number of matchups?

 

[frankxthexbunny]

If you're going to do all star it seems you have to do 4 different characters, that's kind of the whole point. This has way too many problems to ever be the official ruleset though. Mainly character pick order, counter-pick order is going to be a mess and at least double the amount of time we spend in the menu, most likely multiply by much higher than that. There are a lot of rules you would have to test, and re-test and people are too dumb to remember all of that.

When the matches actually play out, you still have to play potentially bad MUs depending on character order. If you go in blind picking characters it's essentially luck, and any other system heavily rewards whoever won RPS or is on the counterpick.

Also how many times do I have to explain that you can't just square the number of characters in the game to count the number of matchups?

ok tell me how many matchups there really are so I can edit it.

I'm considering having a tentative change to the "no repeat rule," only because if it caught on I'm almost definite the meta will demand no repeats anyways. And it would also cut down times since a lot of people will at least at first chose one character, or mayb two. What's important is that players at all times CAN chose two- four characters in a match.

EDIT
Look, here's the thing, the big problem with having repeats be allowed is it doesnt mesh well against team enemies. Having 4 of the same character is really really commital, it's why it causes such an imbalance that makes you lose at character select. It will have a lot of complicated effects against mixed teams, but the crux of the matter is you lose all the advantages that having all star mode no repeats allows. The only way to balance it is to have mixed teams at both sides. This isn't the first game that does it.

 

[AuraMaudeGone]

You have a strange and complicated view of balance... Why would any dev balance a cast of characters by MU, esp. in this game?
Some of the issues I see were already mentioned, but it might be a cool side event if this gets finessed.

 

[Chevy]

Going from first character, let's go with Wario as per stage select: Wario has 41 distinct matchups. Next is Mario, who has 41 distinct MUs, but 1 of those has already been covered with Wario so +40. Next is Luigi, Luigi-Wario and Luigi-Mario have already been counted so +39, etc. Final count is 861 matchups if my math checks out.

 

[frankxthexbunny]

Going from first character, let's go with Wario as per stage select: Wario has 41 distinct matchups. Next is Mario, who has 41 distinct MUs, but 1 of those has already been covered with Wario so +40. Next is Luigi, Luigi-Wario and Luigi-Mario have already been counted so +39, etc. Final count is 861 matchups if my math checks out.

Well maybe, I mean theres an argument that the matchup of fox marth from fox's perspective and from marth's perspective is two distinct matchups, but you're probably right.

You have a strange and complicated view of balance... Why would any dev balance a cast of characters by MU, esp. in this game?
Some of the issues I see were already mentioned, but it might be a cool side event if this gets finessed.

Do you not agree with my theory that PM brackets will be much more relevant to tournament placing than it is for say Melee for years to come? That high level players will be discouraged because they can get knocked out by, dare I say it, jank?

EDIT: One way or another I'm having a really hard time seeing why people shouldnt at all times have the option to use multiple characters in one match. Though I'm positive the key to Project M having any serious longevity is in the ruleset i proposed I at least would like to see more tourneys with all star mode active just in case.

 

[AuraMaudeGone]

Snip

I'm just saying there's better points of view, like let's say how healthy a character's kit interacts in the dev's intended meta. Balancing per MU sounds hair splitting and time consuming.

 

[frankxthexbunny]

I'm just saying there's better points of view, like let's say how healthy a character's kit interacts in the dev's intended meta. Balancing per MU sounds hair splitting and time consuming.

i dont think having extreme matchups against certain characters fits anyone's ideal meta

 

[frankxthexbunny]

Double posting to mention that I edited the rules.

 

[Foo]

HOLY !@#$ OH GOD PLEASE NO! This would kill PM so fast, it would be like it never even existed. If this ruleset was ever implemented I would instantly quit the game forever and never enter a tournament again.

The whole beauty of this game is character specialists. There are so many ****ing characters in this game that can all do some really awesome stuff and watching or experience the people who put countless hours perfecting that one character wreck face is just incredible. The game actually has a self balancing mechanic that would be RUINED by all star, but I'll get to that later. Here is everything wrong with this, that I can think of.

1. It would KILL the game. For reasons that will be stated below, nobody would want to take this game seriously. PM all stars is fun, but it is completely different from regular PM. You'd have to completely switch all your muscle memory, mentality, and game plan every time a stock was taken.

2. Having to understand the matchups for more than 40 characters is hard enough, but having to know it for 4 different characters is absolutely !@#$ing ridiculous and impossible.

3. It would be impossible to balance the pick (for counterpicks) system unless you changed it to 5 stock. Since everyone would be worse at all their characters, games would take EVEN longer and time outs would be a thing. Not to mention, playing 5 characters makes the game exponentially harder to figure out AGAIN.

4. It would ruin the character specialist. That's one of the best things about smash imo, but if you had to put time into at LEAST 4-5 characters, you would never see people be as good at any one character as you see now.

5. It would lower character diversity. If I had to play four-five characters, you know who I and everyone would pick? The easiest/most familiar ones. That means there's be a million Foxes, Roys, Marths, Sheiks, etc. etc. and very few other characters because they are so complex and take a lot of dedicated practice.

6. It would make the game LESS balance. PM already has a self balancing mechanic. While ZSS (imo) loses pretty badly to fox, you see zss player, myself included, beat up foxes close to their level because I play foxes 100 times more than they play ZSS. Since so many characters are viable, it's impossible to have practice against all of them. Therefore, the "low tier factor" comes in strong. I can turn bad matchups into good matchups if they don't know WHY the matchup is bad.

7. In addition to longer games, character and stage select would take much longer.

8. This deserves to be higher, but I'm lazy. The best part about watching games being played competitively is seeing them played at their bests. You want to see someone do everything their character is capable of and push them to the limit. The more characters people are forced to play, the worse they will play each character they have.

9. Stages matter less since you are playing 4-5 matchups where stage preference could change each time a character dies.

Also, to counter a point you made.

"Unfortunately due to the vast number of matchups either way we decide to do the rules we're going to have to main multiple characters. King of Fighters uses the same system as this, and it has a fine competitive scene, and it's also notable to mention that project M has very very easy to learn characters compared to other fighting games. Yes to master a project M character you have to put just as much effort as any other fighting game, but I repeat, Project M is not the first game to have a system like this, and in order for it to survive adopting this system is important. It is the only way to have as many viable characters as we want."

Flat out untrue. Fox has arguably no bad matchups and many many people successfully solo main characters. Also, you can pickup one secondary to cover bad matchups and do alright. Four characters is basically 10 times as much to learn and master and players haven't pushed a single character to their limits yet.


Your theory only works if people had the capacity to master and rapidly adjust between several different characters, but in practice it would just make everyone play the game much worse and make it extremely boring to watch, more "gimmicky" and extremely... lame.

(note: I wrote this before you removed no repeating characters. Currently adding more for new system.)


If you allow repeating characters, this game becomes the ultimate cancer.

Game 1: Everyone picks 4 of one character blind.

Game 2: Loser counterpicks every single character the winner chooses in order.

Game 3: Since the winner of game 3 should lose this one, the reverse happens and HE now gets an absurd advantage.

Basically, loser gets an unbelievable advantage in theory. Most likely, however, almost everyone will just pick 4 of one character every time anyway.

 

[frankxthexbunny]

stuff

Ok you added to your comment but you didn't also subtract. With the repeating character rule change your points 1, 2, 4, 5, and most of 6 are irrelevant.

Your final point I also think is a bit silly, because Loser is free to counterpick even in the current system. How in the hell is counterpicking 4 characters different than counterpicking one character for four stocks? It'll add more variance to the match not less.

It seems that you are the one who assumes people are going to have absurd amounts of character knowledge, rather than just 3-4 mains that they change the order of every time.

there no way in hell giving the optional ability to switch characters mid match would kill project m.

 

[Droß]

Going from first character, let's go with Wario as per stage select: Wario has 41 distinct matchups. Next is Mario, who has 41 distinct MUs, but 1 of those has already been covered with Wario so +40. Next is Luigi, Luigi-Wario and Luigi-Mario have already been counted so +39, etc. Final count is 861 matchups if my math checks out.

Incorrect. The number of match ups in Project M is 41^2 - 41, which is 1640.

By extension, the number of matchups in a 3 player free for all is 41^3 - 41, which is 68,880.

You assumed that a matchup was the same for both players, (ie. Wario vs GnW is identical from Wario's perspective and GnW's perspective.) This is only true if you play both characters extensively enough to know everything about both sides of the coin, which very, very few people can do for all matchups. However, dittos are truly the same from both sides, so we need to subtract 41 from our answer to get our final total.

Edit: To further explain the concept of the matchup, contemplate this: do you play your Bowser vs Squirtle identically to the way you play your Squirtle against Bowser?

 

[LightningDragon]

As interesting as this concept is, I'm very skeptical. Partly because of a skewed perception of how balancing works. You don't balance based on match-up, you balance based on tools and options.

I'd much rather see this tested via side events than see an actual push to make this a widely used ruleset.

By the way, I do have a problem with one of your points:

• Project M was supposed to be like Melee. Changing its ruleset contradicts its goal to emulate a more balanced melee.

I've been over this enough but I just want to be very clear about this: goals or no goals Project M is not like melee. Melee as a competitive system ONLY works as a top 8 scenario. The complicated meta unique to melee exists because there are far fewer matchups than would be if everyone is viable. Being like Melee and having everyone viable are two mutually exclusive states. No game can do melee better than melee, and even trying is a waste of time. So why not stick to the strengths inherent in project m's unique attributes? If it has a rock paper scissors metagame, why not give it a rule set that fits with the rock paper scissors metagame in an interesting and elegant manner that does not in any way make project M less fun to watch, and in fact will improve watchability in some ways. All the best players have pocket characters. Let's just go ahead and embrace that fact and go with it.

Trying to be like Melee doesn't mean being a 1:1 recreation. PM is trying to be similar to Melee, it's not trying to be Melee. There's a difference. It's meant to be more than just Melee, sure, but it's heart still lies in taking what's good in Melee (and the rest of the series), and making something great, along with a larger, more balanced roster. If you're going to address a criticism or potential criticism, at least make sure you're addressing what it actually means.

 

[Foo]

Ok you added to your comment but you didn't also subtract. With the repeating character rule change your points 1, 2, 4, 5, and most of 6 are irrelevant.

Your final point I also think is a bit silly, because Loser is free to counterpick even in the current system. How in the hell is counterpicking 4 characters different than counterpicking one character for four stocks? It'll add more variance to the match not less.

It seems that you are the one who assumes people are going to have absurd amounts of character knowledge, rather than just 3-4 mains that they change the order of every time.

Winner picks a character:
Loser picks the best character they can to counter:
Winner picks character to counter that one:
Loser picks character to counter that one.

etc. etc. I really shouldn't have to explain why that's so imbalanced. If the loser doesn't have the ability to abuse that counterpick system, your system adds nothing. Game 1 will almost always be four of one character anyway, since why would you lead with anything other than your best character in blind pick? Unless people are playing some next level mind games, it wouldn't happen. If loser can't abuse the pick order to get counters to every character his opponent selects, nothing is added to the game, so why change the rules at all?

(here's an example) Let's say each player plays 3 characters Player one plays a, b, and c and player two players a, d, and e.

a-b is 60-40
a-c is 40-60
a-d is 50-50
a-e is 30-70
b-d is 40-60
b-e is 65-35
c-d is 30-70
c-e is 50-50

Player one wins then leads with a
player 2 picks e winning stock
Player one picks to b counter
player 2 picks d to counter

etc. etc. etc. I don't even want to write this out. OR people could pick the same two characters alternating. Either those two happens, or your system adds nothing. This is just silly, and focuses the game even more on it's biggest flaw, matchups. Since this game is already so matchup dependent, do you really want to push people even more towards playing the matchup game rather than the skill game? This system rewards being able to pick counters much more than classic PM and thus takes emphasis away from raw skill, while making it really silly and annoying.

 

[frankxthexbunny]

Winner picks a character:
Loser picks the best character they can to counter:
Winner picks character to counter that one:
Loser picks character to counter that one.

etc. etc. I really shouldn't have to explain why that's so imbalanced. If the loser doesn't have the ability to abuse that counterpick system, your system adds nothing. Game 1 will almost always be four of one character anyway, since why would you lead with anything other than your best character in blind pick? Unless people are playing some next level mind games, it wouldn't happen. If loser can't abuse the pick order to get counters to every character his opponent selects, nothing is added to the game, so why change the rules at all?

(here's an example) Let's say each player plays 3 characters Player one plays a, b, and c and player two players a, d, and e.

a-b is 60-40
a-c is 40-60
a-d is 50-50
a-e is 30-70
b-d is 40-60
b-e is 65-35
c-d is 30-70
c-e is 50-50

Player one wins then leads with a
player 2 picks e winning stock
Player one picks to b counter
player 2 picks d to counter

etc. etc. etc. I don't even want to write this out. OR people could pick the same two characters alternating. Either those two happens, or your system adds nothing. This is just silly, and focuses the game even more on it's biggest flaw, matchups. Since this game is already so matchup dependent, do you really want to push people even more towards playing the matchup game rather than the skill game? This system rewards being able to pick counters much more than classic PM and thus takes emphasis away from raw skill, while making it really silly and annoying.

This is the part where you tell me how that's different than the current counterpick rules.

Way I see it the way things you are now player 2 loses Player 1 leads with A Player 2 picks E and wins EVERY stock. At least my way has a higher chance of varience and matchup normalization than yours where there's some mythical mastermind who happens to have a counter for all of player 1's characters in pocket (in which case he already lost hasn't he?)

 

[Chevy]

Incorrect. The number of match ups in Project M is 41^2 - 41, which is 1640.

By extension, the number of matchups in a 3 player free for all is 41^3 - 41, which is 68,880.

You assumed that a matchup was the same for both players, (ie. Wario vs GnW is identical from Wario's perspective and GnW's perspective.) This is only true if you play both characters extensively enough to know everything about both sides of the coin, which very, very few people can do for all matchups. However, dittos are truly the same from both sides, so we need to subtract 41 from our answer to get our final total.

Edit: To further explain the concept of the matchup, contemplate this: do you play your Bowser vs Squirtle identically to the way you play your Squirtle against Bowser?

I considered that. I would, however argue that when referring to a matchup, most consider two characters and the contained interactions, regardless of side. Yes they are very easily separated into two distinct categories, but for clarity's sake we shouldn't just decide that that's what a MU is when discussion and context has never set that precedent.

In short, it doesn't matter I was just pointing out faulty math because for some reason that has come up a lot lately.

 

[_Chrome]

I would like to point out that in Pokemon (at least the old versions) there are only a few viable Pokemon that are used in each format, save a few unorthodox picks here and there (afaik). That's what would happen, except for the character specialists; which would be reduced to stable characters such as Sheik and MK and Fox and friends.

All information relating to Pokemon goes to @Avro-Arrow btw.

EDIT: It is an interesting concept, and I commend you for putting this much effort into it. We need more people in the community (and life) who are willing to defy the conventional and think outside of the box.

 

[frankxthexbunny]

I would like to point out that in Pokemon (at least the old versions) there are only a few viable Pokemon that are used in each format, save a few unorthodox picks here and there (afaik). That's what would happen, except for the character specialists; which would be reduced to stable characters such as Sheik and MK and Fox and friends.

All information relating to Pokemon goes to @Avro-Arrow btw.

then nothing will be different. All my system does is adds one more option: to have 2 pokemon on a stage instead of just 1. (or 3 or 4 if you're ambitious).

 

[_Chrome]

then nothing will be different. All my system does is adds one more option: to have 2 pokemon on a stage instead of just 1. (or 3 or 4 if you're ambitious).

So what your system tries to do is favour the player who doesn't have a stable main then... and they'll choose a character mid-match who *maybe* hard counters the opposition giving them significant advantage. The solo main had better choose a very stable main then.

 

[Boiko]

Incorrect. The number of match ups in Project M is 41^2 - 41, which is 1640.

By extension, the number of matchups in a 3 player free for all is 41^3 - 41, which is 68,880.

You assumed that a matchup was the same for both players, (ie. Wario vs GnW is identical from Wario's perspective and GnW's perspective.) This is only true if you play both characters extensively enough to know everything about both sides of the coin, which very, very few people can do for all matchups. However, dittos are truly the same from both sides, so we need to subtract 41 from our answer to get our final total.

Edit: To further explain the concept of the matchup, contemplate this: do you play your Bowser vs Squirtle identically to the way you play your Squirtle against Bowser?

If you're talking pure numbers, the formula is 41!/(2!(41-2)!) = 820.

You're right that match ups will be played differently from each side, so player A Bowser vs. player B Squirtle is different than the other way around. But that just makes everything convoluted, and it's still the same two characters on the screen.

 

[frankxthexbunny]

So what your system tries to do is favour the player who doesn't have a stable main then... and they'll choose a character mid-match who *maybe* hard counters the opposition giving them significant advantage. The solo main had better choose a very stable main then.

What my system does is acknowledges that project M is going to become a big ol' counterpick war with or without All Star Mode. My rule set just midicates the damage of counterpicking.

 

[mimgrim]

My rule set just midicates the damage of counterpicking.

If anything it makes it worse.

 

[frankxthexbunny]

If anything it makes it worse.

I'd be very curious to hear how the option for character variety within matches makes matchups more volatile rather than less.

 

[Foo]

This is the part where you tell me how that's different than the current counterpick rules.

Because it's exactly the same, but incredibly stupid unless everyone masters the characters they are playing which isn't possible. Since everyone will be worse at their characters they won't know how to work around bad matchups which makes the matchups matter that much more.

At this point, you should be telling us why we should so this. It doesn't mitigate the rock-paper-scissors element at all, because not starting with four of the same character is just kinda stupid in blind pick (unless you know your opponent always runs a certain order) and in game 3 counters would either work exactly the same, or be really stupid.

I'd be very curious to hear how the option for character variety within matches makes matchups more volatile rather than less.

Because the only time you'd use it is to counterpick more. Counterpicking isn't actually a good thing. It's a necessary evil. You want the game to be as close to skill vs skill with each person's best character. If people are picking up more and more characters to counterpick more with this system, that means the game is being focused more on matchups than peak skill. It doesn't make matchups more or less volatile, it just emphasizes their strategic use.

 

[mimgrim]

Because a greater variance (in this case being able to play different characters in the same match) will generally lead to a higher probability of volatility.

Seeing as blind picking is enforced (not saying it shouldn't be, just stating a current fact) by this rule-set this effectively means each player choosing their characters is going to be random seeing as they have no way of knowing what their opponent will play and let's assume they are of equal skill.

Now in a normal set where they can only chose 1 character for the first game and only that character the variance is rather low (comparably). So like presume player A is playing character X and character X has like 5 bad MUs out of 41 character. This means there is a low chance of Player B's character (character Y) being one of those 5 characters.

However in your method if player A was a character specialist and only played character X but player B played more the one character the chance of him using one of those 5 characters becomes a bit higher.

I of course kept the examples simple (but they are not the meat of what I'm trying to say, only to show what I mean) but the general gist is that your proposed method has a higher variance to it which causes a higher chance of volatility withing sets/matches.

Also what Foo said.

 

[frankxthexbunny]

Because a greater variance (in this case being able to play different characters in the same match) will generally lead to a higher probability of volatility.

Seeing as blind picking is enforced (not saying it shouldn't be, just stating a current fact) by this rule-set this effectively means each player choosing their characters is going to be random seeing as they have no way of knowing what their opponent will play and let's assume they are of equal skill.

Now in a normal set where they can only chose 1 character for the first game and only that character the variance is rather low (comparably). So like presume player A is playing character X and character X has like 5 bad MUs out of 41 character. This means there is a low chance of Player B's character (character Y) being one of those 5 characters.

However in your method if player A was a character specialist and only played character X but player B played more the one character the chance of him using one of those 5 characters becomes a bit higher.

I of course kept the examples simple (but they are not the meat of what I'm trying to say, only to show what I mean) but the general gist is that your proposed method has a higher variance to it which causes a higher chance of volatility withing sets/matches.

Also what Foo said.

You add to the probability of hitting a larger matchup but fail to subtract the decrease in stakes. If you hit a bad matchup you will lose a stock. You have to hit 4 simultaneous bad matchups in order to MATCH the effect of playing one bad matchup in the current ruleset. So yes, you're more likely to find a bad matchup, but so is the enemy, and ultimately you're less likely to lose because of that bad matchup.

To think of it differently: in pokemon you have good matchups and bad matchups. Playing the game with one pokemon 4 times would be stupid because if you chose charmander and they chose squirtle you've lost and that's it.

however if you chose 4 different pokemon that have skills that compliment each other and diverse weaknesses your team will be far stronger than chosing the same character 4 times. You are almost certain to find a bad matchup, but his is part of the plan, and it ends up being more about playing the characters intelligently rather than you losing because they chose your counter.

So no, what you are saying is the opposite of what is mathematically happening here. I just think you don't understand what an increase in variable does in the game of averages (hint, it drives towards the middle, it does not drive towards extremes)

Because it's exactly the same, but incredibly stupid unless everyone masters the characters they are playing which isn't possible. Since everyone will be worse at their characters they won't know how to work around bad matchups which makes the matchups matter that much more.

At this point, you should be telling us why we should so this. It doesn't mitigate the rock-paper-scissors element at all, because not starting with four of the same character is just kinda stupid in blind pick (unless you know your opponent always runs a certain order) and in game 3 counters would either work exactly the same, or be really stupid.



Because the only time you'd use it is to counterpick more. Counterpicking isn't actually a good thing. It's a necessary evil. You want the game to be as close to skill vs skill with each person's best character. If people are picking up more and more characters to counterpick more with this system, that means the game is being focused more on matchups than peak skill. It doesn't make matchups more or less volatile, it just emphasizes their strategic use.

Few things: Your statement that it is impossible to master four characters is wrong. People do it in pretty much every other fighting game ever. King of Fighters. Marvel. Street Fighter. Whatever. Doesn't matter, point is, it's done all the time and unleess youre saying smash has characters far far far more difficult to master than any other fighting game I don't get your meaning. However, you are free to only master two characters. This is fine. Even adding a second character into one match will increase the odds of not getting hard countered at character selection. Or be a character specialist if you want, just be ware of the disadvantages you face against those who actually do learn new characters and then stomp your **** because you only learned one character in goddamn project m.

AS for your last paragraph I cannot even wrap my head around it. The reason you would use counterpick rules identical to counterpick rules already established is to counterpick? Well yeah. counterpicks happen one way or the other, just my way it's less likely for the counterpick to matter unless you somehow happen to be a master all of the 4 counterpicks to their characters. Which is like super unlikely. In which case it's equal, not greater, than any old counterpick from days gone by.

Seriously guys, this system isn't new. It's used in tons of fighting games very successfully, and is an actual strategy used to combat the very challenges project m is facing at this moment. This isn't an original idea. I'm just pointing out the problem and an already established solution to that problem.

 

[Zigludo]

I don't see how using a system of 4 characters each would possibly increase the amount of imbalance inherent in counterpicking. The current counterpicking system already incentivizes the counterpicking player to select the most imbalanced matchup (in his favor) that he possibly can.

By increasing the number of characters necessary for each player to play in each match, you de-emphasize the effect of the most polarized matchups by mixing it in with others.

To put it another way: in an ASV match, the number of matchups ranges from 4 (if someone gets 4 stocked) to 8 (if it's a close game). The average value of any given set of 8 matchups must, by definition, be less extreme than the most lopsided matchup within that set. Furthermore, the matchup spread of any given character with any 4 characters must, by definition, be less extreme than that character's single most lopsided matchup, which is the matchup that the current system incentivizes counterpicking with.

In a sense, you decrease the effect that outliers have on the outcome of the match by increasing the "sample size" (not an entirely accurate term in this case but it works for the metaphor) of the set, which has a normalizing (albeit somewhat volatility-inducing) effect on the flow of the game.

Yes, it increases variance, but it also reduces imbalance without introducing any random factors or elements outside of the players' control.

There's no way that 4v4 counterpicks are more unbalanced than 1v1 counterpicks. Even if you got each of your characters perfectly counterpicked, the average spread would be closer to even than the worst matchup within that set. Mathematically, it's a very sound idea, and I think that if you assume (as some do) Project M to be a game that is designed to disallow dominance with a single character, then this is a valid method of mitigating the effect of drastic counterpicks.

 

[Shellfire]

There is an obnoxious overusage of hyperbole in this thread.

 

[Foo]

You add to the probability of hitting a larger matchup but fail to subtract the decrease in stakes. If you hit a bad matchup you will lose a stock. You have to hit 4 simultaneous bad matchups in order to MATCH the effect of playing one bad matchup in the current ruleset. So yes, you're more likely to find a bad matchup, but so is the enemy, and ultimately you're less likely to lose because of that bad matchup.

To think of it differently: in pokemon you have good matchups and bad matchups. Playing the game with one pokemon 4 times would be stupid because if you chose charmander and they chose squirtle you've lost and that's it.

however if you chose 4 different pokemon that have skills that compliment each other and diverse weaknesses your team will be far stronger than chosing the same character 4 times. You are almost certain to find a bad matchup, but his is part of the plan, and it ends up being more about playing the characters intelligently rather than you losing because they chose your counter.

So no, what you are saying is the opposite of what is mathematically happening here. I just think you don't understand what an increase in variable does in the game of averages (hint, it drives towards the middle, it does not drive towards extremes)



Few things: Your statement that it is impossible to master four characters is wrong. People do it in pretty much every other fighting game ever. King of Fighters. Marvel. Street Fighter. Whatever. Doesn't matter, point is, it's done all the time and unleess youre saying smash has characters far far far more difficult to master than any other fighting game I don't get your meaning. However, you are free to only master two characters. This is fine. Even adding a second character into one match will increase the odds of not getting hard countered at character selection. Or be a character specialist if you want, just be ware of the disadvantages you face against those who actually do learn new characters and then stomp your **** because you only learned one character in goddamn project m.

AS for your last paragraph I cannot even wrap my head around it. The reason you would use counterpick rules identical to counterpick rules already established is to counterpick? Well yeah. counterpicks happen one way or the other, just my way it's less likely for the counterpick to matter unless you somehow happen to be a master all of the 4 counterpicks to their characters. Which is like super unlikely. In which case it's equal, not greater, than any old counterpick from days gone by.

Seriously guys, this system isn't new. It's used in tons of fighting games very successfully, and is an actual strategy used to combat the very challenges project m is facing at this moment. This isn't an original idea. I'm just pointing out the problem and an already established solution to that problem.

It is impossible to do in smash because each character controls so differently and doing the exact same things with two characters require vastly different input timings. In many traditional fighters, people play many characters, yes, but hardly anyone does so in PM or melee. Since it is theoretically advantageous to have several characters under your belt, don't you think more people would do it if it didn't mean the had to be less good at each character?

I'm sorry you can't wrap your head around my other point, it's really not complicated. The only reason to choose multiple characters in a game instead of just one is to counterpick. Therefore anyone taking advantage of you system is doing so in order to counterpick. That means your system emphasizes counterpicking as a strategy to win. As I've said, I think the less relevant counterpicking is, and the more people just play one character as well as they can, the better for the game.

 

[Shellfire]

You know, Melee is still heavily influenced by matchups. Did you forget about 20XX? The real reason Melee doesn't have a counterpick problem is because if someone doesn't like volatility they can just pick Fox. Mango, Armada, Leffen, and Hax have all made Fox their primary character because they were sick of the RPSiness of matchups.

 

[AuraMaudeGone]

It is impossible to do in smash because each character controls so differently and doing the exact same things with two characters require vastly different input timings. In many traditional fighters, people play many characters, yes, but hardly anyone does so in PM or melee. Since it is theoretically advantageous to have several characters under your belt, don't you think more people would do it if it didn't mean the had to be less good at each character?

I'm sorry you can't wrap your head around my other point, it's really not complicated. The only reason to choose multiple characters in a game instead of just one is to counterpick. Therefore anyone taking advantage of you system is doing so in order to counterpick. That means your system emphasizes counterpicking as a strategy to win. As I've said, I think the less relevant counterpicking is, and the more people just play one character as well as they can, the better for the game.

Gonna chime in real quick. You're really overreacting in this thread and the bolded made little to no sense.
Have you played the games he mentioned? Sure, they're balanced with teams in mind, but counter picking is not a reliable or central strategy at all. That is what he's getting at. It's actually harder to counter pick, assuming blind picks are involved.

 

[Zigludo]

I'm sorry you can't wrap your head around my other point, it's really not complicated. The only reason to choose multiple characters in a game instead of just one is to counterpick. Therefore anyone taking advantage of you system is doing so in order to counterpick. That means your system emphasizes counterpicking as a strategy to win. As I've said, I think the less relevant counterpicking is, and the more people just play one character as well as they can, the better for the game.

ASV emphasizes counterpicking to a lesser degree than the current system does lol

 

[frankxthexbunny]

It is impossible to do in smash because each character controls so differently and doing the exact same things with two characters require vastly different input timings. In many traditional fighters, people play many characters, yes, but hardly anyone does so in PM or melee. Since it is theoretically advantageous to have several characters under your belt, don't you think more people would do it if it didn't mean the had to be less good at each character?

I'm sorry you can't wrap your head around my other point, it's really not complicated. The only reason to choose multiple characters in a game instead of just one is to counterpick. Therefore anyone taking advantage of you system is doing so in order to counterpick. That means your system emphasizes counterpicking as a strategy to win. As I've said, I think the less relevant counterpicking is, and the more people just play one character as well as they can, the better for the game.

Your mistake is thinking that ignoring counterpicks and pushing through is a better balancing strategy in a game with 41 matchups per character than accepting that counterpicks matter and eliminating the consequences of counterpicking through simple math. Like really really simple math. Add more matchups, it matters less that you got counterpicked for one of those matchups. Ez

 

[AuraMaudeGone]

Thinking about this more, this ruleset may need its own stage list. Would you form a team based on a few stage picks or sit down it try to build a team where a stage doesn't matter much?

 

[frankxthexbunny]

Thinking about this more, this ruleset may need its own stage list. Would you form a team based on a few stage picks or sit down it try to build a team where a stage doesn't matter much?

I dunno it might have it's own strategy, like you can build your team to be really really good at certain stages or you can build your team to be more flexible. I am not sure at all whether a team having increased flexibility cna use a wider range of stages, or it would be the same. I think the base should be umbreon's at first, and the TO should record people's feelings about that stage list after testing.

I've also considered, though I don't want to make this like a thing because people would lose their minds, that possibly if character select is longer (which I doubt it will be that much longer most the time) and there is some mathematical anti-counterpick rule where odd numbers are superior that I completely don't understand, or that character switches slow down matches in general, experimenting with a 3 stock rule set. I'm pretty sure some All Star Mode tourneys do it. I'd like to shoot for 4 just because I thing the game is well balanced for it, but I'm just saying it's something that could be considered

EDIT: On a similar note I'm experimenting with the ABBA method of counterpick character choice, but i'm worried that it's a bit annoying and complicated. Or maybe I'm overthinking and it's a good idea

 

[Bleck]

the smarter solution is to just remove counterpicking rules entirely; you pick a character and that's the character you play for that match, if not the entire tourney