5150
Banned via Administration
What Sakurai is Doing: An Essay Regarding the Possibilty of Multi-Dimensional Tiers
Version 2.0
Prelude: I'd like it to be known that I do not fully agree with sakurai and what he's doing but this is what I see him to be doing.
*WARNING*
Do not read this if:
-You do not believe in tiers.
-You do not accept that brawl is less technical than melee.
I don't have time for your comments if that's what you believe and I will ignore your post.
Ok, so there are three smash bros games. For each game you can see a rather visible progression of Sakurai's ideas and intentions being put into a game.
Smash bros 64
This game is VERY basic. The entire game can be summarized down to two basic ideas.
#1 The characters are roughly the same in terms of movement (due to same fall speed and near same running speed), but their moves are different.
#2 There is nearly 0 DI, so a character's move-set can be ranked very accurately in terms of worth based on what combos the character possess due to that move-set.
This translated to a tier list that was very polarized. Kirby/Pikachu were quite clearly amazing while characters like link and dk were quite clearly terrible. This is all due to the only major difference being move-set. Although isai was able to make all of the characters amazing with platform cancelling no one else has even come close to what hes able to do.
Smash bros melee
This game is where Sakurai took 64 to a whole new level and tried to introduce more character diversity (despite the addition of clones). He did this by making vague character types, by manipulating the character's size, weight, and speed properties, rather than just the move-set. Here is a rough idea of what he was GOING for.
#1 There are 3 basic type of characters: Power, Speed, and Control characters. Each type reflects different playstyles. Certain types of characters are more inclined to win against different character types. For reference: Power beats speed, speed beats control, and control beats power. It's rock, paper, scissors. There also was an increased variability of defensive techniques and defensive flaws. Some people's recovery became incredibly good and some incredibly bad (not different from 64, but there was still a lot more in-between rather than simply having defense be a measure of how good their upb was). The shield was changed so that certain character's shields were terrible.
#2 There is a good amount of DI, so combos are much more specific and MOST of the time done on the fly.
#3 There are more advanced techs.
This is what happened wrong. Sakurai wanted to balance certain characters he felt lacking by giving them certain moves that had traits of other character types. this is a perfectly fine idea and is how you balance a game. For example, fox's usmash is one of the strongest moves in the game, yet he is one of the fastest characters. Sakurai thought without this kill move option he would not be good enough. He did this for many characters. But what happened he couldn't expect. We took his game and broke his intentional balance changes by abusing many cancels and exploits (like wd and jump-cancelling, etc) that were found to be in the game. So what happened is:
#1 You had types of characters that were doing EVERYTHING, not just their intended playstyle. You had speed characters that were able to move even faster, had great control, and had crazy power. An example is fox. He is super fast, can control using the shine, and has crazy strong kill moves including usmash and uair. You had control characters like shiek being so strong and fast that she became top tier. The advanced techs were meant to increase depth, which they did, but increased the depth for only these hybrid, amazing characters. A good indication, in this game, of how good a character is is how easy it can gimp, or rather kill at a low percent. If it can gimp, then it is most likely a hybrid character.
#2 You had types of characters that did NOTHING, not even their intended playstyle. You had power characters like bowser become absolutely useless because his strength wasn't even close to as good as fox's, when you look at it as a speed : power ratio.
The dimensions are, power, control, defense, and speed and a characters effectiveness/worth is its area. Thus the tier list is a ranking of areas. Character's area can change by changing the different rankings. Shiek used to have the most area, and thus worth, but now fox does. How did this happen?
-player picks up character x
-player uses a move that EVERYONE has access to but uses it differently and better than before
-player does better in tournaments
-people notice
-tier graph is changed
now the question is, what KIND of move was changed? was it a grab? the fsmash? a wierd b move? all of those moves can be classified to strong, fast, or control. so what we would essentially do is give the character a better ranking IN THE AREA THAT WAS IMPROVED due to someone using a move better.
Hypothetical brawl example:
-we figure out a **** use for dedede's downb
-we increase his power rating
-his overall AREA of his graph goes up, and therefore he moves up on the tier list
or a melee example:
-we learn that mewtwo can use his upb to teleport around with near 0 lag
-we increase his defense rating
-his overall AREA of his graph goes up, although only slightly, possibly increasing his tier position (although for some reason we have pichu above him :/)
another melee example:
-fox players (mainly the ec style) started getting very good and started learning how to control their characters very well and precisely (like m2k, kdj, etc).
-Fox's control and speed ranking go up
-his overall AREA of his graph goes up, and therefore he moves up on the tier list
now the ratings dont have to have a limit, they just have to reflect how MUCH of a distance character x is to character y in that type. So we have COMPLETE control over what values characters get, and as long as we make the ratios proper we won't have a problem.
Smash bros brawl
This time sakurai realized that the advanced techs destroyed his balance, and created a game of 5 characters. So what did he do in brawl:
#1 Made the fast characters weaker (example: fox)
#2 Made the strong characters stronger (example: ike, bowser)
#3 Added superarmor for the heavy characters (example: ike, bowser)
#4 Made the control characters control better but kill worse (example: metaknight)
#5 Increased the effect of DI, only increasing the favorability towards strength characters for balance.
#6 Increased the recovery of most characters.
#7 Almost all advanced techs that we KNOW OF have been removed or dumbed down (made automatic or super easy) in order to keep tighter control on the balance.
So this is how the tier list is gonna be. There's going to be a few broken characters, like melee, on the top. I predict peach and ice climbers. From then on the difference in worth of characters will become progressively smaller. Most characters, if ranked, will be grouped into mid or high tier and only a handful are bottom tier quality. But the most important part of the tier list is that it will be more 2 dimensional. I'm not happy that the game is less technical. I'm personally against that as you can easily balance a game WITH more advanced techs, but sakurai realizes that it would be much too hard for him to do as he'd have to keep track of tournaments and actually attend if he wanted to see the game evolve and try to balance it with tournaments in mind. And I completely understand that.
So maybe it might be better to have the tier list be 2 dimensional, rather than 1, in order to show how effective a character is at what it does as well as how good it is OVERALL like the current tier list does? It'd look like quadrilaterals and would be weird at first but is better than what we are using for melee as I think the REAL measurement will be “how effective is this character at doing TRAIT X” where TRAIT X = power, speed, defense, or control. This game will be much more about strategy and counterpicks than melee EVER was.
It also supports the idea that extreme characters dont get a high amount of area and therefore worth. If you stretch a quadrilateral only using one vertex you will get a very thin quadrilateral which has a very small area. however if you stretch it by less on each vertex you will get more area. this translates to "a character that is only good in one thing is not very good overall". and i think this holds true for melee as a real world observation. fox's area is VERY HIGH because all of his factors are "stretched". whereas bowser, who is pretty strong, overall stronger than fox, has much less area because his speed and control ratings are so low.
Here's an example of such, with fabricated "guess" values:
Continuing with this idea, what if these 2d tier graphs could be plotted when you and your partner select your characters for teams. The objective MIGHT be for the perect team to try and cover the most free space with your character object. this means that characters who were really good at control and decent on speed would work perfectly with a character that was terrible at control but has decent speed and is very strong. those characters would fit perfectly together both on the graph and in the actual game because the control character would help set up for heavy hits. if you think about it this idea of 2d tier graphs could REVOLUTIONIZE how we predict matchup wins down to a scientific formula. we could observe patterns between what set of tier space their tier object inhabits and how well that type of space does against an opposite side of tier space, or any other location for that matter. this is applicable for both teams AND singles.
This is an open hypothetical question, but what if you could find the center of mass of a character's object and measure the distance of it to another character's object's center of mass and figure out either an positive or negative relationship. Either the closer they are the closer the matchup or the farther they are the more **** the matchup. Which way the **** is "going" is dependent on what kind of space the character objects occupy. if it's a control type then the general prediction is the farther away a control type object's center of mass is to a power character's center of mass the better the control character would do. so the top tier characters that excell everywhere, their center of mass will be very close no matter what type of character they face.
for that matter, should this center of mass be called a "tier coordinate"?
I will be heading up a Tier Council that will start to experiment with this by adding values and changing formulas till we get a conclusive or inconclusive “it works”, “it doesn't work”, or “it might work”. If you would like to be apart of the council send me a pm or aim me, however it is usually invite only so don't expect much.
The Tier Council
5150
Juke
BigD
Mookierah
Crashman
Glossary
Trait = Speed, power, control, and defense. Each trait encompasses many elements of the game and can be looked up earlier in the essay.
Tier Graph = The graph that shows how well a character does by measuring the 4 traits of a character.
Tier Object / Character Object = The quadrilateral that shows a character's stats on a Tier Graph.
Tier Space = The space on the Tier Graph
Tier Value = The area of a character's Tier Object. Compiling a list of these values and ordering them would create a 1d tier list like we have now.
Tier Coordinate = The center of mass of a Tier Object. May have uses later, but for now is completely theoretical in its application.
Version 2.0
Prelude: I'd like it to be known that I do not fully agree with sakurai and what he's doing but this is what I see him to be doing.
*WARNING*
Do not read this if:
-You do not believe in tiers.
-You do not accept that brawl is less technical than melee.
I don't have time for your comments if that's what you believe and I will ignore your post.
Ok, so there are three smash bros games. For each game you can see a rather visible progression of Sakurai's ideas and intentions being put into a game.
Smash bros 64
This game is VERY basic. The entire game can be summarized down to two basic ideas.
#1 The characters are roughly the same in terms of movement (due to same fall speed and near same running speed), but their moves are different.
#2 There is nearly 0 DI, so a character's move-set can be ranked very accurately in terms of worth based on what combos the character possess due to that move-set.
This translated to a tier list that was very polarized. Kirby/Pikachu were quite clearly amazing while characters like link and dk were quite clearly terrible. This is all due to the only major difference being move-set. Although isai was able to make all of the characters amazing with platform cancelling no one else has even come close to what hes able to do.
Smash bros melee
This game is where Sakurai took 64 to a whole new level and tried to introduce more character diversity (despite the addition of clones). He did this by making vague character types, by manipulating the character's size, weight, and speed properties, rather than just the move-set. Here is a rough idea of what he was GOING for.
#1 There are 3 basic type of characters: Power, Speed, and Control characters. Each type reflects different playstyles. Certain types of characters are more inclined to win against different character types. For reference: Power beats speed, speed beats control, and control beats power. It's rock, paper, scissors. There also was an increased variability of defensive techniques and defensive flaws. Some people's recovery became incredibly good and some incredibly bad (not different from 64, but there was still a lot more in-between rather than simply having defense be a measure of how good their upb was). The shield was changed so that certain character's shields were terrible.
#2 There is a good amount of DI, so combos are much more specific and MOST of the time done on the fly.
#3 There are more advanced techs.
This is what happened wrong. Sakurai wanted to balance certain characters he felt lacking by giving them certain moves that had traits of other character types. this is a perfectly fine idea and is how you balance a game. For example, fox's usmash is one of the strongest moves in the game, yet he is one of the fastest characters. Sakurai thought without this kill move option he would not be good enough. He did this for many characters. But what happened he couldn't expect. We took his game and broke his intentional balance changes by abusing many cancels and exploits (like wd and jump-cancelling, etc) that were found to be in the game. So what happened is:
#1 You had types of characters that were doing EVERYTHING, not just their intended playstyle. You had speed characters that were able to move even faster, had great control, and had crazy power. An example is fox. He is super fast, can control using the shine, and has crazy strong kill moves including usmash and uair. You had control characters like shiek being so strong and fast that she became top tier. The advanced techs were meant to increase depth, which they did, but increased the depth for only these hybrid, amazing characters. A good indication, in this game, of how good a character is is how easy it can gimp, or rather kill at a low percent. If it can gimp, then it is most likely a hybrid character.
#2 You had types of characters that did NOTHING, not even their intended playstyle. You had power characters like bowser become absolutely useless because his strength wasn't even close to as good as fox's, when you look at it as a speed : power ratio.
It's a 2 dimensional graph, really.for those people who dont understand the attributes and what they mean by now:
-speed = the overall ability to move fast and attack fast. This is governed by running speed, wd length, dash dance length, startup time, lag, and fall speed.
-power = the character's overall killing ability. This is ultimately how strong their moves are, but also applies to gimp ko's, such as shine ko's or spikes. This also includes guaranteed combos.
-control = the ability to prevent the opponent from doing what they want and the ability to force them to do what you want. The most common way of doing this is ranged moves and grabs, but applies to makeshift combos as well (consecutive hits), that only occur because the character was able to put them into such a bad situation.
-defense = the ability to prevent you from being hit and/or dieing. This includes but is not limited to your recovery, your weight, your fall speed, your shield size, your dodge speed, your air dodge speed, your roll speed, etc. (defense is not apart of the rock-paper-scissors system, but is important to measure because certain characters ARE lower on the tier list simply because their defense is so terrible and the opposite is true as well)
The dimensions are, power, control, defense, and speed and a characters effectiveness/worth is its area. Thus the tier list is a ranking of areas. Character's area can change by changing the different rankings. Shiek used to have the most area, and thus worth, but now fox does. How did this happen?
-player picks up character x
-player uses a move that EVERYONE has access to but uses it differently and better than before
-player does better in tournaments
-people notice
-tier graph is changed
now the question is, what KIND of move was changed? was it a grab? the fsmash? a wierd b move? all of those moves can be classified to strong, fast, or control. so what we would essentially do is give the character a better ranking IN THE AREA THAT WAS IMPROVED due to someone using a move better.
Hypothetical brawl example:
-we figure out a **** use for dedede's downb
-we increase his power rating
-his overall AREA of his graph goes up, and therefore he moves up on the tier list
or a melee example:
-we learn that mewtwo can use his upb to teleport around with near 0 lag
-we increase his defense rating
-his overall AREA of his graph goes up, although only slightly, possibly increasing his tier position (although for some reason we have pichu above him :/)
another melee example:
-fox players (mainly the ec style) started getting very good and started learning how to control their characters very well and precisely (like m2k, kdj, etc).
-Fox's control and speed ranking go up
-his overall AREA of his graph goes up, and therefore he moves up on the tier list
now the ratings dont have to have a limit, they just have to reflect how MUCH of a distance character x is to character y in that type. So we have COMPLETE control over what values characters get, and as long as we make the ratios proper we won't have a problem.
Smash bros brawl
This time sakurai realized that the advanced techs destroyed his balance, and created a game of 5 characters. So what did he do in brawl:
#1 Made the fast characters weaker (example: fox)
#2 Made the strong characters stronger (example: ike, bowser)
#3 Added superarmor for the heavy characters (example: ike, bowser)
#4 Made the control characters control better but kill worse (example: metaknight)
#5 Increased the effect of DI, only increasing the favorability towards strength characters for balance.
#6 Increased the recovery of most characters.
#7 Almost all advanced techs that we KNOW OF have been removed or dumbed down (made automatic or super easy) in order to keep tighter control on the balance.
So this is how the tier list is gonna be. There's going to be a few broken characters, like melee, on the top. I predict peach and ice climbers. From then on the difference in worth of characters will become progressively smaller. Most characters, if ranked, will be grouped into mid or high tier and only a handful are bottom tier quality. But the most important part of the tier list is that it will be more 2 dimensional. I'm not happy that the game is less technical. I'm personally against that as you can easily balance a game WITH more advanced techs, but sakurai realizes that it would be much too hard for him to do as he'd have to keep track of tournaments and actually attend if he wanted to see the game evolve and try to balance it with tournaments in mind. And I completely understand that.
So maybe it might be better to have the tier list be 2 dimensional, rather than 1, in order to show how effective a character is at what it does as well as how good it is OVERALL like the current tier list does? It'd look like quadrilaterals and would be weird at first but is better than what we are using for melee as I think the REAL measurement will be “how effective is this character at doing TRAIT X” where TRAIT X = power, speed, defense, or control. This game will be much more about strategy and counterpicks than melee EVER was.
It also supports the idea that extreme characters dont get a high amount of area and therefore worth. If you stretch a quadrilateral only using one vertex you will get a very thin quadrilateral which has a very small area. however if you stretch it by less on each vertex you will get more area. this translates to "a character that is only good in one thing is not very good overall". and i think this holds true for melee as a real world observation. fox's area is VERY HIGH because all of his factors are "stretched". whereas bowser, who is pretty strong, overall stronger than fox, has much less area because his speed and control ratings are so low.
Here's an example of such, with fabricated "guess" values:
Continuing with this idea, what if these 2d tier graphs could be plotted when you and your partner select your characters for teams. The objective MIGHT be for the perect team to try and cover the most free space with your character object. this means that characters who were really good at control and decent on speed would work perfectly with a character that was terrible at control but has decent speed and is very strong. those characters would fit perfectly together both on the graph and in the actual game because the control character would help set up for heavy hits. if you think about it this idea of 2d tier graphs could REVOLUTIONIZE how we predict matchup wins down to a scientific formula. we could observe patterns between what set of tier space their tier object inhabits and how well that type of space does against an opposite side of tier space, or any other location for that matter. this is applicable for both teams AND singles.
This is an open hypothetical question, but what if you could find the center of mass of a character's object and measure the distance of it to another character's object's center of mass and figure out either an positive or negative relationship. Either the closer they are the closer the matchup or the farther they are the more **** the matchup. Which way the **** is "going" is dependent on what kind of space the character objects occupy. if it's a control type then the general prediction is the farther away a control type object's center of mass is to a power character's center of mass the better the control character would do. so the top tier characters that excell everywhere, their center of mass will be very close no matter what type of character they face.
for that matter, should this center of mass be called a "tier coordinate"?
I will be heading up a Tier Council that will start to experiment with this by adding values and changing formulas till we get a conclusive or inconclusive “it works”, “it doesn't work”, or “it might work”. If you would like to be apart of the council send me a pm or aim me, however it is usually invite only so don't expect much.
The Tier Council
5150
Juke
BigD
Mookierah
Crashman
Glossary
Trait = Speed, power, control, and defense. Each trait encompasses many elements of the game and can be looked up earlier in the essay.
Tier Graph = The graph that shows how well a character does by measuring the 4 traits of a character.
Tier Object / Character Object = The quadrilateral that shows a character's stats on a Tier Graph.
Tier Space = The space on the Tier Graph
Tier Value = The area of a character's Tier Object. Compiling a list of these values and ordering them would create a 1d tier list like we have now.
Tier Coordinate = The center of mass of a Tier Object. May have uses later, but for now is completely theoretical in its application.