The North Carolina Melee Power Rankings! Updated 8/14/14!

[lord karn]

i don't think there's any data that would put adam above me if he only went to one tournament during this period and has been inactive recently.


and not to sound like lucas or anything, but who the **** is esam





it means somebody with the same name as me lost to a monkey

that ain't me

For some reason I thought you lost to someone lower on the rankings at the ncsu you went to in winners, whereas adam beat everyone but pp at the one he went to.

I also have no idea who you beat oos, so that very well may boost you above him.

 

[Lightsyde]

oversimplifying

sometimes unnecessary

This.

People who think Marth dittos are only about Fthrow/Fsmash =/= People who understand Marth dittos

 

[Divinokage]

I don't think that combo should even work unless the Marth DI inside. lol.

 

[0Room]

If he DIs away no, it won't work.

 

[Divinokage]

If he DIs away no, it won't work.

I learned that the hard way when Marth's used to Fthrow ->> Dair near the ledge, I'm like alright I HAVE to DI away. Ugh.. lol. Just seeing my Ganon drop down to the abyss everytime, at some point it was enough..

 

[Lightsyde]

The beauty of throw mindgames in Marth dittos is part of what makes them so good. If you DI in at some percents or don't DI you get tippered, if you DI hard away you can get dash attacked -> Utilt etc. But other percents make DI'ing in and taking a weak Fsmash better than DI'ing out and taking a dash attack that leads to a combo.

That doesn't even get into the use of wave dash/dash dance pivot grabs which make the correct survival DI for a right facing Fthrow the absolute worst DI for a left facing Fthrow. Not to mention all the tricks with Upthrow -> react to jumps or baited aerials or Bthrow -> Fsmash at later percents or Bthrow -> FH Bair if they jump -> Fair -> Dair etc. Then regrabs sequences where Fthrow -> Fthrow forces a DI out (for a dash attack) or causes them to roll or bait their move, or Fthrow -> Fthrow -> Fthrow allowing variations for their changing DI, or alternating in Uthrow in different ways in those throw chains. And Dthrow forced tech chases or Dthrow pop ups at higher percents. All that stuff (and way more) just from grabs. Crazy changes happen in scenarios by changing one move or the timing of one move.

Then the precision spacing and use of the different types of range (Dtilt for fast low, Ftilt for straight out and slight up, Utilt for archs and behind, turnaround Utilt inverted archs), Nairs for baiting spot dodges. And then there's still other stuff like something I saw Mango's Marth do where he just Ftilt somebody out of a jump and ran a huge distance to chase with a tipper Fsmash.

Ugh there's so much stuff to mess with and react with. I want to play Melee bad now. I think Raleigh should come over tonight and play with me. :[

EDIT: Didn't mean to sound 'Know-it-all'ish, I just got really excited thinking about all the sequences that are possible in dittos between a character with such crazy dynamics (grab game/spacing/movement as peaks, weak from below/perfect floatie combo weight as troughs). Melee is so beautiful for that reason....

EDIT2: Kevin, read this part and tell me if you think this is worth actually doing.

Which gets me thinking now, what if you made graphs for each character that measured aspects of that character? A perfect character's chart would look like a pentagon, if there were, say, 5 objective stats (ie: recovery, grab game, hitbox versality, movement, defensive options) to measure things. If you over laid two different character ability charts (for the sake of naming these) you would perfectly see why a match up was good or bad. The only hard part would be objectively assigning numeric values to how good a given character is at a certain attribute. For anyone who played dance games, you can imagine the groove radars which explained attributes of a song in a similar fashion.

I'm about to make one in MSPaint.

 

[Redd]

Who's going to ROM here?

 

[Juno McGrath]

Who's going to ROM here?

that'd be me.

 

[Dark Hart]

oh oh and me

 

[Lightsyde]

Something like this.

Displayed is a mach match-up chart for Marth vs Kirby.

These values were just guesstimates for the sake of demonstration. I would want to use collected data from top players but this is what I've been imagining.

 

[stingers]

You'd need to define what exactly a 1 or 10 in each category would be wouldn't you?

 

[Juno McGrath]

Me and Josh had lunch today and talked extensively about these charts. I honestly think they are amazing. The Tekken community uses them the way we use matchup charts.

 

[Lightsyde]

Alex, you definitely would need a description of what each attribute is defined as, but also a qualitatively agreed upon definition of each number. I obviously can't do that part just yet, or by myself for that matter. This is just a quick demonstration of the idea. You could definitely have more (or less) categories as well.

It could be really helpful for measuring team dynamics too.

 

[Bl@ckChris]

yes, but once those definitions are made, i can definitely see how these would be useful.

it almost creates a matchup chart for each aspect of the matchup. while still allowing for an overall comparison picture. i approve.

 

[Juno McGrath]

You got it, Chris Nicholas

 

[OmniOstrich]

Where do combo potential and KO option weigh in on that chart?

Do the simply add in to grab game and hitbox versatility (meaning that if a grab leads to a KO or combo then you subsequently raise the grab rating; and if the characters moveset is "versatile" enough to easily hit confirm combos, or KO early that makes it stronger).

Just thinking that over in my head I get a really good idea of what a 10/10 in grab/hitbox categories would mean.

 

[AlcyoNite]

Alex, you definitely would need a description of what each attribute is defined as, but also a qualitatively agreed upon definition of each number. I obviously can't do that part just yet, or by myself for that matter. This is just a quick demonstration of the idea. You could definitely have more (or less) categories as well.

It could be really helpful for measuring team dynamics too.

i do actually think doing one for yourself would be valuable in making more tangible sense of how you personally view character strengths and then being able to relate that to other people. this would be, of course, for reasons other than what you're after.

just a thought

 

[Lightsyde]

Chris: Exactly!

I you overlaid any two characters charts you could not only see overall how those characters stack up against each other, but you could also see WHY one character has an advantage in one area and a disadvantage in others.

Adversely, you could lay character's charts overtop of each other and see if they would make a good team, because they would either have things that showed strong correlations with (ie both have amazing recoveries) or where they can supplement each other (ie one has good movement, one has bad movement).

 

[Lightsyde]

i do actually think doing one for yourself would be valuable in making more tangible sense of how you personally view character strengths and then being able to relate that to other people. this would be, of course, for reasons other than what you're after.

just a thought

I was just thinking about this too!

Imagine if you had every top player make a chart for every character. Think about how hilarious the difference between M2K's chart for Puff and Hbox's chart for Puff would be.

You could demonstrate player bias or incorrect judgments about characters based off how much your charts vary from a standardized chart of a character or match up.

Ben: I'll answer your question in just a second. I need to get caught up at work. But based off your middle paragraph, I think you've kinda got the jist of the way I was breaking stuff up.

 

[Dark Hart]

I like the idea

and for me that's what matters

the idea

it's a great one

defining stuff yes

but then [the proverbial] we're still back to arguing how to rate each of these new things

imo

but [the proverbial] we'll figure it out

 

[Lightsyde]

That's what is awesome about this.

ONLY the standardized diagram for any given character, match up or team will require ANY kind of agreement. We can collect data from a panel of players (could be limited down to, for example, PP, M2K, Armada, Amsah, Mango) and use either median or means of the data to best describe things. Alternatively we could collect a much larger sample size and use the median or means of that.

AND/OR we can compare two people who disagree's charts to each other AND to the standardized charts. It could be a really good way to measure deviations and growth in player mindset and overall metagame.

 

[AlcyoNite]

I was just thinking about this too!

Imagine if you had every top player make a chart for every character. Think about how hilarious the difference between M2K's chart for Puff and Hbox's chart for Puff would be.

You could demonstrate player bias or incorrect judgments about characters based off how much your charts vary from a standardized chart of a character or match up.

Ben: I'll answer your question in just a second. I need to get caught up at work. But based off your middle paragraph, I think you've kinda got the jist of the way I was breaking stuff up.

exactly what i was thinking

 

[Dark Hart]

My only problem is getting accurate information, which is impossible since everyone has different opinions, even among top players.

just my 6cents though

 

[DJRome]

the problem with charts is that there is simply no right answer. everything is an opinion and im just not sure why opinions need to be expressed in charts. sure you could say someone's chart shows that he is better as the sum of 5-10 attributes but characters are not the sum of their parts.

 

[Dark Hart]

I guess that's what I wanted to say.....

 

[Lightsyde]

My only problem is getting accurate information, which is impossible since everyone has different opinions, even among top players.

just my 6cents though

Again, this is not a problem because the means or medians of these tables of data would demonstrate the highest level of the current metagame. Consider that with a group 10 top players the data for Marth's Hitbox Versatility looks like:

Hitbox Versatility: 7, 7, 8, 8, 9, 9, 9, 9, 10, 10

The Standardized Value for that stat would show either
Mean: 8.6
Median: 9

Do that for every number and you get the Standardized Syde Diagram for that character.

With that, you can compare and contrast 1 of the panel members opinions. For example, if Kevin submitted that Marth's Hitbox versatility was a 7 it would show a deviation of 1.6 or 2, depending on whether you were looking at the median or mean charts. From that we can draw 1 of 2 conclusions:

1) The current metagame overestimates how good Marth's hitbox versatility is, according to Kevin.
2) Kevin under values Marth's hitbox versatility in this metagame.

the problem with charts is that there is simply no right answer. everything is an opinion and im just not sure why opinions need to be expressed in charts. sure you could say someone's chart shows that he is better as the sum of 5-10 attributes but characters are not the sum of their parts.

Again, it's not about displaying a right answer because like you're getting at, there is not an actual quantitative value of some of those attributes. Opinions of top level players though are helpful for displaying the current metagame. If we collected data from top players every 6 months or every year then we could have a visual representation of the current metagames take on a match up year to year.

And I think I partly disagree with your last sentence. If you managed to make your categories explicit enough to display all the dynamics of a given character, it would be a very decent representation of what a character is like, especially for the sake of comparing them to other characters. Would those attributes perfectly display a character? Maybe, maybe not, but it can't hurt to find out and it can't hurt to make them available as a learning tool for different levels of players. They would also be super helpful for teaching/displaying perspective and mindsets.

EDIT: Another thing that is cool about this is that these charts would lead to you being able to display match up ratios.

In the Marth Kirby instance I demo'd for the graph, Marth's total attribute points were 41, Kirby's were 22. 41 + 22 = 63

42/63 = ~.0650
21/63 = ~.0349

The match up ratio for my example is that Marth Kirby is 65/35. If this seems erroneous or varies greatly from the accepted standard of a match up, then it's because I was either too generous when assigning values to Marth, not generous enough with Kirby, or that the old ratio is bad.

 

[Dorsey]

aside from attaining accurate information........

IMO this is like a commentator using stats to correlate with wins/losses. It's always easy for them to make the correlation AFTER the game "Yes, it makes sense because they had more 3rd down completions and passing/running yards." , when although true, it really came down to some more crucial/defining moments of the game. I emphasized after because stats have been proven not to be a super consistent measuring tool, even if right the majority of the time. This is why there are so many upsets in sports(statistical upsets). Time is constant in these charts, in regard to both characters using their time 100% efficiently and approaching from stalemate in the best way possible, whether to estimated maximum character potential or the current overall meta-game(additionally stages are factored in too I guess?). You could definitely use it as something to reflect on matches that have already happened, as the commentator does in my analogy... but how good is a chart like this for actually measuring the characters currently in smash bros.?

It's a cool chart, but that's how I look at it anyway.

 

[Dark Hart]

Which gets me thinking now, what if you made graphs for each character that measured aspects of that character? A perfect character's chart would look like a pentagon, if there were, say, 5 objective stats (ie: recovery, grab game, hitbox versality, movement, defensive options) to measure things. If you over laid two different character ability charts (for the sake of naming these) you would perfectly see why a match up was good or bad. The only hard part would be objectively assigning numeric values to how good a given character is at a certain attribute.

but you agreed with me 5 hours ago?

also I don't like that mean median idea since it's not... exact... (not that it's a bad idea)... which is what we need for this to practically work...

but if it's not about getting exact numbers (like you just said to DJ) then I don't even think this is a good idea anymore

since it's all about those numbers being correct [imo]

but I'm still interested in this...

 

[Lightsyde]

Read the edit(s) to my last post Lucas.

And I wasn't saying you were wrong, I do think it will be difficult to define everything correctly but I don't think it's unreasonable or impossible. It will just require an agreed upon set of categories and a good explanation of what every numeric value of a attribute means.

In what ways are neither mean or median not exact if it displays both? They would give two solid numbers that display a Standard opinion and a player specific opinion.

 

[RaphaelRobo]

If you're going to have both the mean and the median, I'd suggest adding the mode too. It's not quite as useful for this situation, but it'd be a nice piece of information to have.

While it's not perfect, nothing is, and I think this would be a great way to show the strengths and weaknesses of different characters. It'd be useful for someone new, like me, to have a quick way to reference what each character is good and bad at. Like I said, it's not perfect, but as long as the reader understands that then it should be fine. If they don't understand it, I think they'll work it out pretty quickly.

 

[Dark Hart]

Read the edit(s) to my last post Lucas.

And I wasn't saying you were wrong, I do think it will be difficult to define everything correctly but I don't think it's unreasonable or impossible. It will just require an agreed upon set of categories and a good explanation of what every numeric value of a attribute means.

In what ways are neither mean or median not exact if it displays both? They would give two solid numbers that display a Standard opinion and a player specific opinion.

this is why I don't agree with what I don't agree with

but you should really stop trying to explain it to me since I'm not even smart enough to do anything other than point stuff out and ask questions

I'm not worth it

 

[Bl@ckChris]

off the top of my head, i would say that the same people who had the authority to decide the tier list may be able to come up with a solid idea of how this could all fit together.

i think the matchup ratio concept will be very good as well. it plays into the idea of how if different characters control different aspects of the matchup, they will get more or less points in that region.

it may also show incoming players what parts of their game to focus on in a particular matchup, although in reality all points are most certainly important. But if it's evident that a character has a much stronger grab game than their opponent, it may be able to show a point that is abusable.

similar to the matchup chart, it may be something left for each of the character boards to decide. but i'm not sure how i feel about that either. The ganon boards have a much similar mindset to each other and how our matchups are and how our character is than most other characters do, i imagine.

 

[Lightsyde]

If you're going to have both the mean and the median, I'd suggest adding the mode too. It's not quite as useful for this situation, but it'd be a nice piece of information to have.

While it's not perfect, nothing is, and I think this would be a great way to show the strengths and weaknesses of different characters. It'd be useful for someone new, like me, to have a quick way to reference what each character is good and bad at. Like I said, it's not perfect, but as long as the reader understands that then it should be fine. If they don't understand it, I think they'll work it out pretty quickly.

The bolded part is exactly what is kinda driving me to work on this. It could help people learn so much faster and could settle a lot of debate.

And I think you're right about mode. If we can display means and medians, we could display mode too. Depending on what a player thinks is most accurate (or what makes most sense to them) will determine which sorting method they reference.

this is why I don't agree with what I don't agree with

but you should really stop trying to explain it to me since I'm not even smart enough to do anything other than point stuff out and ask questions

I'm not worth it

And I disagree! I think that is your biggest misunderstanding of me. I DO think you're worth explaining to, or else I wouldn't bother continuing to explain further. :]

 

[Dark Hart]

actually Josh I think we both understand that we disagree on that part

lol

no misunderstanding currently XD

 

[Lightsyde]

Lmao that sentence was soooooo confusing, kinda loved it.

Smashfest tonight, if you didn't see it in the other thread. Please honor me with your presence and your ridiculous characters at my home this evening.

 

[bossa nova ♪]

updated 11/10/11

1. me
2-10. n****z

 

[Lightsyde]

Cam, read all my chart stuff, you goof.

And find your way to Raleigh tonight for my Smashfest.

Also, I more or less agree with your update. I don't know why the panel has such a hard time realizing that. >________>

 

[bossa nova ♪]

****. i literally JUST missed my chance at a ride.


next time bro

 

[Dr Peepee]

LOOOOOOL twitch you salty mofo

man I can't wait to play you again

 

[$mike]

For some reason I thought you lost to someone lower on the rankings at the ncsu you went to in winners, whereas adam beat everyone but pp at the one he went to.

I also have no idea who you beat oos, so that very well may boost you above him.

at that ncsu, pp is the only person i lost to (in winners and GFs). i didn't drop a game to anyone else in pools or bracket.



@ josh

omw babe