Let Us Discuss Match-Up Numbers

[Crystanium]

When it comes to match-ups between this character and that character, I see people saying things like, "This character has a slight advantage. I'd say 60-40." I read somewhere that these numbers represent percent. So, for example, character A will beat character B 60% of the time. It doesn't mean that B can't win. It just means that B won't win as much. So, unless I am ignorant about all of this and if there is a thread that actually discusses the numbers, please link it to me. Otherwise, I'd really like to get some insight on what all of this really means and how one even comes to the conclusion of 60-40, it being in A's favor.

 

[Crizthakidd]

im not too sure about how the numbers work also but there aree many things to consider. matchups help determine tiers a lot too. thats why people say MK has no bad matchups or if he has its barely even. so take Mk vs captain falcon the number u get from that matchup will be generated by both = 100

mks attacks stop all of c. falcons so ull give him 10-0 so far
mk has a better recovery - add like 20 .
c.falcon can kil mk early but not too early and his smashes are hard to pull off

u add all these things up and it turns out to be such a bad matchup like 75-25

 

[Crystanium]

im not too sure about how the numbers work also but there aree many things to consider. matchups help determine tiers a lot too. thats why people say MK has no bad matchups or if he has its barely even. so take Mk vs captain falcon the number u get from that matchup will be generated by both = 100

mks attacks stop all of c. falcons so ull give him 10-0 so far
mk has a better recovery - add like 20 .
c.falcon can kil mk early but not too early and his smashes are hard to pull off

u add all these things up and it turns out to be such a bad matchup like 75-25

Well see, how does that work? Give Meta Knight a 10% increase in winning. Why 10? Why not 5? Why not 1? If you were to increase the percent for each attack, then that would go beyond this 0-100, 10-90, 20-80, &c.

 

[Samuelson]

Lets say the matchup is 60:40. If the 2 characters play 100 matches and the 2 players are at equal skill then character A will win 60 matches. It is an estimate and there is no formula for coming up with these numbers.

 

[Crystanium]

Lets say the matchup is 60:40. If the 2 characters play 100 matches and the 2 players are at equal skill then character A will win 60 matches. It is an estimate and there is no formula for coming up with these numbers.

Doesn't that make it a bit arbitrary, then? You see, when it's a ditto match, you will automatically say that character A versus character A is an even match. You'd say it's 50:50. The reason is because both are equal in offense and defense. Both weigh the same, both can use the same tactics, &c. If we have Snake versus Donkey Kong, what is it that will determine one having the match in their favor more than the other? I just find it strange that people will throw out numbers like 55:45 or 60:40. It's especially difficult if the people strongly disagree with the opposing side, only because they like their character. So bias ensues.

 

[MK26]

50-50: Neutral
Characters are evenly matched
Winner will be determined by the skill of the players
Smart counterpicking will give one player the advantage

55-45: Slight Advantage/Neutral
One character has a small advantage
Among players even in skill level, the player with the advantaged character will win slightly more of the time, but tournament sets can go either way
Smart counterpicking will tilt the advantage in favour of the counterpicker

60-40: Advatage
One character has one large or multiple small attributes that gives it the advantage
Among players even in skill level, the player with the advantaged character will win more of the time, and has a tournament set in his favour
Smart counterpicking will even the matchup

70-30: Large advantage
One character has several large attributes that gives it the advantage, or a single attribute that completely shuts down the opposing character
Among players even in skill level, the player with the advantaged character will win most of the time and may sweep a tournament set or three-stock his opponent
Smart counterpicking will decrease the disadvantage, but not even the matchup

I honestly just thought of this^^ right now, by myself

Does it make sense?

 

[Genos]

Yes, it is completely arbitrary. It supposedly comes out to what Samuelson said. But, people typically associate various percentages with how much better one character is than another, rather than thinking "X character will win 60/100 matches while Y character will win 40/100." According to DanGR, they end up like this.

30/70-10/90(Red) -big disadvantage
40/60-35/65(Orange) -disadvantage
55/45-45/55(Yellow)-even
60/40-65/35(Green)-advantage
70/30-90/10(Blue)-big advantage

As can be seen from his matchup thread that you've no doubt seen.

 

[ixdnL]

Valid point. I've wondered this as well.

Since the two responders already have differing views, what can we say these estimates come from?

Edit: **** I type slowly.

 

[Crystanium]

Does it make sense?

That's not what I am asking, though. In fact, what you wrote up is the very thing I would like to be addressed.

Yes, it is completely arbitrary. It supposedly comes out to what Samuelson said. But, people typically associate various percentages with how much better one character is than another, rather than thinking "X character will win 60/100 matches while Y character will win 40/100." According to DanGR, they end up like this.

30/70-10/90(Red) -big disadvantage
40/60-35/65(Orange) -disadvantage
55/45-45/55(Yellow)-even
60/40-65/35(Green)-advantage
70/30-90/10(Blue)-big advantage

As can be seen from his matchup thread that you've no doubt seen.

If it is arbitrary, then why do we have it in the first place? If we were to figure out who was better than who by means of KO potential, recovery, &c., we could do just that. However, it wouldn't explain where these numbers for the match-ups come from.

 

[choknater]

Sheik is 100:0 to Ganondorf

that's a real matchup number

 

[MK26]

That's not what I am asking, though. In fact, what you wrote up is the very thing I would like to be addressed.

Genos wrote the numbers out and explained them. I wrote them out, explained them, and explained what the explanations mean...Take for example, Link vs. Kirby

That matchup is 70-30 in favour of Kirby. Why?

Kirby has several big attributes that give him the advantage, such as 0-40% throw combos, aerial control, gimping (main one)
When two players of equal skill face off, the Kirby will nearly always win, almost assuredly win a best-of-3 tourney set, and has a real chance at a three-stock due to these attributes
Even if Link counterpicks *insert Link's best stage here*, he will still not even the matchup, because Kirby has so much on him that the stage they play on will be moot

In short, the numbers are just an arbitrary (and opinionated) way of saying who has the advantage in a matchup, and by how much. Sometimes, people - even with their own personal biases - will agree on the difficulty of a matchup, and these numbers are what they agree on.

Now does it make sense?

EDIT: lol@posts@same time

10ats

 

[ph00tbag]

I've always considered them to be an estimation of the extent to which a player's ability to win will be assisted by their character's natural strengths and weaknesses in direct competition with their opponent's character's natural strengths and weaknesses. A ratio is used for the sake of universality, and having a base for measurement, and no other reason.

 

[choknater]

People need to make up statistics and be more accurate rather than general and make outrageous claims like Lucas is 83:17 on Zelda, provide a long and drawn out explanation on how Lucas' ftilt has the speed, range, and priority to stop all of Zelda's approaches and defenses.




..

Lol seriously though all this stuff is so arbitrary

there really should be only three matchup mindsets, all drawn from experience and game knowledge, not just what people tell you about matchups:

1. i have an advantage i think, so i'm good
2. this seems about even, so i just gotta play better
3. seems like i'm at a disadvantage, so i'll try to figure this out

notice the nuances about that mindset:

there's no ABSOLUTE about a matchup, it's never 100:0, 50:50, or 0:100 (unless it's sheik vs ganondorf which is definitely 100:0)

so who cares? figure stuff out...!

man theorybros is so dumb sometimes

 

[ph00tbag]

83:17

No one ever makes up numbers like this, because they're too precise. In fact, most never really go too far beyond the precision you're suggesting. My general understanding is:

70/30: I definitely have an advantage here, but this doesn't mean I can let up.
60/40: I think I have an advantage, so I should work to hold onto it.
50/50: This seems to be even, so I should pay special attention in this fight.
40/60: I'm pretty sure this is a disadvantage, so I should study this match-up in greater detail.
30/70: I definite have a disadvantage here, so I have to work my butt off, or maybe just find a good second.

 

[Crystanium]

Genos wrote the numbers out and explained them. I wrote them out, explained them, and explained what the explanations mean...Take for example, Link vs. Kirby

That matchup is 70-30 in favour of Kirby. Why?

Stop here. You can offer numbers. If the numbers are arbitrary, isn't it better to just say, "The match is in A's favor" without resorting to numbers? You offer examples of what Kirby can do against Link, but I don't see how you get to the numbering system. Aren't we better off figuring out who has better recovery, range, KO potential, &c?

I've always considered them to be an estimation of the extent to which a player's ability to win will be assisted by their character's natural strengths and weaknesses in direct competition with their opponent's character's natural strengths and weaknesses. A ratio is used for the sake of universality, and having a base for measurement, and no other reason.

And how do we get to this point?

 

[DanGR]

The numbers in my chart weren't totally arbitrarily picked. I actually thought about it for a while. In fact, I initially had the big disadvantages at 75-25 and higher. From all the threads I viewed in making the chart, though (and believe me, I've seen it all), I came across some general matchup numbers that expressed a certain difficulty level. Even matchups were commonly showed as 55-45 through 45-55 either way. Matchups that were "harder than hard" per se were given odds around 70-30 and above. Here's some examples... (remember, this was a while ago)
Yoshi v Wario- around 70-30
Luigi v Wario- around 70-30
Lucas v Marth- around 70-30
Olimar v Ganon- around 70-30

I mean can you really call Wario v Luigi just a disadvantage? With the support from some other people, I just decided to draw the line there.

I know this isn't really answering the questions in the OP, but I thought I'd share my experiences.

 

[ph00tbag]

Stop here. You can offer numbers. If the numbers are arbitrary, isn't it better to just say, "The match is in A's favor" without resorting to numbers? You offer examples of what Kirby can do against Link, but I don't see how you get to the numbering system. Aren't we better off figuring out who has better recovery, range, KO potential, &c?

Numbers are more concise than a statement, and if those numbers are taken to mean something universally, that makes them valid.

And how do we get to this point?

By considering things like approach methods, defensive abilities, OOS options, shield pressure options, effective tactics, etc. From this, you can determine the options one character has against the other. If both characters come out of this roughly equal in options, the match-up is roughly even. If one seems to have a few more effective options, they have a slight advantage. If one character easily shuts down another with more options, or some highly effective tactic, they have a major advantage.

Of course, you can sometimes in extreme cases extend it to match-ups where one character shuts down the other's approach and defense and has some highly effective tactic to boot, and the match-up becomes one of the rare unbeatables, eg D3 vs DK, which has been called anything from 80/20 to 90/10, from what I've seen.

 

[choknater]

Sheik:Ganon 100:0 come on get it through your thick heads

Sheik:Falcon 83:17

 

[Sonic The Hedgedawg]

Sheik is 100:0 to Ganondorf

that's a real matchup number

it's so true.... so true it's sad.

 

[Blackbelt]

Think of the numbers as a scale.


50-50



Obviously, the scales would be equal.


55-45

A slight tip in the scales, but nothing truly notable


60-40


The tip is much more notable



Keep following that to



100-0

The scale is tip completely to one side.




That's how I always saw the numbers.

 

[feardragon64]

It's not that arbitrary. Two characters of HIGH AND EQUAL SKILL(<=important) battle 100 times. What would be a LIKELY outcome? It's a theoretical battle of the characters and their options. The point is that players don't actually get involved.

You might think it's arbitrary, but it's a hell of a lot more informative than saying, "SNAKE>C4PT4IN F4LC0N!!1!" That doesn't do anything. This gives a general consensus from the community. Through experience in the community the numbers actually come to mean something.

 

[HeroMystic]

Personally, I've always saw it as this.

Netural = Player skill alone.
60:40 = A decent player with "X" character will beat decent player with "Y" character.
70:30 = A decent player with "X" character will beat good player with "Y" character.
80:20 = A decent player with "X" character will beat excellent player with "Y" character.
90:10 = ****.
100:0 = Does not exist.

 

[JackieRabbit5]

we just feel the need to put numbers to a term like "slight advantage" to compare it to others though it probably is arbitrary and inaccurate in reality

i'd complain about the system myself but theres not really a better way of doing it unless u wanna try and get REALLY technical with statistics and experimentation (which doesn't seem like it would work anyways with so many variables)

btw i think matchups > tier list

 

[~ Gheb ~]

100:0 = Does not exist.

lol

DK vs DDD

 

[choknater]

of course 100:0 exists! Sheik vs Ganon, obviously.

(i seirously wanna get a ganon in here to refute this and then ask him to MM)

 

[Blad01]

of course 100:0 exists! Sheik vs Ganon, obviously.

(i seirously wanna get a ganon in here to refute this and then ask him to MM)

Ahah, i'm also a Ganon main, chok è_é

Sheik - Ganon is 80-20, no more :D

But that's not the point >_>

 

[Steeler]

i like what overswarm said in his rob matchup thread before he bailed on rob. (jk)

it was something along the lines of...

as the metagame progresses, these numerical analyses will likely shift slightly, but a general "large/small (dis)advantage" won't, barring some huge at advancement in one character's game.

so os just didn't use matchup numbers in that thread.

besides, do we really need to know whether a matchup is 70:30 or 75:25 or 80:20? the general point is that this is a bad matchup and you should use another character or learn a secondary for this matchup. this is the whole point of matchup threads. these numbers get the point across but aren't necessary. there will almost always be debate on whether something is 65:35 or 70:30, which is the common line between large and slight advantage...

i think it's best to keep it in the single digits, ie 5:5, 6:4, etc. i think 55:45 and 65:35 are too precise.

 

[3GOD]

One way I've handled this is to first group the match-ups into good, neutral, and bad. Then compare the match-ups within each category and list them from easiest to hardest. Use this as a sort of guide for feeling out the numbers.

Of course this is based somewhat on personal opinion, but it's no more subjective than other arguments I've seen.

 

[HeroMystic]

lol

DK vs DDD

Unless DDD's matchup guide is outdated, I'm reading 80:20 here.

 

[~ Gheb ~]

That's just a number. DDD will always win vs DK, so it'S 100:0 for DDD...

 

[Emblem Lord]

Numbers represent probability, but it also represents how much the scale tilts in the favor of one combatant over the other.

Once you hit 65/35, the character that's at disadvantage will rarely win.

 

[choknater]

Pretty much Sheik vs Ganon and DDD vs DK are 110:-10

 

[ShadowLink84]

Not necessarily true, DDD doesn't land the grab that easily. I am quite sure it wouldn't be a bad disadvantage if the infinite wasn't there.

 

[DanGR]

I've beaten a Sheik in friendlies with Ganon. ^_^b

 

[~ Gheb ~]

Not necessarily true, DDD doesn't land the grab that easily. I am quite sure it wouldn't be a bad disadvantage if the infinite wasn't there.

But it is there

 

[Wayland]

Ok, people are hooked on this idea of "Probability of Winning between 2 equally skilled competitors." That's pretty much impossible to determine. I dare you to find 2 perfectly evenly skilled competitiors. But that's not enough. You'll have to find a boatload of them, get them to play a bunch of matches with every character, then run the numbers to find probabilities. Even then, tell me how those numbers are representative of the general smash community. It isn't. Not even a little.

The numbers are a rough numerical estimate of the comparative weight of the potential advantages of a character in a given match. That's all. It's not about winning and losing matches, it's about how a character's moveset and properties give them an advantage/disadvantage in the series of situations that constitute a match.

Edit: ****, that's confusing to look at. Pretty much, if you liked the scales meataphor, you win, if you went with probability, better luck next time.

 

[Sonic The Hedgedawg]

I've beaten a Sheik in friendlies with Ganon. ^_^b

it was an awful sheik then.

Sheik can punish everything ganondorf does and edgeguard him like a monster. I suppose he might be able to flame choke SD if he's lucky to even out the stock occasionally, but if he misses once, sheik has a stock advantage and it's sheik's game to lose.

I've played good ganon's with my decent sheik and have completely torn up. The matchup is totally onesided.

 

[~ Gheb ~]

Many match-up's vs Ganon are one sided. All of the spacies have 80:20 match-up's or higher

 

[Dark Sonic]

Okay, here's a simple way to explain why we use numbers.

Yes, the numbers are a bit arbitrary, but their only purpose is to show how severe one character's advantage over another is.

For example, Marth vs Ness can be viewed as anywhere for 80:20-90:10 in favor of Marth. What does this mean? It means that as a character Marth ***** Ness. It's not something the Ness player should ever be expected to win if the players are even remotely close in skill, it's just that bad.

Marth has more range, racks damage faster, kills earlier, and in general outclasses Ness in almost every category. And on top of all that, he has a release grab infinite, that chains into down smash or f-smash for the kill.

If I had to express this severity of advantage on a 0-10 scale, I'd give Marth 9/10. Now to put this as a numerical ratio comparing Marth's advantage to Ness's disadvantage it would look like this

9:1

And that's it. Now when I want to describe how bad Ness is against Marth, I can just say it's 9:1 Marth's favor. I don't have to list all those reasons why, I can just list that ratio, and everyone will understand that Marth ***** Ness. If you disagree with the number, then you look up the reasons yourself.

The numbers are only there to serve as a quick notation. When you compare numbers, you're actually comparing the severity of advantages in each matchup. A 1:9 matchup is much harder than a 4:6 matchup, and the numbers just represent that difference in difficulty

 

[FBM]

That's just a number. DDD will always win vs DK, so it'S 100:0 for DDD...

No.

IMO, a character matchup is inversely proportional to the skill difference between the two players needed to make the matchup even. If D3 vs DK is 80:20, then the DK player has to be like 80:20 in terms of skill to the D3 player to make up for it. 100:0 would preclude winning, which is stupid, because it happens.

In that thread about infinites, some guy was saying he was a D3 player who went to a tourney and faced off against Bum's DK. He thought it would be easy with the infinite, but Bum played so carefully that by the time the guy got the infinite off, he'd already lost a stock himself and went on to lose the tourney set. The difference between the player's skill had to be huge to compensate for the character matchup, but it happens.

The numbers are only there to serve as a quick notation. When you compare numbers, you're actually comparing the severity of advantages in each matchup. A 1:9 matchup is much harder than a 4:6 matchup, and the numbers just represent that difference in difficulty

Agreed.