-
Welcome to Smashboards, the world's largest Super Smash Brothers community! Over 250,000 Smash Bros. fans from around the world have come to discuss these great games in over 19 million posts!
You are currently viewing our boards as a visitor. Click here to sign up right now and start on your path in the Smash community!
The Match-Up Chart Project
- Thread starter Tero.
- Start date
Skler
Smash Master
Non-MBR people should be included if they're truly good at discussing and not just being biased. Unfortunately people outside the MBR usually aren't like that.
We should discuss the controversial ones. You know, like Mewtwo vs Zelda.
We should discuss the controversial ones. You know, like Mewtwo vs Zelda.
Luma
Smash Lord
its not like we dont have (or used to have) alot of biased bull**** in here aswell =P
i think we first should do a vote and if some matchups just get complete different votes we should discuss it
i think we first should do a vote and if some matchups just get complete different votes we should discuss it
D
Deleted member
Guest
this is a really horrible idea.
tero what are you thinking
tero what are you thinking
Tero.
Smash Champion
- Joined
- Jul 18, 2008
- Messages
- 2,688
First: that was not my idea, I was just the one willing to actually do something.
Seccond: How can the idea of making a match-up chart for the community be bad? Sorry man, I don't get it.
But yeah obviously you're right... what was I thinking trying to actually do something with this lazy ****ed up BR.
Seccond: How can the idea of making a match-up chart for the community be bad? Sorry man, I don't get it.
But yeah obviously you're right... what was I thinking trying to actually do something with this lazy ****ed up BR.
N64
Smash Champion
Axe, zig, chad, and I are in the process of fleshing out pikachu's matchups for teh_icy's matchup chart in the general melee discussion room, and I'll be happy to share the results here if needed. I feel they represent pikachu best in the current melee community and have no issue putting my confidence in the result.
As far as creating an MBR matchup chart, I don't realistically see how we'd likely be able to come up with one that would be agreeable as we can't seem to agree on a tier list (which imo takes less work/discussion). I think it's an ambitious but ultimately fruitful venture.
I'm also not sure where it was explained why removing the lower portion (nonviable) from the matchup chart is needed. Is it because we are assuming their matchups have no effect on competative play? or that we are much less likely to have as much info on their matchups because they are less commonly played (and would thus be more likely to misrepresent them)?
As far as creating an MBR matchup chart, I don't realistically see how we'd likely be able to come up with one that would be agreeable as we can't seem to agree on a tier list (which imo takes less work/discussion). I think it's an ambitious but ultimately fruitful venture.
I'm also not sure where it was explained why removing the lower portion (nonviable) from the matchup chart is needed. Is it because we are assuming their matchups have no effect on competative play? or that we are much less likely to have as much info on their matchups because they are less commonly played (and would thus be more likely to misrepresent them)?
Fly_Amanita
Master of Caribou
The initial idea was this:
It would be nice to create a weighted match-up chart to see how the characters ultimately stack up in regards to how they perform against each other. However, since there are many characters and since many match-ups between the lousy ones aren't well-understood, undertaking a complete match-up chart isn't very practical. However, we could approximate the upper results of such a chart with a smaller chart involving only some of the better characters. That this chart would yield a decent approximation of the results from a normal chart relies on the assumption that the represented characters' match-ups against the unrepresented ones ultimately would have a negligible effect on their rankings in a complete weighted match-up chart. For example, that Sheik ***** Pichu worse than Fox does should have very little influence on their relative placings.
I think there are at least 11 characters who are definitely good enough to warrant representation on such a chart. Characters that randomly have decent match-ups against some of the better characters should also be taken into account; e.g. ICs arguably lose to a couple characters that are generally perceived as notably worse than them, which is a factor that should be represented, so those characters may also appear on the chart.
It would be nice to create a weighted match-up chart to see how the characters ultimately stack up in regards to how they perform against each other. However, since there are many characters and since many match-ups between the lousy ones aren't well-understood, undertaking a complete match-up chart isn't very practical. However, we could approximate the upper results of such a chart with a smaller chart involving only some of the better characters. That this chart would yield a decent approximation of the results from a normal chart relies on the assumption that the represented characters' match-ups against the unrepresented ones ultimately would have a negligible effect on their rankings in a complete weighted match-up chart. For example, that Sheik ***** Pichu worse than Fox does should have very little influence on their relative placings.
I think there are at least 11 characters who are definitely good enough to warrant representation on such a chart. Characters that randomly have decent match-ups against some of the better characters should also be taken into account; e.g. ICs arguably lose to a couple characters that are generally perceived as notably worse than them, which is a factor that should be represented, so those characters may also appear on the chart.
Skler
Smash Master
SleepyK
Banned via Administration
- Joined
- Mar 26, 2006
- Messages
- 5,872
hey, let's do this, this is a good idea.
do we need to gather empirical data for this?
fly wat u think about maskedsheik's numbers
Fly_Amanita
Master of Caribou
I'm not sure which numbers you're referring to. I do recall seeing him make a stat-heavy topic quite a while ago, but I don't remember exactly what it was about.
I don't think any empirical data is essential for this chart, although it couldn't hurt to see how players of X character tend to perform against character Y in tournament at a high level in the past several years. It's not something I'd prioritize since honest input from high-caliber players should tell us more or less the same thing.
If we actually go about doing this, here are a few things we'll need to clarify/figure out:
-What the match-up ratios actually mean. I've seen a lot of different takes on what they should represent; the most natural thing to me for a ratio to represent is the frequency with which character X will win against character Y in a tournament set (whether it's bo3 or bo5 hopefully doesn't matter too much, although it is something worth considering) at a high level of play in accordance with the current recommended ruleset.
-What scale we'll use for match-up ratios. The scheme we see all the time implicitly allows 21 possible ratios per match-up, which leaves room for way too much nitpicking. I'm partial to using the ratios n : 10-n for 0<=n<=10, although going slightly lower probably won't hurt much.
-Exactly which characters to represent. There are some characters that are uniformly considered good and deserve representation, but then there are also the characters that we shouldn't neglect that do okay against some of the good ones. As an example, DK might deserve representation since he apparently does very well against Samus. However, we wouldn't then put characters that do well against DK on, since that sort of thing could lead to including way too many characters; we'd only add characters that have pretty good match-ups against some of the "viable" ones. On another note, I think the chart should only aim to rank the "viable" ones and not the other characters thrown on for accuracy purposes, since it would be odd to include some mediocre characters in the rankings dictated by match-ups when other similarly mediocre (and maybe even better) characters aren't on the list.
-What weighting algorithm we'll use. If there's anything that's standard out there for weighting charts once all of the numbers are in, I'd like to know about it. I couldn't find anything with my weak Google skills. It doesn't seem like something that would be hard to come up with and I'm sure I or someone else here could think of something decent should we need to use our own algorithm.
I don't think any empirical data is essential for this chart, although it couldn't hurt to see how players of X character tend to perform against character Y in tournament at a high level in the past several years. It's not something I'd prioritize since honest input from high-caliber players should tell us more or less the same thing.
If we actually go about doing this, here are a few things we'll need to clarify/figure out:
-What the match-up ratios actually mean. I've seen a lot of different takes on what they should represent; the most natural thing to me for a ratio to represent is the frequency with which character X will win against character Y in a tournament set (whether it's bo3 or bo5 hopefully doesn't matter too much, although it is something worth considering) at a high level of play in accordance with the current recommended ruleset.
-What scale we'll use for match-up ratios. The scheme we see all the time implicitly allows 21 possible ratios per match-up, which leaves room for way too much nitpicking. I'm partial to using the ratios n : 10-n for 0<=n<=10, although going slightly lower probably won't hurt much.
-Exactly which characters to represent. There are some characters that are uniformly considered good and deserve representation, but then there are also the characters that we shouldn't neglect that do okay against some of the good ones. As an example, DK might deserve representation since he apparently does very well against Samus. However, we wouldn't then put characters that do well against DK on, since that sort of thing could lead to including way too many characters; we'd only add characters that have pretty good match-ups against some of the "viable" ones. On another note, I think the chart should only aim to rank the "viable" ones and not the other characters thrown on for accuracy purposes, since it would be odd to include some mediocre characters in the rankings dictated by match-ups when other similarly mediocre (and maybe even better) characters aren't on the list.
-What weighting algorithm we'll use. If there's anything that's standard out there for weighting charts once all of the numbers are in, I'd like to know about it. I couldn't find anything with my weak Google skills. It doesn't seem like something that would be hard to come up with and I'm sure I or someone else here could think of something decent should we need to use our own algorithm.
ArcNatural
Banned ( ∫x, δx Points)
I like the idea of doing a Top-Mid Tier viable character Match-up Chart project. I will say though, that the one in Melee Discussion is one of the really good projects they have been working on. I almost don't like the idea of coming up with a chart since it would detract away from one of the truly good threads in the Melee Discussion.
The tier list has always been something the MBR has done as well as having the public Tier List thread. Again I'm not sure I want to detract from the Melee Disc. Match-up Chart.
The tier list has always been something the MBR has done as well as having the public Tier List thread. Again I'm not sure I want to detract from the Melee Disc. Match-up Chart.
Druggedfox
Smash Champion
@Flyamanita
I've always interpreted the ratio to mean how many times you'll win out of the total.
If a matchup is 50:50, two players who know the matchup equally well and have equally good characters will each win 50/100 matches.
If a matchup is 65:35, under the same conditions as above, one character will win 65/100 matches.
It's obviously approximate, but that way the numbers aren't especially relative to each other. This is also a rather simple way of looking at it, so we would not get into any particularly convoluted specifics about what the ratios mean.
I've always interpreted the ratio to mean how many times you'll win out of the total.
If a matchup is 50:50, two players who know the matchup equally well and have equally good characters will each win 50/100 matches.
If a matchup is 65:35, under the same conditions as above, one character will win 65/100 matches.
It's obviously approximate, but that way the numbers aren't especially relative to each other. This is also a rather simple way of looking at it, so we would not get into any particularly convoluted specifics about what the ratios mean.