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!
Simplified for you.
The point is that the "average ranking points" is not useful information on its own, because it is entirely dependent on the cutoff - but I feel like you're not understanding why I'm complaining about that, so let me demonstrate with examples.
When there's people around who don't get math, and math provides answers relevant to what those people are talking about, yes, it is always math class.Master Raven said:
I think a few smashers (especially the lazy ones) could use another one of those.
It's still a bothersome matchup for the ICs nonetheless, as is Peach because of their sheer ability to constantly separate the climbers and avoid grabs. Don't let matchup ratios fool you.
Sometimes figuring out what's best for the community requires us to treat things like it is a class.
Agreed.
I'm just going to hop in and say that I disagree with this.
Perhaps, but what I said still applies. You're also ignoring stage influence on the matchup.
You're missing the point.
A completely balanced game breaks the CP system too.
That's my bad, I read the post that you had quoted in your post and accidentally hit the reply button to yours.
That's very similar to what I was saying, though it's more degree and number.
With GREAT difficulty, at some point I hope to kick off a project to reduce it to pure math using character attributes, but there really is no short answer.
Then the problem becomes "what players do we understand as reflecting the MU at a relevant level of play".Oh, just saw some people are bringing out a bell curve. I like drawing curves. Doing that can explain why this skewness thing is a problem.
Here's what people are used to thinking about:
Nice and symmetric; taking averages and stuff works nicely. But that's not what happens when you pick the N best players of each character and plot their scores. It looks something more like this:
Again, I'm eager to get the real data, but this is as CLEAN as the data could possibly be (i.e. best case scenario for Omni's types of methods).
Anyway notice the horizontal lines I drew. Depending on the number of players you choose to average the scores over, you get very different averages FOR EXACTLY THE SAME SHAPE, i.e. exactly the same level of "popularity" which is what Omni claims he's removing from the data.
The smallest N I drew was N=2; that's how many Omni gave Snake. Notice how high the "average score" winds up being in this case.
The highest N I drew was N=10. Omni gave Meta Knight N=13.
That is false, throw in 20 skilled link mains and the data will be skewed unless the data is averaged period. Of course how large of a skew depends on how inherently good the character is, but the only perfect way to judge is equal skill.
So what you are trying to say is that we should take the average performance of all MK mains and compare them to the average score of every other character's mains?You're missing the point.
The argument was that Metaknight the character has such a huge representation compared to all of the other characters on the cast, that sheer numbers alone guarantee him to be placing higher than he normally would if he had bad or even average reputation - regardless of how good he is.
This is why tournament rankings and such are of little to no use when determining the actual tier spot of the character. There are way too many variables involved in tournaments to reliably report the usefulness of all of the characters.
Yes, the data is skewed because you have on one end players that do very very well with MK, and on the other end players who literally bring down MK's tournament average. The two extremes are too far apart to allow the average as a whole to be reliable.
A good example is this. Take the income per family graph I made below. Note how just a few high incomes make the mean - the fulcrum point that balances the incomes above and below - deceptively high. Why is this?
Because of one person essentially ****ing up the entire mean average. The dude at the end (let's call him Bill Gates) makes the graph deceiving and basically inaccurate for real world applications.
I wonder who we know that's like that in the Smash community? Replace income in the graph with top spots occupied by MK mains.
I already accounted for outliers; I used the top 3 scoring for MK, Diddy, and Snake from Ankoku's list then I did it again but took out M2K, Ally, and ADHD (Metaknight lost the most points from this). MK still loses.You're missing the point.
The argument was that Metaknight the character has such a huge representation compared to all of the other characters on the cast, that sheer numbers alone guarantee him to be placing higher than he normally would if he had bad or even average reputation - regardless of how good he is.
This is why tournament rankings and such are of little to no use when determining the actual tier spot of the character. There are way too many variables involved in tournaments to reliably report the usefulness of all of the characters.
Yes, the data is skewed because you have on one end players that do very very well with MK, and on the other end players who literally bring down MK's tournament average. The two extremes are too far apart to allow the average as a whole to be reliable.
A good example is this. Take the income per family graph I made below. Note how just a few high incomes make the mean - the fulcrum point that balances the incomes above and below - deceptively high. Why is this?
Because of one person essentially ****ing up the entire mean average. The dude at the end (let's call him Bill Gates) makes the graph deceiving and basically inaccurate for real world applications.
I wonder who we know that's like that in the Smash community? Replace income in the graph with top spots occupied by MK mains.
Hey you! Get on skype; it's podcast time! Lol.
Now that you clarified your position I actually agree with this, but this is only a factor, not a cutthroat reason.
Third worst in the game?
Who does he use?
Whoa, this is neat. While I was making my other post you also made a version of it.You're missing the point.
The argument was that Metaknight the character has such a huge representation compared to all of the other characters on the cast, that sheer numbers alone guarantee him to be placing higher than he normally would if he had bad or even average reputation - regardless of how good he is.
This is why tournament rankings and such are of little to no use when determining the actual tier spot of the character. There are way too many variables involved in tournaments to reliably report the usefulness of all of the characters.
Yes, the data is skewed because you have on one end players that do very very well with MK, and on the other end players who literally bring down MK's tournament average. The two extremes are too far apart to allow the average as a whole to be reliable.
A good example is this. Take the income per family graph I made below. Note how just a few high incomes make the mean - the fulcrum point that balances the incomes above and below - deceptively high. Why is this?
Because of one person essentially ****ing up the entire mean average. The dude at the end (let's call him Bill Gates) makes the graph deceiving and basically inaccurate for real world applications.
I wonder who we know that's like that in the Smash community? Replace income in the graph with top spots occupied by MK mains.
People say this all of the time but lack legitimate proof that people are horribly downplaying MU's. People say this game has MU's much worse than people say yet no one explains why they are even as bad as people say.
