Official MBR 2010 NTSC Tier List

[Deleted member]

He only asked like 12 people or something, and they were all MBR members + Armada.

I'm sure he'd greatly appreciate your input. Shoot him a PM or ten.

armada is in the MBR. and he's a greatly welcome member.

 

[Fortress | Sveet]

armada was only brought into the mbr for the tier list

i think it'd be a shame if he wasn't allowed in full time though

 

[Beat!]

armada is in the MBR. and he's a greatly welcome member.

Oh yeah, you're right. Mah' bad.

 

[JPOBS]

Who are the 12 or so people making up the voting panel?

 

[Froggy]

IMO 12 is far too low a number to make a valid tier list. In statistics, a poll needs at least 30 samples to be considered statistically significant. Having less than half of that is near useless. Do he tier list right or don't bother doing it at all.

 

[Construct]

The actual tierlist that is going to be drawn up and published is made by established posters and players in the community so that randoms scrubs like you and I don't throw the vote.

 

[RaphaelRobo]

That's why I'm going to be voting. I'm an established poster and trash-talker in the community. Plus, I represent the new generation of smashers. You know, the ones that would be referred to as random scrubs if they weren't me.

 

[choknater]

i agree that 12 is too few

 

[Purpletuce]

IMO 12 is far too low a number to make a valid tier list. In statistics, a poll needs at least 30 samples to be considered statistically significant. Having less than half of that is near useless. Do he tier list right or don't bother doing it at all.

You're saying "in statistics" as if you took a class on it. . . BUT, there is no magic "You need 30 samples". Also, who are you to say what makes the list near useless? The more samples the better, as well as the biggest variety among knowledgeable smashers, to avoid sampling bias. It is also to be noted that for every player asked, they are likely pooling knowledge from many more players.

 

[Bones0]

In statistics, a poll needs at least 30 samples to be considered statistically significant.

According to a study by 27 statisticians.

 

[Seartu]

According to a study by 27 statisticians.

At least 27>12

 

[Fortress | Sveet]

Arbitrary numbers are arbitrary. There are only 9 supreme court judges. The chicago PR has been done with 3-5 panelists. This isn't an opinion survey.

 

[Froggy]

The actual tierlist that is going to be drawn up and published is made by established posters and players in the community so that randoms scrubs like you and I don't throw the vote.

You've missed the point entirely. Having only 12 pros vote makes the tier list dumb. The tier list is far too easily skewed because of so few votes.

You're saying "in statistics" as if you took a class on it. . . BUT, there is no magic "You need 30 samples". Also, who are you to say what makes the list near useless? The more samples the better, as well as the biggest variety among knowledgeable smashers, to avoid sampling bias. It is also to be noted that for every player asked, they are likely pooling knowledge from many more players.

Um...yes. I did take a class on it. Three in fact(including one that would make your head spin). 30 samples is the widely established number of samples for the correlation coefficient and the t-statistic to be representative of the real population and approach a normal distribution. (Any less and the data is far too easily skewed)

I am someone who apparently knows a whole lot more about statistics than you, and it's not a rule I came up with. In it's every statistics text book. I'm not saying that you need 30 samples for a poll to be useful. But 12? Less than half of the statistical significance threshold? That would be useless, and a waste of time really.

 

[ShrieK1295]

The number 30 is completely irrelevant in this case. This isn't a sample that is supposed to be indicative of some fact about a population. This is asking people their opinions about balance in a game. If you ask one smart person, this is potentially as revealing as asking 1,000. Given, having more people is less prone to bias, but acting like this is related to what you learned in stats is just wrong.

 

[tarheeljks]

Um...yes. I did take a class on it. Three in fact(including one that would make your head spin). 30 samples is the widely established number of samples for the correlation coefficient and the t-statistic to be representative of the real population and approach a normal distribution. (Any less and the data is far too easily skewed)

I am someone who apparently knows a whole lot more about statistics than you, and it's not a rule I came up with. In it's every statistics text book. I'm not saying that you need 30 samples for a poll to be useful. But 12? Less than half of the statistical significance threshold? That would be useless, and a waste of time really.

not dealing with random sampling here, so what shriek said

edit: whether you agree w/polling "experts" is largely a different issue i think. no one has claimed that the distribution of mbr opinions on the tier list is "normal". kind of a weird thing to say anyway unless the question is something like "what % of players think x is the best character" or you are looking at individual permutations of tier lists as outcomes and saying the distribution of these lists is normal. and in the latter case we actually have many reasons to believe that a large scale poll of tier lists would not have normal results b/c we would expect more than one mode (look at how long this thread is), and heavy skew (it is widely recognized that a handful of characters are much better than others)

 

[KrIsP!]

At least 27>12

Your right man, 27 is larger than 12.

On that note, I think more than 12 people will be involved.

 

[RaphaelRobo]

27 is larger than 12.

I disagree with this statement.

 

[Sangoku]

I love all the butthurt kids who can't admit they know nothing about the game and should let the ones who supposedly know a lot more make a HIGH LEVEL OF PLAY tier list.

 

[N.A.G.A.C.E]

I dont really care how they choose who gets to vote on the next teir list b/c unless its random i dont believe i would get a vote.

But i dont understand the people who think that just b/c someone is good at the game that that person has a deeper understanding of the game/characters on the whole. Its like in sports many top analyst dont play and where not former players they are just people who study and have a deep understanding of the game.

 

[Seartu]

I love all the butthurt kids who can't admit they know nothing about the game and should let the ones who supposedly know a lot more make a HIGH LEVEL OF PLAY tier list.

it doesn't appear to me that people are really trying to involve themselves in the actual vote.

I know there are others like myself who understand the game mechanics to a decent extent and enjoy playing, watching, and discussing the game. While I certainly am not qualified to vote in anything this official I think there are probably more than 12 individuals who ARE qualified and would be willing to participate in such a vote.

/endsomewhatformalwriting

 

[Froggy]

The number 30 is completely irrelevant in this case. This isn't a sample that is supposed to be indicative of some fact about a population. This is asking people their opinions about balance in a game. If you ask one smart person, this is potentially as revealing as asking 1,000. Given, having more people is less prone to bias, but acting like this is related to what you learned in stats is just wrong.

Barring extreme circumstances any poll will be related to what I learned in stats class.

I doubt you understand the concept. Unless the goal of the tier list is to learn the opinion of the tier list from a few random members of the mbr as opposed to the mbr as a whole, then asking 12 members out of that population is useless.

And lol about asking one smart person. Looks like I am going to have to explain skew to you.

Let's say 1 person with a few peculiar views(for example Puff as 1 and Marth as a mid tier) votes in this. Even with the other 11 votes with more conventional views such as Fox at #1 and Marth as high tier, the mean(average) calculation will be moved away from the convention because of this 1 data point. Even in a group as small as 12 this isn't such a big deal, but what happens if more than one person with unconventional views vote? What if it's 4 or 5 or 6 with peculiar votes? Then what happens is that the curve of the data points gets skewed so badly from what it should be because nearly half of the votes are strange and not representative of the top players because some idiots decide that having 12 data points would be enough.

I know I'm not the best teacher but I hope you get the point now.

If however

not dealing with random sampling here, so what shriek said

edit: whether you agree w/polling "experts" is largely a different issue i think. no one has claimed that the distribution of mbr opinions on the tier list is "normal". kind of a weird thing to say anyway unless the question is something like "what % of players think x is the best character" or you are looking at individual permutations of tier lists as outcomes and saying the distribution of these lists is normal. and in the latter case we actually have many reasons to believe that a large scale poll of tier lists would not have normal results b/c we would expect more than one mode (look at how long this thread is), and heavy skew (it is widely recognized that a handful of characters are much better than others)

The point of a distribution being normal isn't to say it's normal. It's so that all related coefficients, measures of variance, confidence thresholds, etc provide a useful calculation. That is all lost if we do something as stupid as making the tier list composed of 12 sample points. And as mentioned in my response to Shriek above it leaves the tier list open to rather frightening data skew possibilities.

 

[Bones0]

You're assuming the opinion of people with outliers in their data are somehow less correct than the majority. If 9 people think Puff is 5th and 1 thinks Puff is 1st, you can't just ignore the 1 simply because they are the minority. Maybe they are the most knowledgeable or least biased about that character and everyone else is just wrong.

No matter what, I'm not particularly interested in what the tier list will look like. Players have too much bias (both from experience and precedence of previous tier lists), and the character representation is skewed horribly towards certain characters making results almost useless.

 

[Varist]

I dont really care how they choose who gets to vote on the next teir list b/c unless its random i dont believe i would get a vote.

But i dont understand the people who think that just b/c someone is good at the game that that person has a deeper understanding of the game/characters on the whole. Its like in sports many top analyst dont play and where not former players they are just people who study and have a deep understanding of the game.

no bc in sports all the athletes suffer brain damage so they can't analyze **** or else they'd be the best

--

i'm trying real real hard to side with the 12 knowledgeable people thing because I hate scrub bias but I can't ****ing do it. 12 is too ****ing small. I mean god damn look at the ssbpd top FIFTY, almost all of those ***s should be asked. There's no reason not to. 12 can't be the lowest you can muster up, jesus. Who cares if it throws **** off, there should be no one to decide where it's being thrown off from. Understand that sentence. That's the REAL bias. MBR is a bunch of ******s. **** the MBR, next tier list is going to be ****. and puff better go down.

 

[Bones0]

We are the 99%!

 

[Elyssa Xey Hexen]

G&W > Roy since Roy gets punished for actually hitting his opponent.

 

[RedGamer]

why doesn't somebody make one?
I promise peach has moved up because of armada.
Falcon has gone down for lack of Scar/Silentspectre/Darkrain/Hax not placing top/Me not being able to get to tourneys/other crap.
I'm pretty sure falco is above fox again (only to be put back below him).
where the heck are the IC's now?
I think we need a 2012 tier list.

 

[Fortress | Sveet]

Lol redgamer is out of the loop

no bc in sports all the athletes suffer brain damage so they can't analyze **** or else they'd be the best

--

i'm trying real real hard to side with the 12 knowledgeable people thing because I hate scrub bias but I can't ****ing do it. 12 is too ****ing small. I mean god damn look at the ssbpd top FIFTY, almost all of those ***s should be asked. There's no reason not to. 12 can't be the lowest you can muster up, jesus. Who cares if it throws **** off, there should be no one to decide where it's being thrown off from. Understand that sentence. That's the REAL bias. MBR is a bunch of ******s. **** the MBR, next tier list is going to be ****. and puff better go down.

Not all the people in the top 50 are able to participate. I had many people decline participation (including PP). Others don't have an account (Mango). Some are known for subpar post quality (no names). Some people have said they would participate but havent posted once (hbox).

To be honest, I don't really care. You guys can fight among yourselves and argue over which way is better, but you know whats worse than both of them? Not doing the project at all. If you're not satisfied with the results, feel free to organize your own tier list project. I will not be offended, I promise.

#idgaf

 

[N.A.G.A.C.E]

Idk why sub par post quality would make someones opinion on a tier list invalid. Just b/c someone may not be the best writing does not make them dumb nor does it make their opinions wrong. Also for something like a tier list which seems to just be about voting i dont see an issue with a smart person who is bad at writing being allowed to have a say.

 

[Fortress | Sveet]

Because the project was intended to be discussion focused. If one cannot meet a certain standard of posting, they cannot discuss properly and thus aren't valuable to the project. Especially people who choose to insult rather than discuss ideas. Also, discussion is open to everyone in the MBR and I have already told them if they participate I will add them to the final panel. Also, most of the discussion has been happening in this thread to improve transparency and allow anyone to post their ideas. In a sense, everyone in the community who is interested in presenting their opinions or views has means to sway the list.

So no, this isn't an opinion survey. This is an active discussion with a select number of capable judges.

 

[Deleted member]

G&W > Roy since Roy gets punished for actually hitting his opponent.

so does marth

/strawman

you should explain what you mean better.

joe i told you that you'd have problems with your approach. if you feel that you have to strong-arm through some X number of people, you should take the heavy opposition as a sign that you're doing something wrong.

w/e cactuar will name search and find this thread and say something ending with a

cactuar's posts are worth 4.2 cactuars.

 

[N.A.G.A.C.E]

I thought the tier list was voted on.

Anyway i would agree that people who insult should not get a say, but i am kinda curious what you consider "cannot discuss properly"? Like what level is the cutoff?

Also are you saying i could write my opinions for what i believe the next tier list should be, and as long as my reasoning makes sense it will be taken into consideration?

 

[Bones0]

Also are you saying i could write my opinions for what i believe the next tier list should be, and as long as my reasoning makes sense it will be taken into consideration?

lol NO. This is SmashBoards.

 

[Varist]

. . . Also, most of the discussion has been happening in this thread to improve transparency and allow anyone to post their ideas. In a sense, everyone in the community who is interested in presenting their opinions or views has means to sway the list.

So no, this isn't an opinion survey. This is an active discussion with a select number of capable judges.

who are the 12 judges

 

[Fortress | Sveet]

joe i told you that you'd have problems with your approach. if you feel that you have to strong-arm through some X number of people, you should take the heavy opposition as a sign that you're doing something wrong.

Yeah, I wish I would have talked to you before I got started. In hind sight, I think your way is better, but to change it now would be to invalidate all the work that has been put into it thus far. I think the method of selecting panelists is only marginally important, though, so it shouldn't make an impact on the final list.


@NAGACE
These are questions that have been posed before. If you want to talk about it more feel free to PM me or IM me and we can chat

 

[RedGamer]

Lol redgamer is out of the loop

The project was intended to be discussion focused. If one cannot meet a certain standard of posting, they cannot discuss properly and thus aren't valuable to the project. Especially people who choose to insult rather than discuss ideas. Also, discussion is open to everyone in the MBR and I have already told them if they participate I will add them to the final panel. Also, most of the discussion has been happening in this thread to improve transparency and allow anyone to post their ideas. In a sense, everyone in the community who is interested in presenting their opinions or views has means to sway the list.

So no, this isn't an opinion survey. This is an active discussion with a select number of capable judges.

I'm not trying to give an opinion, I'm trying to be informed about the current position of the tier characters.

 

[Deleted member]

Yeah, I wish I would have talked to you before I got started. In hind sight, I think your way is better, but to change it now would be to invalidate all the work that has been put into it thus far. I think the method of selecting panelists is only marginally important, though, so it shouldn't make an impact on the final list.

There's nothing wrong with starting over if you don't like what you have so far.

 

[Fortress | Sveet]

I'm not trying to give an opinion, I'm trying to be informed about the current position of the tier characters.

nah its cool, but theres been a tier list project in the works for a while now. Next week was supposed to be the "final voting" time. Thats all i meant by you being out of the loop.

For the sake of transparency...

[collapse=Compiled Preliminary Tier List]S:
Fox 1.57
Falco 1.86

A:
Sheik 3.14
Marth 4.43
Puff 4.64
Peach 5.79

B:
Falcon 7.00
ICs 7.93
Doc 8.86

C:
Samus 10.64
:ganonmelee:Ganon 11.36
Pikachu 11.86
Luigi 12.85 (1 abstain)
Mario 13.71

D:
DK 15.46 (1 abstain)
Y Link 16.29
Link 16.38 (1 abstain)
Zelda 18.08 (1 abstain)
Yoshi 18.85 (1 abstain)

E:
Roy 20.67 (2 abstain) / Mewtwo 20.67 (2 abstain)
:gwmelee:G&W 21.33 (2 abstain)
Ness 22.54 (1 abstain)

F:
Kirby 24.15 (1 abstain)
Bowser 24.25 (1 abstain)
Pichu 24.62 (1 abstain)[/collapse]

This is before the discussion has begun. Feel free to hold public discussions and critiques of this list and we will try to take things into account.

 

[JPOBS]

as long as peach > marth in the new list i'll be happy

 

[tarheeljks]

nm, not the place to continue this

 

[Geenareeno]

as long as peach < marth in the new list i'll be happy

Fixe d.