Success/Popularity stuff. Final update coming later today!

[John12346]

But the anti-ban side claimed that winning 30% or less in tourney "swiftly dismisses any claims of dominance!"

Anyway, here's the update for the week.

Update Information

Spoiler

Characters
MK: 695.27
Diddy: 132.85
Falco: 139.02
Lucas: 7.35
Wario: 183.65
Zelda: 17.35
Dedede: 15.85
Snake: 321.2
Marth: 118.67
Olimar: 89.2
Wolf: 66
ICs: 55
Lucario: 50.6
Pikachu: 29.5
Fox: 10
Peach: 46
Yoshi: 16
TLink: 6
Z-Sheik: 2.5

Users
MK: 900.6
Diddy: 231.1
Falco: 313.1
Lucas: 44.1
Wario: 220.4
Zelda: 54.1
Dedede: 61.1
Snake: 321.2
Marth: 180
Olimar: 89.2
Wolf: 66
ICs: 110
Lucario: 67.6
Pikachu: 59
Fox: 40
Peach: 46
Yoshi: 24.5
TLink: 12
Z-Sheik: 2.5

Players
Jtails MK: 63
Jtails Diddy: 63
NinjaLink Diddy: 44.1
NinjaLink Falco: 44.1
NinjaLink Lucas: 44.1
NinjaLink Wario: 44.1
NinjaLink Zelda: 44.1
NinjaLink Dedede: 44.1
Dekillsage Snake: 12.6
Malcolm Wario: 6.3
Mew2King MK: 264
Ally MK: 147
Pierce7d Marth: 88 <-- Tournament was probably stacked
Fino Olimar: 60
Holmes Wolf: 15
Arty Falco: 15
Blue Rogue Wario: 170
Lain MK: 85
Lain ICs: 85
Seagull Wolf: 51
Argent Lucario: 34
Argent Pikachu: 34
Fatal Snake: 140
Bizkit Snake: 60
Pelca Falco: 40
Pelca Fox: 40
Pelca Diddy: 40
Pelca MK: 40
Bloodcross Falco: 20
Bloodcross MK: 20
IcyLight Olimar: 10
Riot Zelda: 10
UltimateRazer Snake: 84
Gnes Diddy: 84
Trela Lucario: 33.6
Denti Olimar: 19.2
Fogo Snake: 9.6
DPhat MK: 9.6
Jem MK: 120
Nerd MK: 60
Nerd Falco: 60
Filip Falco: 30
DarkLink MK: 92
DarkLink Falco: 92
DarkLink Marth: 92
Kyle Peach: 46
Hades Dedede: 17
Hades Yoshi: 17
K9 Falco: 12
K9 TLink: 12
ESAM Pikachu: 25
ESAM ICs: 25
MVD Snake: 15
Poltergust Yoshi: 7.5
Ed Z-Sheik: 2.5

Tournaments Used
http://www.smashboards.com/showthread.php?t=306354
http://www.smashboards.com/showthread.php?t=306380
http://www.smashboards.com/showthread.php?t=306393
http://www.smashboards.com/showthread.php?t=306373
http://www.smashboards.com/showthread.php?t=306371
http://www.smashboards.com/showthread.php?t=306372
http://www.smashboards.com/showthread.php?t=306410
http://www.smashboards.com/showthread.php?t=306458

Tournaments Unused
http://www.smashboards.com/showthread.php?t=306475 - Added to the MK banned listings
http://www.smashboards.com/showthread.php?t=306317 - Free tournament
.
.
.
.
.
30+ Man Tournaments

Characters
MK: 453.5
Marth: 88
Olimar: 60
Wolf: 66
Falco: 15
Wario: 170
ICs: 42.5
Lucario: 14
Pikachu: 14

Users
MK: 496
Marth: 88
Olimar: 60
Wolf: 66
Falco: 15
Wario: 170
ICs: 85
Lucario: 34
Pikachu: 34

Tournaments Used
http://www.smashboards.com/showthread.php?t=306380
http://www.smashboards.com/showthread.php?t=306393
.
.
.
.
.
MK Banned Tournaments

Characters
Snake: 80
Diddy: 80
Lucario: 80

Users
Snake: 80
Diddy: 80
Lucario: 80

Character Usage
Snake: 2
Diddy: 2
Lucario: 1
Olimar: 1
Wario: 1
GaW: 1
Falco: 1
Marth: 1
Ike: 1
TLink: 1
Sheik: 1

Update Commentary
- Meta-Knight breaks $22000.
- Marth breaks $2000!
- Mr. Game and Watch loses viability... again. How many times has this happened now?
- Zelda's doing... alright, compared to the rest of the characters around her tier.
- All charts and graphs have been reorganized to match Tier List v6.

Please remember to Right Click and hit "View Image" on any image that isn't displayed large enough to be seen properly.

 

[Conviction]

But Snake is making 4x more than any other character! Let's ban him next

That's funny cause MK stops Snake's worrisome match ups from showing up

 

[Deleted member]

John, is this the first time Holmes has brought in money for Wolf?

 

[John12346]

No, there were a few other times, why?

 

[Deleted member]

Ah, okay. Just wasn't sure if he did after you started collecting this data.

 

[san.]

The data with MK removed is really interesting. I need to look at it a little more. As an aside, the Kirby numbers increased?

 

[Ishiey]

As San said, Kirby made more money with MK taken out... I don't think that's possible lol. Also, how did Zelda/Sheik's, Bowser's, Falcon's, Link's, and Jigglypuff's % increase with MK taken out? @___@

Using these statistics, if you take out everyone that uses MK, you take out three out of every four dollars won in tournament... well damn.

 

[Conviction]

I search "Well damn" now. So I'm here to co-sign Ish.

Well damn.

 

[Ishiey]

Just for you Iblis <3

 

[Flayl]

Also, how did Zelda/Sheik's, Bowser's, Falcon's, Link's, and Jigglypuff's % increase with MK taken out? @___@

Really? If they have the same amount of money, and the total sum of money is smaller, then they have a larger percentage of the money.

100 is a bigger portion of 500 than 100 out of 700.

I have no idea what the heck is going on with Kirby though, I assume it's a clerical error.

 

[Ishiey]

Really. Flayl, look at the chart again. Note how total %s are different for the two columns. Either way, even if that wasn't the case, it doesn't check out when you do the math.

 

[John12346]

Uh, that was my bad, lol. I made a few typing erros.

Gimme a sec there, k?

And just so everyone knows, the percentages are based off the original pot sum, which is currently $59709.35. The sum of money used WITHOUT splitting should NEVER be used when calculating any percentages, EVER. The only reason I left them in this time around was to show how drastically it changes when MK users are taken out of the equation.

 

[John12346]

Alright double post, w/e sorry

Here's the updated version. Not much changed, but hey, it's more accurate now. And it takes into account this week of tournies, so that's cool too, I guess.

 

[dainbramage]

Are you removing everyone who ever used MK, or just MK for a given tournament?


like if I used all falco in one tourney, then mk+falco in a second, would my first tourney count towards this?

 

[Conviction]

He's taken MK for the whole thing I believe on that chart.

 

[Flayl]

And just so everyone knows, the percentages are based off the original pot sum, which is currently $59709.35.

That makes no sense at all. You're using phantom data. To answer the question "What fraction of the money have Falco users won if you remove MK?" you need to take out the money won by MK. What you have there is an answer to a different question - if it weren't for the 0 next to MK: "What fraction of all the pot money have Falco users that don't use a MK secondary won?"

The sum of money used WITHOUT splitting should NEVER be used when calculating any percentages, EVER. The only reason I left them in this time around was to show how drastically it changes when MK users are taken out of the equation.

Yeah I'm not sure why that's there. This column of data is a Venn's diagram, there are overlapping outcomes but the total is the same as with splitting. Using %s makes sense if you use the original sum of money - I'm guessing that's what you did.

Anyway nitpicking aside great job so far.

 

[Ishiey]

And just so everyone knows, the percentages are based off the original pot sum, which is currently $59709.35. The sum of money used WITHOUT splitting should NEVER be used when calculating any percentages, EVER. The only reason I left them in this time around was to show how drastically it changes when MK users are taken out of the equation.

Whoops... my bad :x

So let's say that without any split, we took out all of the winnings earned with at least one successful use of MK in tournament. What percent of the total winnings would that be?



EDIT: thanks for the fixed data any chance you could make a chart with the split data like this, and have the %s sum to 100 in the column with MK users removed in it?

 

[John12346]

That makes no sense at all. You're using phantom data. To answer the question "What fraction of the money have Falco users won if you remove MK?" you need to take out the money won by MK. What you have there is an answer to a different question - if it weren't for the 0 next to MK: "What fraction of all the pot money have Falco users that don't use a MK secondary won?"

That was the request. >.<;

They wanted me to take all players who have won money with MK, and remove ALL of such players' characters from the listing, and recalculate the "No Split" column;

Yeah I'm not sure why that's there. This column of data is a Venn's diagram, there are overlapping outcomes but the total is the same as with splitting. Using %s makes sense if you use the original sum of money - I'm guessing that's what you did.

The only reason the sum of the No Split and Percentage columns are there is to show how much it changes when MK is taken out of the equation, that's all. If you notice, the total Percentage drops by over 100, while we lose about $60000 from the total set in the No Split column.

So let's say that without any split, we took out all of the winnings earned with at least one successful use of MK in tournament. What percent of the total winnings would that be?

any chance you could make a chart with the split data like this, and have the %s sum to 100 in the column with MK users removed in it?

Reword your request, please, it's kinda confusing to understand...

 

[Ishiey]

My bad lol, I don't try to make sense when I'm mobile xD

Basically, you said that I shouldn't use no split when calculating the amount of money won in tournament by people who won at least one match with MK. Is there any way you can collect that data? Like, make the chart you just made using split data instead of no split data?

Also, Wolf for 5th most viable character when all MK users are removed

 

[san.]

The data for characters with the non-MK money shouldn't be related to each other, but with the MK data for the same character. You can't really draw any conclusions for viability of characters, well other than how much it seems MK is needed to do well in tournaments in an MK dominated scene.

 

[John12346]

Basically, you said that I shouldn't use no split when calculating the amount of money won in tournament by people who won at least one match with MK. Is there any way you can collect that data? Like, make the chart you just made using split data instead of no split data?

Unfortunately, no, I can't. The only data I have to go on are the charts I present each week, along with character breakdowns, which does not include pot splitting. I haven't been keeping track of how much of a given character each player uses, so...

In either case, any interpretations can be reliably made with the "No Split" aspect of the charts, so it shouldn't be much of a problem.

The data for characters with the non-MK money shouldn't be related to each other, but with the MK data for the same character. You can't really draw any conclusions for viability of characters, well other than how much it seems MK is needed to do well in tournaments in an MK dominated scene.

You're gonna have to reword this if I get it wrong, but I think what you want is how much each character has when all of the NON-MK users are removed from the chart?

 

[san.]

I wasn't asking for a request. I was mostly responding to Ishiey's claim about Wolf being 5th most successful, saying you really can't compare it that way. Sorry about that.

 

[Ripple]

hey john, on the chart you gave me, the %a aren't recalculated once m2k and aly are removed. can you fix this please

 

[Blacknight99923]

removing everyone who uses metaknight at all is going to create false data considering how many people will have a tendancy to switch to metaknight because of being or cping someone to counterpicks (or in cases of fox mains switching to metaknight against shiek/pikachu/ic's)


I'd be interested in seeing data where people only used character MAINS only though (including metaknight)

abiet it to would be a bit skewed because people do use metaknight

 

[san.]

That's exactly the data I was looking for, Blacknight, to gather info on MK secondary usage and to see how many characters had a tendency to switch/rely on MK to move ahead and make money. It could have been any other character people chose, but MK seemed very high.

Based on the way data was gathered, you couldn't tell who mained MK and used those other characters a few times, or the other way around.

 

[John12346]

hey john, on the chart you gave me, the %a aren't recalculated once m2k and aly are removed. can you fix this please

Yes they are.

Keep in mind that I'm dividing over the ORIGINAL POT SUM from all tournaments, which is a constant no matter how much money I may end up removing from any part of the column. Since removing M2K only removed money from MK, the charts of "w/o Ally" and "w/o Ally and M2K" are almost exactly the same.
.
.
.
.
.
As far as the inherent problems pointed out by Blacknight and San, I'll see what I can come up with. I should be able to rig a formula to get this to work the right way.

...If you have any suggestions as to what that formula might be, don't hesitate.

I was thinking I could rig a logic sequence like this:

Spoiler

How does that sound?

 

[Ripple]

updated! plus I added %s in this time to let you know exactly how popular each character is.

news:

Mk went up again but only like .07%

falcon was the only character that didn't have any tournament uses in an update period.

 

[John12346]

Alright, this chart should be something to ease all of your concerns. However, because my charts tend to confuse a lot of people for some unknown reason, I'm explaining how and why each value in each column has come to be, because this chart is really freaking important.

We will assume the Character Column is Column 1, and work from there.

Column 1 - The list of characters, set in the order of Tier List v6.
Column 2 - Completely empty, save for the total set at the bottom. That value is the total amount of money that has circulated 2011 tournaments so far. The constant of $57909.35 is that such value.
Column 3(User) - A column you know all too well. The column that shows how much money each character has won in 2011, when we neglect splitting between characters.
Column 4 = (User/57909.35)*100

Column 5(Special) - How Special is Calculated:

Spoiler

Column 6 = (Special/57909.35)*100
Column 7(Difference) = (User - Special)
Column 8: Special 2 = (Difference/57909.35)*100
Column 9: Special 3 = (Difference/User)*100

Column 9 is very important, because it indicates the percentage of money lost from each character's original total, after hard MK users(denoted by Special) are removed from the standings. The fact that there's wild fluctuations of large percentages flying around, coupled with the enormous Total in Column 8, attests to the reliance that players have on MK in addition to their other characters.

In addition to that, when we consider Column 6, which shows the percentage of money that the users of each character have after hard MK users are removed from the standings, the game's balance when MK is out becomes extremely clear.

 

[Kaffei]

this thread is awesome. ty john numbers

 

[Ishiey]

I was just joking San, I know that it doesn't really correlate to viability

@ John#s: So if I did...

(total cash lost from formula - MK's winnings) / (total cash won - MK's winnings)
or
(61114.18 - 31746.63) / (94218.47 - 31746.63) = 47%

... does that mean that, using your formula, MK's assistance almost doubles the winnings earned from the rest of the cast?

 

[John12346]

Er, if you'd want to do that, you would probably want to go with this formula:

(Total Cash Lost/Total Cash Won by Users)*100

MK's winnings shouldn't really be factored in at any point. You just need a point of reference to compare the overall amount of money lost to. In this case, we use the sum of money won by the users of each character.

(61114.18/94218.47)*100 = 64.8643%

There's no real way to interpret this value in terms of users, and we can't use the split earnings sum(57909.35) as our denominator, because that would give inaccurate results(exceeding 100%, actually, which is not possible).

What we CAN assume, however, is that if we were to use the split earnings sum(again, 57909.35) as our denominator, we could assume the numerator to be reduced by a proportional amount if I were to perform the same function on a list of players' characters taking splits between characters into account. For leniency on the overall interpretation of the percentage, and to take standard deviation into account, we can just give or take 5% in either direction.

So, we could say that:

If we were to remove all traces of hard MK users - secondary characters included - from the Brawl community, we would lose somewhere between 59.86% and 69.86% of all tournament money that has been in circulation for 2011.

...And if you haven't picked up on it at this point, that percentage is WAY TOO ****ING BIG HOLY ****

 

[Ishiey]

Well, idk... doesn't that inflate the percentage since without removing MK's winnings from each category, we're also adding in the percentage of money lost by taking MK's winnings out since all of his mains are hard MK users? I suck at explaining things Like what you said for alternate accounts skewing votes though. 7/10 < 17/20. Y'know? I'm pretty certain about this one :x

Either way, yeah, since it's without a split the numbers won't really be precise, but they still give a decent indication.

 

[John12346]

This is starting to make my head hurt... I know what exactly what you're saying, I'm just not sure which of our formulas are correct... I'm gonna try to lay it all out on the table and see what we can come up with.

My formula: If we were to remove all traces of hard MK users - secondary characters included - from the Brawl community, we would lose somewhere between 59.86% and 69.86% of all tournament money that has been in circulation for 2011.

Your formula: If we were to remove all traces of all hard MK users' secondary characters from the Brawl community without actually removing MK himself, we would lose somewhere between 42% and 52% of all tournament money that has been in circulation for 2011.

My formula takes MK into account, while yours does not. Both are equally important, because mine shows just how much money MK users as a whole are taking(way too much, as you can see), and yours shows how much of a splash effect of using MK alongside other characters actually does to these other characters' winnings.

Both interpretations are correct and awesome, and show how much of a ****** MK is.

HOWEVER! These interpretations can be combined. I'm thinking of a way to do that now, but it's going to take a lot of time. This is really complex now... -___ -;

 

[John12346]

Sorry about the double post, but this would be very out of place if I edited this into the previous post.

I'm just going to go ahead and put down all of the true statements we've figured out.

1. Meta-Knight users have won 53.69% of all tournament money in 2011(as indicated by my chart in the 2nd post).
2. Meta-Knight, along with the secondary characters used by Meta-Knight users have won between 59.86% and 69.86% of all tournament money in 2011.
3. The secondary characters used by Meta-Knight users have won between 42% and 52% of all tournament money in 2011.

This all sounds legit so far, but then I noticed that, when I subtract Statement 1 from Statement 2, I get your Statement 3, but with a percentage range of 6.17% to 16.17%.

My info seems to indicate around $5000, while yours seems to land in the ballpark of $22500 for Statement 3. Keep in mind that these percents and dollar values are now based off all of the money that has circulated Brawl tournaments for 2011; $59709.35.

What do you think, Ishiey? Any input...?

 

[Ishiey]

My formula takes MK into account, while yours does not. Both are equally important, because mine shows just how much money MK users as a whole are taking(way too much, as you can see), and yours shows how much of a splash effect of using MK alongside other characters actually does to these other characters' winnings.

Both interpretations are correct and awesome, and show how much of a ****** MK is.

I see :3 Yeah, that sounds right. I was looking into MK's effect on the winnings from the rest of the cast, so... basically

I have work until the evening :/ but when I get back I'll look over your second post and comment on that. At a glance, it seems like the error in statement 3 is that the 42-52% is in relation to the rest of the cast's winnings, not MK's winnings. So it would be closer to ~60% of the range, or 28.2 +/- 5%. There's still a big gap though... X_x I'll think about it and get back to you after work.

One more question. Let's say I'm using the split column to calculate this. If MK makes 37% and the rest of the cast makes 63%, but 47% of the rest of the cast makes their money with MK's assistance, 37% + (63%)(47%) is approximately 66.6%. So... the use of MK is involved in 2/3rds of tournament winnings? Would this logic be correct?

 

[BSP]

John, I think you should keep Special 3 in from now on.

Also, this is sad

 

[John12346]

Alright, well, this is probably the last issue that needs to be worked out, then we can start drawing conclusions...

It just doesn't seem that natural for the other characters to have won THAT much to the point that it almost rivals the amount of money MK has made... it makes it seem like the other characters were responsible for bringing in all that money, rather than MK himself.

How much money has MK earned for the other characters in the game? $5000(indicated by my 11%), or $22500(indicated by your 47%)? When I was removing characters according to Special, we generally lost a very, VERY large majority of the characters from the Top 11 along with Toon Link. That's a total of 11 characters, which means that MK earned around $454.55(indicated by my 11%) or $2045.45(indicated by your 47%) for each individual character in the game.

Just eyeballing that, it doesn't seem possible for the latter to be true, because a lot of characters(in the Split column) haven't even won that much money yet.

So... I'm gonna be bold and say the following to be true: Meta-Knight users have won 53.69%($32057.95) of all tournament money in 2011, and as a result of MK mains using side characters, other characters in the game have earned somewhere between 6.17% to 16.17%(roughly $5000) of all tournament money.

The red text is what's under scrutiny right now. If I've made an error somewhere, Ishiey, point it out, alright?

 

[Ishiey]

Well, let's see. I'm on the same page with your first two statements from before. For the third statement, I posted this on the last page (just reposting to make sure you didn't miss it by accident, with some edits because I messed up the numbers >_>)

At a glance, it seems like the error in statement 3 is that the 42-52% is in relation to the rest of the cast's winnings, not MK's winnings. So it would be closer to 66.3% of the range multiplied by 53% of their winnings being made without MK being involved, or 35.15 +/- 5%.

So far, without factoring in any error, we have:

MKs winnings + the winnings of other characters that are won with the side use of MK: 64.86% (from post 791, I recalculated this with a different method and got the same number. Using the data we have, this is DEFINITELY as accurate as it gets)
The winnings of other characters that are won without MK: 35.14%

Sum of the two is 100.01%, which is basically where it should be, so I think these are correct. This means that 66.3% - 35.14% = 31.16%, same as 66.3% times 47%, which is the percentage of money made by characters in tournament without using MK. Idk man, I'm a good math person and all of this is checking out perfectly, but I still can't find a way to explain why your 11% isn't correct X______x

 

[John12346]

Alright, you've convinced me that what you did is the correct formula, but just so we're on the same page, take that cyan and red statement from my previous post and implant the correct numbers, so I completely understand what you're referring to.

 

[Ishiey]

Gotcha. There is some slight rounding here, because... I can just chopping off decimals, but all the data is in the collapse tag in case anyone wants to follow up. Keep in mind, this is using no split data, so there may be some errors and everything is taken off of the split total ($94218, 157.8%) as a basis for 100% of the tournament money. I think this inflates MK's numbers a bit because he's the most common secondary and blabla but too braindead to bother proving or explaining it, all the math is correct though. Too lazy to put in a +/- spread, people can do that part themselves but it should be assumed that these numbers are not 100% accurate concerning the metagame.

-------------------------------------

Meta-Knight users have won 33.7% ($31746/$94218) of all tournament money in 2011, and as a result of MK mains using side characters, other characters in the game have earned 31.2% (~$29368) of all tournament money.


Numbers / Calcs:

Winnings for MK only: 31746/94218 = $31746, 33.7%

Winnings for characters besides MK that were won with the use of MK as determined by the formula: (61114-31746)/94218 = $29368, 31.2%

Winnings for all characters (including MK) with the use of MK as determined by the formula: (29368+31746)/94218 = 61114/94218 = $61114, 64.9%

Winnings for characters without the use of MK as determined by the formula: (94218-61114)/94218 = $33104, 35.1%

Sum of the previous two categories, or total tournament winnings: 100%

[collapse=Data Used]

We will assume the Character Column is Column 1, and work from there.

Column 1 - The list of characters, set in the order of Tier List v6.
Column 2 - Completely empty, save for the total set at the bottom. That value is the total amount of money that has circulated 2011 tournaments so far. The constant of $57909.35 is that such value.
Column 3(User) - A column you know all too well. The column that shows how much money each character has won in 2011, when we neglect splitting between characters.
Column 4 = (User/57909.35)*100

Column 5(Special) - How Special is Calculated:

Spoiler

Column 6 = (Special/57909.35)*100
Column 7(Difference) = (User - Special)
Column 8: Special 2 = (Difference/57909.35)*100
Column 9: Special 3 = (Difference/User)*100

[/collapse]