Character Ranking List - Project M - updated 4/4

[Overswarm]

Credit for this original idea goes to Ankoku, now Agidyne I believe. I will be copying some of his style, as it proved to be fairly accurate, and some of the text he used.

What this is:

Not a tier list. Merely what characters HAVE been successful, and how successful.

The math behind the list is as follows:
Base values are
1 for top eight
4 for top four
7 for second
10 for first
Base values are then multiplied by number of entrants and entry fee, then divided by 160. This helps account for larger tournaments being more relevant than smaller ones. If two characters are listed, both gain half the points. If you have three, you gain 1/3rd for each.

If you have more than 3, I cut you off at 3.

Please note that all data being used is referring to the Top 8 players only. If there were 50 links in your 58 man tournament, but none of them made top 8, Link receives 0 points and no mention in usage.


I will not use tournaments that

• Have less than 10 entrants
• Have less than a $5 entry fee
• Have anyone in the money split (not play out their set/forfeit to exchange money)

How are categories formed?
Fairly arbitrarily, using the Average as a baseline for "C" and averages of relevant clusters for further gaps. As such, a characters "tier" is determined by the success or lack of success of others. Because this is a zero sum environment with no easy metric in which to base it off of, we have to either do it completely arbitrary (this % is S tier, this % is A tier, etc.) or do it relativistic to the surrounding data points with less arbitration and I choose the latter. No idea how Ankoku did it but I'm assuming something similar.

Format on detailed list is as follows:
# Character (character placements) - points - previous rank #


Orange means the character moved up in Rank significantly
Lime means the character moved up in points; a ^ signifies an increase in standing compared to others
Red means the character moved down in points or rank due to removal of old data
Yellow means the character did something silly, like go up in points but down in category, or vice versa
Cyan means the character was previously unranked
Green means the character is Luigi

A tier <overused>: 27.74
- 88 (.11.33%) 20.13
- 66.48 (8.55%) 10.84
- 61.10 (7.86%) 8.97
B tier <Standard> 44%
- 48.09 (6.19%) 6.19
- 46.87 (6.02%) 5.61
- 36.97 (4.50%) 0
- 35.04 (4.50%) 3.60

- 33.69 (4.33%) 22.78
- 33.34 (4.30%) 7.53
- 33.12 (4.26%) 5.36
- 29.98 (3.86%) 9.19
- 20.53 (2.64%) 3.44
- 20.09 (2.58%) 2.41
- 18.25 (2.35%) -0.08
- 18.18 (2.34%) 0
- 18.08 (2.33%) 5.19

C tier <Borderline>12.4%
- 15.99 (2.08%) 9.55
- 14.89 (1.92%) 2.69

- 14.07 (1.81%) 2.33
elda - 13.61 (1.75%) 1
- 13.54 (1.74%) 1.17
- 12.39 (1.59%) 7.69
- 11.70 (1.51%) 3.48


D Rank <Underused> 6.92%

- 9.25 (1.19%) 0
- 8.92 (1.15%) 1.31
- 8.61 (1.11%) 4.21


- 7.47 (0.96%) 1.25
- 7.29 (0.94%) 2.625
- 6.23 (0.80%) 0.53
- 6 (0.77%) 2.42
E Rank <Never Used> 1.55%
- 5.61 (0.72%) 3.75

- 4.08 (0.52%) 0.33
- 2.42 (0.31%) 1

F Rank <Pardon the Dust> 0%


Captain Falcon is the first to lose points because I noticed a tournament that had "Falco" instead of "Falcon", thus giving Falcon some points he should not have had. I have since fixed this error, my fault for not calling him "Captain".

Tournaments Used: 23 (+4)
Project M Showdown
Sa2vation
Smash Rehash 11
Sweet IX: Winter Opener!
Rain!
Lordy's Monthlies v5
Kirkgo's Biweeklies
Plenty of Corn
Sweet X: Clash of the Dons
Summit 6!
Bowser's Revenge 6
January Gameclucks
February Gameclucks
Melee in the Hold 8
Something2Do
Dope 7
Rolla Monthly 7
Xanadu Games Harlem Shake
Yamcha Sucks
Rain 2
Lake Effect
East Texas Fight Club R1
SCSYN5
Kirkgo's Pancake Kittens
Bowser's Revenge 7
SiD14 Quebec Returns
BC Brawl Monthlies
Blue Screen Monthlies
Above the Influence
And then I dragged
Outrage I
Lordy's Monthly v7
Sweet XII
Gameclucks March
Melee In the Hold IX


History
Feb 17th, 2013

Spoiler

A tier <overused>: 24.23%
- 39.7- (12.59%)
- 36.70 (11.64.%)
B tier <Standard> 45.26%
- 22.26 (7.06%)
- 21.08 (6.69%)
- 13.64 (4.33%)
- 13.58 (4.31%)

- 12.99 (4.12%)
12.93 (4.10%)
- 12.49 (3.96%)
- 11.92 (3.78%)
- 11.47 (3.64%)
- 10.32 (3.27%)
C tier <Borderline>13.38%
- 8.67 (2.74%)
elda - 8.57 (2.72%)
8.39 (2.66%)
8.30 (2.63%)
8.28 (2.63%)
D Rank <Underused> 11.37%
7.16 2.27%)
6.21 (1.97%)
5.78 (1.82%)
4.78 (1.52%)
4.55 (1.44%)
3.75 (1.19%)
3.65 (1.16%)
E Rank <Never Used> 3.02%
- 2.84 (0.90%)
- 2.42 (0.77%)
- 2.19 (0.69%)
- 2.08 (0.66%)
F Rank <Pardon the Dust> 1.4%
- 1.53 (0.49%)
- 1.47 (0.47%)
- 0.9 (0.29%)
- 0.46 (0.15%)
- 0

Feb. 26th

Spoiler

A tier <overused>: 30%
- 62.5 (11.26%) +22.80
- 53.56 (9.65%) +16.86
- 50.46 (9.09%) +38.54
B tier <Standard> 47.11%
- 36.97 (6.67%) +29.81
- 34.62 (6.24%) +12.36
- 28.22 (5.09%) +15.30
- 25.50 (4.60%) +12.51
- 23.60 (4.25%) +15.30
- 23.52 (4.24%) +2.44
- 20.20 (3.64%) +6.56
- 18.33 (3.30%) +4.76
- 17.64 (3.18%) +7.31
- 17.09 (3.08%) +8.70
- 15.66 (2.82%) +4.19

C tier <Borderline>11.69%
- 12.48 (2.25%) +0
elda - 11.80 (2.12%) +3.23
- 10.95 (1.97%) +8.53
- 10.91 (1.97%) +4.69
- 10.11 (1.82%) +1.83
- 8.66 (1.56%) +0

D Rank <Underused> 8.83%
- 6.98 (1.26%) +2.44
- 6.45 (1.16%) +.072
- 6.22 (1.12%) +1.44
- 6.19 (1.12%) +5.28
- 5.85 (1.05%) +3.77
- 4.79 (0.86%) +1.14
- 4.70 (0.84%) +2.52
- 4.13 (0.74%) +2.66
- 3.75 (0.68%) +0

E Rank <Never Used> 1.05%
- 3.17 (0.57%) +0.33
- 2.67 (0.48%) +1.14

F Rank <Pardon the Dust> 0.08%
- 0.46 (0.08%) +0
- 0 (0%) +0

to April 4th

Spoiler

A tier <overused>: 28.8%
- 68 (11.14%) +5.5
- 55.64 (9.12%) +2.07
- 52.14 (8.54%) +1.67
B tier <Standard> 47.04%
- 41.26 (6.76%) +6.64 ^
- 36.96 (6.06%) +0
- 31.44 (5.16%) +5.94 ^
- 31.40 (5.15%) +3.17
- 27.76 (4.6%) +4.17
- 25.80 (4.23%) +2.28
- 20.79 (3.40%) +.59
- 18.33 (3%) +0
- 18.18 (2.98%) +.54
- 17.69 (2.90%) +2.03 ^
- 17.09 (2.80%) +0

C tier <Borderline>13.41%
- 12.90 (2.11%) +4.06
elda - 12.61 (2.06%) +.81
- 12.37 (2.02%) +1.42
- 12.20 (2%) +5.22 ^
- 11.73 (1.92%) +1.63 ^
- 10.91 (1.79%) +0
- 9.25 (1.51%) +.59

D Rank <Underused> 9.07%

- 8.22 (1.35%) +2.38 ^
- 7.61 (1.25%) +1.42 ^
-6.45 (1.06%) +0
- 6.22 (1.02%) +0
- 5.70 (0.93%) + .91
- 4.70 (.77%) +0
- 4.67 (.765%) +.54
- 4.40 (.72%) +1.73 ^
- 3.75 (.61%) +0
- 3.58 (.59%) +.41
E Rank <Never Used> .54%
- 1.86 (.31%) +1.41
- 1.42 (.23%) +1.42


F Rank <Pardon the Dust> 0%

 

[Overswarm]

Last Edited: March 12th

Grand Chart (Point Total and Times Used):

Money Won:

Points Earned:

Times Used:

Point Increase: 3/12-4/4 update

Fun Information:

Total Money Won: $3850.70 +$440
Total Points: 777.5625 +167.65

To determine % of total points earned or money won for your favorite character, simply divide the money or points your character has earned by the total money or points, respectively.

Total number of Solo Character players - 190
Total number of Dual character players- 68
Total number of Tri character players - 22

Top 10 characters % of money won:
- 14.33%
- 10.18%
- 12.09%
- 7.06%
- 6.82%
- 5.88%
- 5.67%
- 4.68%
- 3.71%
- 2.98%


Tournament Winners:
Wavebuster
Chillindude
PP
Mango
Axis
SuperTH
Metroid
Kels
Mew2King
SDSG
Sethlon
Gimpyfish x2
Silent Wolf
Overswarm
Geno Blast
Fishburne
Dr. Peepee
Majora88
HankyPanky
Slayertokey
Dakpo
Sethlon x2
Kirk
HolyNightmare
Firefly
InternetExplorer
Reflex x2
Axis
SuperTH
Mew2King
Silent Wolf
Blunted_Object

Use the list above to see which players should potentially follow, and also as a quick way to see if I'm missing a tournament!

Characters that have won a tournament SOLO:




















Implications 4/4:

The game is still going through its growing pains and there's still few enough data points to where a single decent tournament victory can catapult a character (see: Falco, Ivysaur)

Given the trends I've seen, I think we can expect many of the "B" tier character to remain where they are or go down as time goes on. Without significant changes, Bowser, Fox, and Sonic are likely to be high tier characters for the foreseeable future. They have all had success with multiple players, meaning that they have proven themselves against the B tier characters who ALSO have multiple players (for the most part).

I'd have to say D rank surprises me the most. ZSS, Lucas, Ness, Dedede, sure. They're all obviously bad in comparison at this point to the top tier characters that are doing well and don't have quite the same possibility for improvement. But Pit and Snake? Both characters have seen prominence in Europe and have at the very least have what appears to have potential... but they have shown little in the way of success here.

All in all a pretty boring update. A few characters with small successes got a bit of a bump, but nothing too substantial has occurred. Bowser, Fox, and Sonic have continued to show them all being above the rest. Power, speed, and more speed!

3/12

Spoiler

Small update!

Sethlon brings Marth into a tournament victory, giving Marth an additional 5.9375 points. If you'd like to imagine Sethlon would have done the same with Sonic you can mentally add that to Sonic's tally.

Starting to see some gradual breakaway from the initial results. Whenever you see the % in the A rank go DOWN and the lower ranks go UP, the more variety the game has. As the game goes on people will drop "bad" characters and pick up good ones save for a few people who just love the character more than winning. When this happens you'll notice homogenization of tournament results, similar to Melee and Brawl. Lots of top tiers, little else. When that happens you should see A rank get an INCREASE in points, possibly B rank, and the lower ranks get less points while gaining additional characters to their roster.

ROB has been an interesting character to watch; he's decently popular, being used the same amount as Sonic (10th or 11th most used), but can't quite make a splash. While ROB is being used a lot and getting decent placings, there doesn't seem to be a Sethlon or Gimpyfish with a strong tournament presence for ROB which is holding back his placement. Whether this is a result of the character's limitations or the limitations of the players or the region they play in remains to be seen. If we see ROB players switch and pick up a new character and do better, we can likely attribute ROB to the discrepancy. Whenever you see a character used a decent amount and getting decent placings, they always deserve close observation.

If a character is used rarely and gets awesome placements when used, it's likely the player.
If a character is used often and gets crappy placements when used, it's likely a bad character.
If a character is used rarely and gets crappy placements, it is a bad character.
If a character is used often and gets awesome placements, it is a good character.

If a character is used a decent amount and gets in-the-middle kind of placements with a spattering of top 3 here and there, it doesn't necessarily imply that the character itself is good or bad. People drop bad characters pretty fast and those that don't become irrelevant pretty quick. People pick up good characters pretty fast and the points and usage reflect it pretty quick. But if ROB's numbers/placements remain static, he fits in a grey area.

Pikachu got his first tournament result, sharing a 2nd place with another character, but still is in dead last.

Bowser, Sheik, and Marth were the only ones to INCREASE the amount of money they won in this last cycle. There weren't a lot of tournaments, but it is important to note that of all the characters that did win money, only 3 of the top 10 managed to make the money placements. If you see this happen consistently over more tournaments, expect a drop in their point performance over time as well.

Captain Falcon is kind of a failure in terms of money won, which is a little surprising! He's doing about as well as Zelda. Mario himself is the absolute biggest failure. There are 10 tournament Marios to date, and he has made $0. Comparing this to Bowser who has 14 tournament players and has made $516 and you can see how big the gap is. It may not be safe to say that Mario is not a tournament viable character from his results alone, but it can be said that none of his players are high level players. When a character is rarely played by top players, it often means that character is strictly limited in some way.

All in all, the metagame still looks healthy and, while it has settled somewhat, it's still gradually changing and is not in a static state.

The only "unhealthy" thing current in the game is how immensely popular Fox and Sheik are. This would normally imply that they are versatile or easy-to-pick-up characters, but we already know the real reason is that they are good and they were big in Melee.

Given this, the most important matchups to know in general will be the Fox and Sheik matchup.


SLIGHT UPDATE with some new info:

I was curious as to how "healthy" Project M's game was, statistically speaking, so I looked at the variance.

For those unfamiliar, it basically boils down to "the larger this number is, the greater variance"

Usage variance: 29.90
# of Characters that can be considered "popular" using the mean as a baseline: 14 (Fox to Peach)
Points variance: 269.82
# of Characters that can be considered "viable" using the mean as a baseline: 12 (Bowser to Captain Falcon)


We have a relatively health game at this point, with 12 characters being considered mathematically viable and 14 being considered popular.

I don't have usage statistics for Brawl or Melee, nor do I have Character ranking data relevant to our tiem period, but I DO have Character Ranking data from a few years into Brawl.

Post-March 2010 Brawl Points Variance: 619,444.2

Compare that to Project M's 269.82 and you can safely say holy crap.

Now this is primarily due to Meta Knight's dominance and Snake's sloppy seconds.

Removing Meta Knight Brawl Points Variance: 181,146 (~440,000 less)
Removing Meta Knight and Snake Points Variance: 76176.85

This is good information, as Brawl Post-March 2010 was a dominated game that was basically already solved. The changes haven't been that extreme since then and most involve characters shuffling places rather than anything else.

So if in the future I or someone else uses data in this way and says "Bowser's OP, he wins too many tournaments" or something of the sort, you can point to the variance present.

Project M's variance WILL increase as people figure out matchups and characters more, but I do not forsee the same variance we saw in Brawl.

Spoiler

After studying Agidyne's/Ankoku's list during the Brawl era (heavily during the MK debate) I noticed several trends throughout time. A sort of collection of "rules" that are rarely broken.

Rule #1: If a character's "times used" is significantly lower than their "point total", you have a near guarantee that this is a top character or at least has the potential to be in the hands of a good player. The greater the disparity, the more potential this character has.

Rule #2: The lower a characters "Times used" is, the more likely its place on the list will change drastically in the future.

Rule #3: The earlier in the list, the high of a boost "simple" characters will get. It took nearly 2 years for Snake to be dethroned in Brawl, despite there being several characters that can compete. Looking at this list I can forsee a similar thing with Bowser.

Rule #4: If a character's "times used" and "point total" is very close to one another (grand chart) for any significant amount of time, it is likely that this character is on a downward slope. Looking at characters like Ike, Ivysaur, or ROB and you can see good examples. Good characters that are being adopted typically shoot up the list quickly, while bad characters will slowly fall.

Rule #5: If a characters' "times used" significantly overshadows their "point total", this character is likely a 'noob character'. Commonly loved, but not really as great as people tink. Think Link in Melee. Looking at this we can see Mario, Squirtle, Toon Link, ZSS, Wario, and Dedede as good examples. As we get more data, it will become more apparent if these characters are actually underpowered compared to the rest of the cast or not. Currently they are some of the least successful in comparison to their usage.

Read the "times used" as a "what matchups do I need to know when I get to the top 8".
Read the point total as "Who is actually being successful?"
Read the money won as a similar indicator, but also as a balance against points.

Hope you enjoy this!

 

[Overswarm]

I have some more specific data (in non-chart form) I'd like to add to this, but know there are many tournaments that Metroid still needs to add to his thread. Until then, what kind of info would you like to see?

 

[Jesseyo]

Interesting stuff

 

[BTmoney]

Ooooh, I can't wait for someone to get mad over this

 

[dookdigity]

SO MAD!

 

[Tlock]

You need to scale the money won in the grand chart somehow. If there are just one or two large tournaments with pots over $1,000 you will not even be able to see the #of times used at all.

 

[Oro?!]

Sonic is about to jump to top of money won...

 

[Kink-Link5]

Something really aught to be done showing the point totals and usage ratio.

Also show the sample sizes for both entrants and total tournaments factored in.

 

[Gimpyfish62]

http://www.youtube.com/watch?v=Ll3uipTO-4A

 

[Spiffykins]

http://www.youtube.com/watch?v=Ll3uipTO-4A

Why else would anyone play stupid fake fighting games like smash anyway

 

[Overswarm]

You need to scale the money won in the grand chart somehow. If there are just one or two large tournaments with pots over $1,000 you will not even be able to see the #of times used at all.

The Grand Chart is meant just as a curiosity. It is never going to be pretty, and each aspect of it has its own chart below. It's just meant as something you can look at quickly and go "oh, neat" or maybe learn you want to look at something else more closely.

Something really aught to be done showing the point totals and usage ratio.

Also show the sample sizes for both entrants and total tournaments factored in.

Point totals already factor in sample size and entrants, and points are cut in 1/2 or 1/3 (as is money won) when a player uses a secondary or tertiary character. I'm not interested in showing some sort of point/usage ratio because that'd be accounting for usage twice in some areas and would inflate the results of crappier characters.

G&W sucks, right? Well G&W won a tournament recently. Do we need to inflate G&W's win there to compete with how many more Foxes there are? There'd be no point. It'd be artificially trying to make the results more "even".

This isn't a tier list, it's just a reflection of what is actually happening. There is no theory craft here. Fox is the most popular character. Fox has the highest point total, meaning he has gotten many high placings in tournaments that have factored entry fee and entrants in (under the presumption that larger entry fees mean harder opponents, as it weeds out the lower level players). Bowser has won the most money.

This can all change (and likely will when Metroid updates his main thread and I can finish this), but this is a reflection of the currently used data. What you want to pull from this is up to you.

 

[Kink-Link5]

What's actually happening is that tournaments have results past top 8. You are only including the instances of characters getting those results in your data? Since this was not made clear in either post, it leaves your data skewed and unclear. When you have a chart showing "Times used" it would stand to reason that this would show the times a main of the character enters a tournament, NOT how many times the character has gotten top 8. Include all instances of a character being used in tournament in "times used" when possible and it won't artificially inflate the results of ****ty characters, it will more accurately reflect the fact that, hey, maybe Fox wins so often because he's used so often. Or not. I don't know how often Fox is used because you won't show it.

It still doesn't show the total tournaments and entrants.

 

[Overswarm]

The tournaments used are listed in Metroid's post; I pull every tournament from there that fits my criteria.

I also don't care that there were 5 Dededes going 0-2 in bracket. Top 8 is considered "good" in tournament play and means that usage of that character is signifcant.

1 for top eight
4 for top four
7 for second
10 for first

I am only using top 8 placements. This is a tried-and-true method over the course of Brawl's lifespan and it accurately showed which characters were good and which were bad and showed metagame trends before they occurred.

If you think that more people using Fox and going 0-2 in a tournament somehow makes Fox more likely to gain significant points using this model, you are mistaken. It is a merit-based achievement, not a lottery.

 

[Kink-Link5]

Okay.

It's just weird to me to intentionally leave out relevant information regarding character usage, without stating the sample size in any form, and awarding points without really explaining why the points awarded were chosen. It looks sloppy from a statistical angle.

 

[Overswarm]

Okay.

It's just weird to me to intentionally leave out relevant information regarding character usage, without stating the sample size in any form, and awarding points without really explaining why the points awarded were chosen. It looks sloppy from a statistical angle.

It's been used for like 4 years and has been more accurate than the Melee tier lists for their entire existence, so... eh. :B

I don't have all the stuff up yet that I'm going to be adding yet. This is just a preview while I wait for Metroid to update, as I don't want to re-do the same amount of work for no reason. There are several tournaments that still need to be added.

 

[Kink-Link5]

I find it kind of hard to believe that the Brawl list is very accurate when it lists Olimar at second and Ice Climbers as low as fourth.

All I want is to know the totals for a character being entered and the number of tournaments being used. You don't need to recalculate or run anything more through, just show your data points.

 

[Overswarm]

http://smashboards.com/threads/165954/

I think you're looking at the wrong list. Also, just wait for Metroid to update his thread.

 

[Juushichi]

I also would like to think that, had that list been updated for as long as the metagame has continued it would have listed Yoshi, Olimar, Zero Suit Samus, Wolf, and etc as being stronger than they were circa Spring 2010.

Because all of these characters (especially Olimar after APEX 2011?) had drastically improved in placings since that point with the surge of success seen by multiple players of those characters, I think the above would have been reasonable to assume.

 

[Zankoku]

Four years later and my legacy of an idea for having a ranked list before the release of an official tier list still sees use. I'm so proud.

 

[Overswarm]

Four years later and my legacy of an idea for having a ranked list before the release of an official tier list still sees use. I'm so proud.

It was a good influence and shut a lot of people up. It was also an incredible warning system for Meta Knight. While this game won't have a problem with bannable characters due to the developers being able to change the offending character, it will help them realize when a character might be too strong/weak/unpopular.

Pikachu being a good example. He currently is super unpopular. Isn't even discussed, really. Might mean he needs some changes just to make space in the game itself! This list can track that stuff. You did good sir. You did good.



Also, really wish I had the excel file you used for formulas. I had to figure out how to set this all up so it was automatic and that was a pain! It's all good now though.

 

[Zankoku]

Ahahaha, Excel file with formulas? Until like mid 2010, I did everything with Notepad and Windows' Calculator, then copied the numbers into Excel. I did eventually have everything done in a program, but that wasn't until quite a bit later.

 

[Overswarm]

Holy geeze. It took me an afternoon to learn how to do it in Excel and now it only takes me a few seconds to update it. Efficiency man, efficiency!

Although manually doing the numbers is very soothing.

 

[Kink-Link5]

http://smashboards.com/threads/165954/

I think you're looking at the wrong list. Also, just wait for Metroid to update his thread.

Used for four years? It was last updated three years ago and Brawl is only 5 yeasrs old.

Did the Brawl community really think Snake was #2 at one point?

 

[Red(SP)]

Used for four years? It was last updated three years ago and Brawl is only 5 yeasrs old.

Did the Brawl community really think Snake was #2 at one point?

Yeah, he used to be #2 back in the early days.

 

[Kink-Link5]

Huh; well goes to show how much influence early tournament leaders and the evolving metagame can have.

 

[Zankoku]

You question the past, yet based on your join date surely you lived it?

There were many who did not see Snake as #2 in terms of character capability, but he was definitely #2 in terms of successful representation back then.

 

[Kink-Link5]

I didn't pay much attention to Brawl results, and only recently looked into the game's matchups and stuff. I played the game for about a month before realizing that Olimar's design was the most toxic thing I've seen in a video game.

But then I met 2.5 Sonic.

 

[PMS | LEVEL 100 MAGIKARP]

I find it kind of hard to believe that the Brawl list is very accurate when it lists Olimar at second and Ice Climbers as low as fourth.

All I want is to know the totals for a character being entered and the number of tournaments being used. You don't need to recalculate or run anything more through, just show your data points.

Except Olimar is second. No doubt.

 

[Kink-Link5]

I'm not really here to discuss the Brawl tier list.

That said, losing to Ice Climbers and going even with Marth and Falco doesn't really scream second best character to me. Third best sure. Ice Climbers have even or advantageous matchups across the board.

Also the matchup chart exaggerates matchups quite a bit. IC's are certainly not advantaged over Pikachu to the point that it's a notable matchup shift, and no way in hell does MK go anything but even against most if not all of what is listed as being a +1 (And if he does go advantaged enough for it to make a real difference, then the matchup isn't shown accurately in the chart). MK is still far and away the best character, but take the matchup chart with a grain of salt between -1 and +1 for what is listed as the matchups.

My main problem with this thread has already been stated; there isn't a total sample size or data points shown so we only get to know that of a few tournaments Fox is winning and Bowser is getting a lot of top 4s. Great to look at trends over time, (Just look at the Brawl one when Snake was apparently doing stuff), bad to use as an analyses of "character rankings."

 

[JOE!]

So, if fox is used 500 times but wins 8 major tournaments between 6 players, is he good or bad kink?

 

[Zankoku]

The point from the beginning was to watch trends, but calling a list that ranks characters by points something other than a character rankings list would be somewhat silly.

 

[Overswarm]

Kink, I think you're misunderstanding both the purpose and the effectiveness of this endeavor. To be frank, the list is smarter than you and everyone else on smashboards because it is accurately representing what is actually occurring rather than what we believe should occur. There will be minute errors (such as people picking a secondary they don't need, people sandbagging in tournament and dropping out earlier than they normally would, etc.), but over time as more tournaments are added they will be insignificant.

You ask for sample sizes and data points, but in truth the sample size is all tournaments. Whether or not they are added is due only to Metroid's thread and if they meet my criteria.

As good as you may think character X or Y should be, this list merely says who is the most successful. You can say "Olimar loses to Marth, Falco, and ICs, how can he be #2" but the list would simply say "he is". Snake was #2 for a significant amount of time because Snake was doing very well.

While you may worry about something like, say, Fox being over-represented and thus getting more points than he should, you will find that Fox's points are going to be linked only to successful tournament entrants in the first place. If there are 50 Foxes in the month of January and only 5 Bowsers, the 5 Bowsers could easily get more points than the 50 Foxes.

This is due to the fact that characters used as a secondary (common with Fox) receive only half the total points and money earned and that your points is calculated by multiplying your base value (1st for top 8, 4 for top 4, 7 for 2nd place, 10 for 1st place) by the entry fee and amount of entrants, then divided by 160.

This means that in a 20 person, $5 entry fee tournament, if Bowser got first he would receive 6.25 points. If Fox was "over represented" and appeared heavily in the results, we need only really be concerned if he was filling up the bottom spots. If Fox consistently got 1st, 2nd, 3rd, and 4th, his high placement surely would be deserved. Regardless...

1st Bowser - 6.25 points
2nd - 4.375
3rd/4th - 2.5 (5 total)
5th/6th/7th/8th -0.625 (2.5 total)

Fox taking up 5th through 8th would result only in a total of 2.5 points for Fox assuming he was a solo main for each. If he also got 3rd and 4th, he would have 7.5 points (1.25 more than Bowser).

As you can see, over representation isn't an issue. Because the point system is weighted, it is impossible for a character that isn't good or successful to overtake those that are actually successful over a long period of time. The only exceptions to this are:

1) Tournaments with a ridiculous amount of entrants
2) Tournaments with a ridiculous entry fee
3) Tournaments at the very beginning of a measurement

Now over representation of a character can be possible on the flip side: one player doing extremely well with a character on his own. During the Brawl list, Meta Knight and Snake were given so many points by Mew2King and Ally that they themselves would have been ranked A tier characters alone!

But this is only an issue if you consider this list to be a tier list, which it is not. It is merely a measurement of success. That said, I will be noting oddities I come across.

 

[Kink-Link5]

What does times used represent then because all it says is "Times used" and that apparently doesn't mean "The number of times this character has entered a tournament."

 

[Overswarm]

What does times used represent then because all it says is "Times used" and that apparently doesn't mean "The number of times this character has entered a tournament."

Times it has appeared in the top 8, either as a primary, secondary, or tertiary. All the data being shown is referring to the top 8. That is, the important placements we care about.

Times used does not inflate points to any reasonable degree. One strong player with character A will give more points to their character than 3 mediocre players any day of the week. All "times used" is meant to show is popularity. This can make it more likely that good players will play that character, but it doesn't meant they will necessarily win.

Think of it as "matchups I should know", in order of appearance.

 

[Kink-Link5]

Wouldn't a show of popularity be, you know, showing all the times a character is entered?

You don't have to do **** with that data, just show it please.

Don't run it with the ratio of how much they win

don't do it with how much money they get

Just please I am begging you, and getting obscenely frustrated that you obstinantly refuse to, compile the data for total tournament entrants of a character. Because if Pikachu isn't being used at all that's one thing. If Pikachu is being used but just not making it to top 8 that is another.

But we don't get to know if Pikachu is getting used, because it isn't conveyed to your audience.

I mean **** at least let me know WHY you absolutely refuse to show that peice of information in a thread made as a comprehensive reference of information.

 

[Oracle]

If some randoms are entering with pikachu and losing, then I don't think that's really valuble tournament data.

 

[Kink-Link5]

There's a big difference between randoms at least using a character and that character getting literally no use at all. Of course it's vuable data to know if a character's preformance is due to just bad players, no players, or being a bad character.

 

[Overswarm]

I have updated the first and second post significantly! If you have any questions, let me know.

Wouldn't a show of popularity be, you know, showing all the times a character is entered?

You don't have to do **** with that data, just show it please.

Don't run it with the ratio of how much they win

don't do it with how much money they get

Just please I am begging you, and getting obscenely frustrated that you obstinantly refuse to, compile the data for total tournament entrants of a character. Because if Pikachu isn't being used at all that's one thing. If Pikachu is being used but just not making it to top 8 that is another.

But we don't get to know if Pikachu is getting used, because it isn't conveyed to your audience.

I mean **** at least let me know WHY you absolutely refuse to show that peice of information in a thread made as a comprehensive reference of information.

Because no one cares and it is not relevant and it is impossible. All of those things compound together to form a giant ball of "who gives a damn".

I was able to predict the rise and fall of characters with alarming accuracy using that list during the Brawl era without having more than top 8, so having those entrants does literally nothing except satisfy your curiosity. Worse still, it might actually ruin a successful model having extraneous data involved.

No one, and I mean no one, records the characters of everyone at their tournament. Because of this, past top 8 would be a collection of ???'s and only one of the characters players used and oddball guesses that may or may not be correct. This is bad data, difficult and annoying to accomplish, and takes forever.

Finally, no one cares. There is not a single instance in history where someone said "oh, do you have the characters for all the people that went 0-2 in your 32 man bracket" because it has no bearing on us. A random noob playing any character doesn't have any bearing on that character's success nor does he indicate any sort of trend.

Top 8, however, represents actual tournament threats to real tournament players. Top 8 isn't hard to get into and has proven itself to be a reliable source of information when weighted properly. It's also an appropriate range given that top 4 is a common payout scheme, and this gives us exactly 4 more players to bring us to a total of 8.

 

[Kink-Link5]

Is Axe a random irrelevent player?