Mathematically Calculated Tier List. SECOND RESULTS IN! First post updated.

[Nubbin]

Not buying it...

I'm not terribly impressed, sure it's been a lot of work, but what does it all mean? By reducing all the factors of the characters to a single number, and then ranking them, the characters lose all their depth.

When you reduce all the factors of a character to a single number, the number becomes inaccurate. If say, Doc (hypothetically) had the best ranking in this system, does that also mean that Doc should have the best recovery as well? What about his juggle game, or edge-guarding? Following the chart, one would assume that Doc is the best in all of these categories, even if the reality is very different.

I feel that the only way to make it remotely accurate would be to have many different rankings, in a variety of categories and settings, such as a separate ranking for a best 2/3 match, a best 3/5 match, and a best 4/7 match, then in each category have separate rankings for aspects such as: Ground game (a separate ranking sub group for each type of opponent, heavy,light, floaty, fast fallers, etc), air game (again, a separate ranking sub group for each type of opponent, heavy,light, floaty, fast fallers, etc), edge-guarding (a separate sub-group for guarding various types of adversaries recovering from various positions on the field. below the stage, left, right, etc, and light, heavy, floaty, fast faller, etc for each of the positions as well.), Recovery (a sub-category for each of different recovery scenarios), as well as other rankings for other aspects of characters which have escaped me at this time.

Well, that's my two cents anyway. Bravo on the efforts so far regardless.

 

[Mogwai]

I'm not terribly impressed, sure it's been a lot of work, but what does it all mean? By reducing all the factors of the characters to a single number, and then ranking them, the characters lose all their depth.

When you reduce all the factors of a character to a single number, the number becomes inaccurate. If say, Doc (hypothetically) had the best ranking in this system, does that also mean that Doc should have the best recovery as well? What about his juggle game, or edge-guarding? Following the chart, one would assume that Doc is the best in all of these categories, even if the reality is very different.

I feel that the only way to make it remotely accurate would be to have many different rankings, in a variety of categories and settings, such as a separate ranking for a best 2/3 match, a best 3/5 match, and a best 4/7 match, then in each category have separate rankings for aspects such as: Ground game (a separate ranking sub group for each type of opponent, heavy,light, floaty, fast fallers, etc), air game (again, a separate ranking sub group for each type of opponent, heavy,light, floaty, fast fallers, etc), edge-guarding (a separate sub-group for guarding various types of adversaries recovering from various positions on the field. below the stage, left, right, etc, and light, heavy, floaty, fast faller, etc for each of the positions as well.), Recovery (a sub-category for each of different recovery scenarios), as well as other rankings for other aspects of characters which have escaped me at this time.

Well, that's my two cents anyway. Bravo on the efforts so far regardless.

Feel free to do that if you want, but that's not the point of this. Teirs are specifically refering to how viable a character is in tournament, not who's "the best". To some, these mean the same thing, but as you pointed out, "the best" can mean a lot of different things. Assuming that a matchup chart such as Phanna's can be accurate, I've assembled this system to determine ultimately how viable every character should be. This is a making a pretty big assumption that a lot of people will never agree with, so I'm not hoping to win everyone over, just to give a truly objective, mathematical approach to how one would calculate tiers. The fact that Falco currently comes out on top doesn't suggest he's the strongest, fastest, has the best recovery, looks the coolest, and has the best taunt. . .

Rather, these results specifically say, that with the current Phanna chart as a guideline for how matchups should go, Falco will ultimately have the best expected matchup against a field of purely logical players who select their characters in a very specific way. Of course, this is not a terribly realistic model, but the results still carry meaning and the model has been tuned and will continue to be tuned to immitate reality as best we can.

 

[pchoo]

That's not the point of the project at all. The point is to establish a mathematically based tier list for winning matches... not a mathematically based all inclusive every-single-imaginable-attribute ranking.

For example, when you look at the normal tier list, where Fox is at the top, do you then assume that Fox has the best recovery? No. How 'bout his juggle game or edge-guarding? No. You would most definitely not assume that Fox is the best in all of those categories since you know that reality is very different.

The tier list doesn't need to include every aspect of the game and will still be very accurate at what it is trying to do. Don't be silly :S.


edit.
Or what wesley said. Since he got to it first :o.

 

[Nafix]

Just read the first post Nubbin.

 

[im2emo4myshrt]

If you take what Shai Hulud did with the three separate tiers and then took like an average of that. You could do it by giving a point value for the position a character has in each tier than adding the point values from all three tiers together and creating one final tier with who ever having the most points as the top and the least at the bottom. For example, Sheik was first in the first tier and second in the last two so that means sheik would have 26+25+25=76 total points and you do that for every character and make the grand daddy tier.

 

[Mogwai]

If you take what Shai Hulud did with the three separate tiers and then took like an average of that. You could do it by giving a point value for the position a character has in each tier than adding the point values from all three tiers together and creating one final tier with who ever having the most points as the top and the least at the bottom. For example, Sheik was first in the first tier and second in the last two so that means sheik would have 26+25+25=76 total points and you do that for every character and make the grand daddy tier.

I disagree. Most tournaments are best of 3, therefore, I think that data is the only data worth considering. Even using best of 3 right now is shady because of the %s in Phanna chart being so inprecise. . .

 

[Wave⁂]

But in finals it is out of 5.

 

[jaywinner]

But in finals it is out of 5.

But most of your matches will be best of 3. Having the finals best of 5 shouldn't hurt the math much.

 

[Wave⁂]

Yes, but it wouldn't hurt to do that, would it? I'm not saying you should... it'd probably be a waste of time.

 

[squaminator]

kudos! like this a whole lot. help phanna's chart everybody!

 

[Zenjamin]

bump because i think this is a great thread.



also. i just started playing a good mewtoo.

can anyone tell me what makes ganon so good aganst him. i have no experence with this match-up

 

[Pakman]

I am not a big fan of the Phanna chart.

 

[Mogwai]

The phanna chart was too easily influenced and while it was a good idea, I'm not sure if a good version of it could be made. Something like it is necessary for this to be ressurected, but until we get a nice looking matchup chart, there's really no point.

 

[Jihnsius]

Haven't read the whole thread but have a comment on P.
Does the power relate to a character's ability to deal damage and/or knockback factor? If so then I think you should also figure a way to factor in priority and speed. That'd barely be scratching the surface, though. You'd also have to include recovery, edgeguarding ability, tech ability, and combo ability. For example: Tekken's tier list is mathematically set up to include how each move can combo to the next, giving each player a balanced moveset. Including this function in your current algorithm would involve making a ranking system of combo ability for each character based on knockback, power, and trajectory.

 

[Mew2King]

this is so dumb

 

[Ekaru]

this is so dumb

In your opinion, that is. In my eyes, it's interesting. It's ALL opinion.


Anyways, this was a nice read, to say the least. Interesting...

 

[Smashboozer]

Watch out! M2K has connections. He'll get you banned and/or killed, and not neccasarily in that order.
Seriously though, this is pretty cool and I want you to know you have all my noobish support, even though this list makes the all mighty Boozer look bad.

 

[Oskurito]

this is so dumb

I was going to say the same thing

 

[Mogwai]

this is so dumb

I was going to say the same thing

whatever, I tried something back when I actually cared about teirs and matchups and stuff, sue me.

I also still think the theory behind this was solid, but that matchups just aren't concrete enough for it to work.

 

[jaywinner]

I also still think the theory behind this was solid, but that matchups just aren't concrete enough for it to work.

Agreed.

Sadly I see no way to fix this problem.

 

[technomancer]

The way to fix this problem is to go contribute to phanna's thread and make it as tight as possible.

The top 4 tiers should be 5:5 or 5.5:4.5's across. All of those matchups are VERY even at high levels (fox falco 5:5, fox shiek 5.5:4.5, fox marth 5:5, Marth falco 5.5:4.5 or 6:4, Marth Sheik 4.5:5.5 or 5:5, Sheik Falco 5:5).

Also this sort of mathematical thing encourages the thinking that the best characters always come out on top and discourages philosophizing and experimenting with matchups to find ways to win against certain characters and strategies. Back to your gamecube noobs

 

[squaminator]

its a good idea. its just that the phanna chart doesn't go far enough. If we had some more specific numbers (4.75 instead of 4 or 5), then it would be a lot more accurate.

edit: posting after not refreshing the thread for over 3 hours FTL.

 

[Mogwai]

why is this thread even still alive? Stop posting please...

 

[StellaNova]

I don't this can work. The base Power number is derived from what? The Phanna Chart... but where does the Phanna Chart get its numbers from? The whole numbers are not specific enough for idk where you would get a rating for a character. Besides, it is all about who is using the character, not the "power"rating of the character input into a formula. I am all for being mathematical with things, but I don't see how you can get a base rating for a character that determines how he will perform, without taking into the equation the human player controlling the power.

 

[BigRick]

I don't this can work. The base Power number is derived from what? The Phanna Chart... but where does the Phanna Chart get its numbers from? The whole numbers are not specific enough for idk where you would get a rating for a character. Besides, it is all about who is using the character, not the "power"rating of the character input into a formula. I am all for being mathematical with things, but I don't see how you can get a base rating for a character that determines how he will perform, without taking into the equation the human player controlling the power.

GTFO dude.

Next time, try reading b4 posting stuff.

I know I'm being rude, but I think u deserve it.

 

[Brookman]

M: Matchup vs. a given character (Taken from Phanna chart)

I was pretty excited to see this post from you, only to see that.

Phanna's match up chart is mistaken in a lot of match ups (unless it's had a serious overhaul in recent weeks; which I doubt), but I guess that it's a good start.

 

[Zankoku]

I have the entire formula written out and ready for execution on an Excel spreadsheet, but I too lack faith in the Phanna Chart. If anyone's willing to provide a complete chart on matchups in terms of win percentage, I'd be happy to input it and give the tier results.

 

[Brookman]

"I lack faith in the phanna chart" is a good way to put it. It's like a religion for smash noobs.

 

[Mogwai]

"I lack faith in the phanna chart" is a good way to put it. It's like a religion for smash noobs.

Totally true.

I still sort like the phanna chart as a jumping point for discussion, but the extent to which some people believe in it is really ridiculous.

 

[Brookman]

We could definitely make an accurate match up chart though, and I think it would be a good idea too. So long as it's accurate, and people who know what they are talking about contribute to the list instead of randoms from across the universe.

 

[squaminator]

lets get that jason s person to contribute to it i heard he four stocked cort.

 

[Micheloxx]

lol, i bet that u dont do the same effort in mathematics, jaja..
no seriusly, pretty nice investigation man..
didnt even think about that formulas..

 

[no. 1 smash daddy]

hey, nice chart dude, a better way to show who has more chances to win

 

[StellaNova]

K. BigRick. Go ahead and worship the chart and believe its the only outcome. I'm glad I read through the entire thread and still posted what I did, as IT STILL MAKES NO DIFFERENCE as to what character you use. It is the person in control of the character.

 

[Gimpyfish62]

really? is this really still alive?

 

[Pat/Pro]

If phanna made his chart better this would be legit. There is some crazy stuff on the chart. Like most doc players would tell you that shiek and ICs are a harder matchup than Marth. In fact many do very well in the matchup. Also Marth counters falco and Shiek definitely doesnt 7-3 counter marth. Its 6-4 at its peak. Also like people said it would be better if there were decimals and at least half points. Like falco doesnt 6-4 counter fox. Most people think its even but i think its 5.5-4.5 for example. I think that if it was done with a perfect matchup chart it would be more like this.

Marth
Fox
Shiek
Falco
Peach
IC's
Doc
CF
Samus

Even though I feel that shiek is the best character when played perfectly for his ability to ghey and outprioritize fox and falco, his slight counter against marth, and counter against peach samus and cf.

 

[Gimpyfish62]

marth is the best character

 

[takieddine]

Marth's power level is over NINE THOUSAAAND!

 

[NES n00b]

There is worse stuff than that. Look at the Mario matchups: It says that Mario has the advantage over Luigi, is even with Ganondorf, and not only does Mario have the disadvantage against Doc, but also Doc has the same amount of advantage that Falco has against Mario. CF and Ness are completely even with Mario.

Link matchups: It says that Link has a disadvantage against ICs and it is about as bad as a matchup for him as Captain Falcon. Supposedly Samus is as good as CF against Link. Link is even with Peach. Ness is a Link counter and is just as good as CF in this matchup. >_>

Kirby is only at a slight disadvantage against Sheik. Ness beats all low tiers. Young Link counters Peach. . . . . . . .

As you can see, all those matchups that I got off from this chart are pretty rediculous. This matchup chart is filled with errors.

 

[Pat/Pro]

Most of those things you mentioned arent that bad like young link is a peach counter maybe not as bad as it says though and link and peach are pretty even but more in peaches favor i would say. Mario does have a slight disadvantage to doc. ETC ETC. Kirby and shiek is pretty ******** though and maybe a couple more are bad.