GG:AC Official Tier List Thread

[Hellmonkey]

Destin, the Japanese tier list was made up of a bunch of people on a forum, with tons of people giving their input (not necessarily top players). This all happened very early in the game (mostly pre-SBO 07) So yes, those tier lists are outdated. The last time I saw discussion on the BBS about ranking was someone arguing some balance point and being told that it was silly to discuss at this point in the game. I agree, as there is no merit in making or changing a tier this late in this game. People have a good sense of character's general strengths at varying levels of play, but at top level play, which tier lists are meant to describe, it is impossible to make any sort of accurate list because it is so much more dependant on the player's performance than the character they're playing. Results reflect this inaccuracy. Also yes, I tier whore eddie all day. Though when I was in Japan it was WAY easier for me to win with Baiken, tier whoring just doesn't do much for you there against the better players.

[Ryu]

Since then low and mid tiers have consistently been winning in top level play.

Thats not enough matches. If low tier were actually winning more than High tier wouldnt they be high tier. :S

seriously arguing that there are just more players of middle and especially low tiers that cause them to win more often

...YES... 6/23 for high tier character if theres one person per character.

but if a character really has a bunch of 4-6 should be extremely unlikely.

Its only a 6% difference.

For a 32 man tournment.

.6 *.6 *.6 *.6*.6 = .07776

.4 *.4 *.4 *.4*.4 = .01024

Edit: left off a number. Was thinking 16 man when I said 32.

it is impossible to make any sort of accurate list because it is so much more dependant on the player's performance

Tier list doesnt account for player's performance.

and this

Ryu wrote:

That aside there are times when new strategies are found for X character which could change match ups and affect the tier list. These become more and more rare as time progresses. In short, potential growth per character decreases over time thus we are left with a more accurate tier list than we started with. So I dont see why it wouldnt be possible to determine whos the best character

Humans do not always play at their best. Shoots worst day could be on the same day as Lime Key's best. So no it isnt an exact science

[Teyah]

Hasty? Recent? I made that other post in what, February? Since then low and mid tiers have consistently been winning in top level play. How can you keep saying that I'm being biased and considering only a few recent results that don't really matter and then provide nothing to support your criticism?

Yes it is hasty, because you haven't even bothered to look into the statistics of the tournament records of various top players. And of course, I'm providing 'nothing to support my criticism', since you don't understand the relevance of what myself and others have said about the character and matchup statistics so far.

Saying things like "a player of [Ogawa's] caliber in a game where those tiers really played a role at top level play would win much more", fails to account for the fact that Ogawa is generally playing against other top players, and only needs to uphold roughly a 60% ratio of moving beyond the first round of a major tournament to satisfy the level of play of a top-player that plays a high-tier. Ogawa being knocked out in the second round of a tournament such as SBO would have been the most statistically likely event, but even that didn't happen. Yet you still deride him (and in that line of thinking, other high-tier players) for not winning enough in the past year, as if winning is the decisive measuring stick here. A more prudent person would not have made this mistake.

Saying the "level obtained by their most skilled player" is not nearly specific enough. We can analyze characters' strengths separately, but to combine everything in to one nice list is all opinion.

Success and skill of top players of each character is directly observable, recordable, and easily tracked over time. It doesn't matter what you think of this method, because, short of a theoryfighting robot (lol), this is as accurate as it gets. The Japanese certainly find it specific enough to use this method in their tier listings, and that's what counts.

With between 1/3 and 1/2 of all of the players playing the small portion of the cast supposedly A and S tier, are you seriously arguing that there are just more players of middle and especially low tiers that cause them (somehow? you're arguing that there are more top players of them because of a bigger pool of people playing them?) to win more often? Your entire point about high tiers being a minority is backwards.

No comment was made about the number of top players, but considering how 'top level play' is defined by the Japanese, the answer is fairly clear. Yes, 'between 1/3 and 1/2 of all players' is a minority. I sincerely hope this is not indicative of your knowledge in the field of statistics.

At top level play the math behind the tiers fail,

... LOL. And you wonder why your main argument catches so much criticism. If you've got a probability table or control chart (or anything, really) to go along with this unfounded assertion, it'd be a lot more convincing.

[reaVer]

Thats not enough matches. If low tier were actually winning more than High tier wouldnt they be high tier. :S

This is kinda what hellmonkey is implying. Except that this applies the other way around as well, making it more or less even.

Tier list doesnt account for player's performance.

A character's strength is based on multiple players' performance, so yes it does.

Humans do not always play at their best. Shoots worst day could be on the same day as Lime Key's best. So no it isnt an exact science

So I guess it was Ogawa's bad day while Efute and Lime-Key had a good day? What about Sawa? Did he have a bad day when nearly reverse OCVing N-Otoko's team?

[Ryu]

^^Not every character can be high tier.

A character's strength is based on multiple players' performance, so yes it does.

Nope. The tier list is made based on the assumption that every character is played to its max potential thats realistic.

So I guess it was Ogawa's bad day while Efute and Lime-Key had a good day? What about Sawa? Did he have a bad day when nearly reverse OCVing N-Otoko's team?

Thats just a example. Just saying not everyone plays their best all the time. The names dont matter

[Hellmonkey]

post

Feel free to get back to me when you can say something other than "no, you're wrong! provide proof!", when you have provided less than nothing to justify any point you've brought up. For someone who apparently doesn't follow the scene in Japan, you sure have a lot of confidence in arguing about it.

As for math, when around 44% (this SBO) of the players are playing roughly a fourth of the cast, this means there are over two high tier players for every player of other characters. Such a minority.

[reaVer]

^^Not every character can be high tier.

Nope. The tier list is made based on the assumption that every character is played to its max potential thats realistic.

And how do you know what the maximum potential of a character is?

[Teyah]

Feel free to get back to me when you can say something other than "no, you're wrong! provide proof!", when you have provided less than nothing to justify any point you've brought up. For someone who apparently doesn't follow the scene in Japan, you sure have a lot of confidence in arguing about it.

You're the one raising the point that tiers don't exist, not me. The burden of proof is on you if you ever want to prove anything aside from what has historically been accepted by the majority. This is pretty standard fare for any debate.

As for math, when around 44% (this SBO) of the players are playing roughly a fourth of the cast, this means there are over two high tier players for every player of other characters. Such a minority.

The number of high-tiers used in this SBO is 39.6% (38 of 96 players), from the chart of matchups posted in this thread. And uhh yes, this "a fourth of the cast" is a minority in this case, using either of these numbers. Less than 50% = Minority. Do you seriously not understand this basic principle. Well, it would appear so, from your lack of addressing anything actually related to statistical analysis and not just raw numbers.

Please don't bother posting again if all you have to say is something along the lines of "I'm right, you're wrong, I'm not going to explain why or go into any analysis of available data, kthxbye"

[Hellmonkey]

I'm not going to waste more time finding the numbers again but even with 40% there are just under 2 players of top tiers per every other character. Despite being slightly less in numbers overall how can you argue this is any significant minority, they are being represented more than everyone else individually by far and still falling short in major tournaments all over. Yes, I have provided no explanation or analysis of data. Sorry, it's not worth continuing this with your comprehension.

[fatewalker]

quick Q: was the teir list that was originally printed in the May 2008 issue of Tougeki Damashi magazine set up like we have it here or was it done like a match up chart? if it was a match up chart can some on post a scan of it please? thanx.

[Hellmonkey]

The only matchup chart is horribly outdated.

[Teyah]

Yes, I have provided no explanation or analysis of data.

Good. I hope you understand for the next time you want to put forth a (rather major) argument, some explanation is in order. When the opposition manages to refute pretty much every single one of your supporting points, you can't get by on nitpicking on minor details and then calling it a day.

Sorry, it's not worth continuing this with your comprehension.

There is nothing wrong with my comprehension. I read what you post; you don't post much new material (other than to reword your argument several times), so there isn't much for me to draw from.

If I see any further comments like this one, your posts won't remain here, as this is just plain trolling now. Consider this a warning.

[Teyah]

People, please post only on-topic posts, there's no sense in devolving this thread yet again.

[>_<]

IMO chipp and sol have rose in tier rank!

from what I've seen recently chipp is atleast lower mid, and sol a little above B tier. just my Opinion based on what I've seen the past couple months.

[TITANIUM BEAST!!!]

Sol's not going up anytime soon. Pretty much nothing new for him at this point.

[Ryu]

And how do you know what the maximum potential of a character is?

Max potential thats realistic. And, oh, I dont know, use the best players for your data maybe.

Edit:

This was most likely going to be my next post so Ill go ahead and say it. "How come the tier list doent account for skill level? Whats Max potential thats realistic blah blah blah."

Ok lets say you took all the I-no players and wanted to find their matchups. Well you dont want to use just one person for all your data. You will skew the bell curve. For example, if you only used Koichi to find all I-no's matchups you wouldnt find I-no's matchups/teir, you'd find Koichi's I-no matchup/tier. Since Koichi is really really good he make I-no look like S rank because he's at the tip of the bell curve. To get rid of that you want to take the average of all the top I-no players. Then you have, at top level play this is what I-no's matchups are regardless of skill.

[reaVer]

Regardless of the size of your input, you'll always be basing it on player capabilities which results that in a tierlist you do take player skill into account. This is simply something that can't be avoided and this is also why tierlists change.

[Ryu]

Because Im lazy Im going to agree with you. If you take Beginners and make a tier list based on them Potemkin and Slayer are S rank. That kind of tier list is worthless though. Doing it the correct way reduces the margin of error.

[DoomieJ]

Because Im lazy Im going to agree with you. If you take Beginners and make a tier list based on them Potemkin and Slayer are S rank. That kind of tier list is worthless though. Doing it the correct way reduces the margin of error.

IMO, Slayer + pots are stier. i mean, they are no eddie/test, but they are rather good. Sol is the beginners char of choice along with ky though, and beginnners can mash effectively to get reversals and such.

Dunno, i like the new list though. Definatly an improvement.

[MGibson]

Do tiers exist? Of course, by definition no character is equal therefor tiers surely exist.

Do tiers make a difference in game play? Definitely, although near impossible to measure accurately, it's there.

Do tiers play a large role in the outcomes of tournaments? http://dustloop.com/forums/showthread.php?t=4582 I'd put my money on no.

Part of the beauty of GGXX is the plethora of options to overcome character differences.

[FleetingPenguin]

Could someone explain why Chipp and I-no both dropped in tiers, and Anji went up, in the latest tier list? I'm curious.

[AtTheGates]

i'm just speculating, but chipp still had potential back when all the char specific alpha blade combos weren't figured out. unfortunately, his combos are very hard, just look at the stuff samitto pulls off, even he screws that up from time to time. you need to play like him to justify chipp being maybe C+. Samitto is the best, and look how he fares against worse players with better chars :/

[afterdark]

ei is there a recent "tier" list made by jap players?

[JinSaotome]

TGhe first list on page one is from May, and is the most recent

[Destin]

Oh yeah I think horizontal positioning does matter.