Can we talk about the structure of tiers for a second?


#1

I think that the way tiers are quantified is a brilliant method. Setting the matchups, establishing the odds and compounding it at the end.
However, noob that I am, I’m having a little trouble what everyone else sees.

If I understand it correctly, tier matchups are done in an “out of 10” manner, so that a 5-5 matchup means that, if two players of equal skill fought each other with the matched up characters, it is reasonable to expect that one would win five and the other five due to the nature of the characteristics of each character. Correct?

Once again, the model is very very nice. I’m just having a little trouble seeing it applied to real life. Tiermakers are hesitant to attribute 7-3 matches, and 8-2 are extraordinarily rare. Many matchups are even quanitifed with decimals, liek 5.5-4.5. My beef is that in real life, some matches seem to be far, far more unequal than the tiers would make.

When you look at tournaments there seems to be a lot of 7-3 matchups out there. A few times, it seems like picking one character almost automatically means a loss for the other.

The EventHubs tier says that Hakan and Dhalsim are 5-5. From what Infiltration did to Filipino Champ ( one of the best Dhalsims in the world), it doesn’t feel that way.

Zangief vs. Sagat is 7-3. Meaning that two players of equal skill fighting each other with these characters would result in Zangief winning 3 out of 10 matches. But would he really? That feels extremely unlikely. As a matter of fact, when it happens in a tournament it’s pretty much night night Gief.

I know there are many knowledgeable guys making these tiers, I am merely stressing the point that, between top notch players, these discrepancy between uneven matchups is far greater than what you’d find in a tier.

Dan is supposed to be the worst character in the game but his tier is mostly 4-6 against everyone, except the obvious massacres ( Dhalsim, Zangief, Seth, etc.).

Honestly, if his matchups were really that close ( winning 4 out of 10) we’d see a lot more competitive Dans out there.

So perhaps you guys could enlighten me on this discrepancy, show me what I’m not seeing here, because I’m legitimately curious about this.


#2

I’m not the most knowledgeable dude out there, but if I understand well, the tiers (mostly the matchups) are a way to determine which character is the best without any external variable. If I’m not wrong, it is mostly calculated with thing like range and options against the opposing character.

However, tiers does not determine the winner of a match opposing 2 humans, because there is external variable that change the value of these numbers. Some are game-related like your combo skills, you knowledge about set-ups etc. And some are not directly linked about the game like your ability to play under tournament pressure, your ability to bait and your opponent ability to bait, etc.

So, tiers only affect game when the 2 players are about the same level of skills, not only technical, but also their knowledge about the matchup and about the opposing player. One of the greatest assets of Infiltration is his knowledge of his opponents. Because of this, his overall level of skill gets higher than most of his opponents, meaning that the theoretical matchup loses precision.


#3

This is true, and it is calls for reflection, but someone recently did a chart with the most popular/victorious characters in tournaments in 2012 and it deviates considerably from what the tiers would lead one to expect. Seth, in particular, almost never appears, whereas other B and C characters often do.


#4

I think this could have been discussed in the tier list thread. I will post here some things I wrote in the past: "usually tier lists are based around match ups (and our tier list follow this patern). We take every character, match him against the others, and try to decide how many matches he would win in a 10 games set, considering the characters are played at the highest level by 2 equaly skilled players and then asign a number (a 6-4 match means a character wins 6 games out of 10).
But how do we do that ? well, this is how LoyalSol (one of our contributors) explained: “Winning match up is based off how many options you shut down versus how many options your opponent can shut down. The more options you have the better the match up is for you and the fewer options the worse the match up is. Horrible match ups are the ones where even your best options gets beat.”

Now, regarding your questions:

  • when a tier list is done, we match 2 equally skilled players. Problem is, in practice, very seldom 2 opponents have the same exact skill level, so the matches might look more loopsided not because the match is inherently worse, but because one player is simply better than the other. It’s impotant to keep in mind that when someone loses, it’s not necesarly because his character has a bad match.
    Actually, I believe tiers have a less important role than many people asume. In my oppinion, the reasons why someone loses are:
  1. his opponent is better than him - the most important reason, and the one that decide the outcome in msot of the matches
  2. he doesn’t know the match that well (or at least worse than his opponent) - an important factor as well
  3. his character has a bad match.
    Tier lists matter only when a. opponents have similar skill level, b. both opponents know the match up, c. the match up is not really loopsided (7-3 or worse).
    Basically what I want to say is - when you see someone winning, you have to ask yourself why does he win, and not simply asume his character has good matches - for example, Justin Wong blows a lot of people, but that might have to do more with Justin being very talented than with rufus having a lot of loopsided matches.
  • people are hesitant to call big numbers because SF4 is a very balanced game. The problem is, many games that preceded this one were way more unbalanced - for example, ST has a lot of 7-3 and 8-2 (and maybe worse) - and since SF4 is considered more balanced, people think that matches should be called that worse. Ex: in ST, Honda murdered T Hawk and Gief, but was murdered by Guile and Ryu. By comparisson, in SF4 the matches are not that loopsided, so if in ST Honda vs Hawk was 8-2, in AE people call it better.

  • Another thing to consider match up numbers are constantly changing. Eventhubs tier list was made a year ago, rougly a few months after v.2012 patch came. The knowledge about Hakan is very limited also. UltraDavid also called the match 5-5 back then, but now he changed his oppinion.
    And again - don’t base your oppinion solely on one match - yes, Infiltration destroyed F Champ, but check out those 3 reasons I gave above: maybe infiltration is a better player than F Champ. Maybe F Champ doesn’t know Hakan match that well (most likely - how many good hakans are in Usa ?). These 2 factors might have contributed in the outcome of the match more than we think.


#5

The question is do tiers matter at a low/mid level? I’m not a high level player by any means but my best/worst matchups as Seth are precisely the same to what the matchup numbers refers to for Seth.


#6

Match-up numbers are or have some subjective background, it is hard to come to a agreement in the majority of the match-ups and the knowledge of certain characters is very low.However keep in mind, every player has his or her own match-up char and this is also region depending. For example Sagat vs Zangief is easy to classify, but i guess Zangief vs Hakan is harder to classify.

The tier list was created before Evo2012 and SCR13 happened, back then nobody knew, how this match up is:



Now things have changed(http://www.youtube.com/watch?v=8JM-cleOi64), btw Filipino Champ has played some matches against Hakan players, however according to some people , online matches(Fchamp had some) don´t count. It is funny to see that some people still say that he has only won, because Infiltration is the better player or that Fchamp didn´t had match-up knowledge, even at an advantageous match-up for Hakan or the fact that Hakan doesn´t lose match-ups very hard like Gief does.

Edit: Amazing is that despite the fact that Infiltration doesn´t know much about Hakan´s advanced techs, he still can easily defeat opponents with Hakan, maybe he hides those techs?


I don´t know, but if the Sagat player doesn´t do any mistakes, then he shouldn´t lose those matches. Gief can only win, if the Sagat players screws up(depends on the mistakes), but even then Sagat has good chances.


Edit:http://www.youtube.com/watch?v=Vv6l3W_LO-Q
Just look around the 4:50 mark in the second video, but to be honest Gief is and was never a reliable tournament winning character and then the Gief mainers or fans complain about so called “bullshit” match-up against Sagat´s(But mauling Yang´s is Ok for them) ,that they can´t win tournaments with him.


#7

Tiers only truly matter at the highest level of competition in my opinion. The lower down the “skill bracket” you travel, the more likely your opponent is to make mistakes you can capialize on. A lot of matches that would generally be considered “bad” become a lot more winnable when you can easily block and punish ill advised Dragon Punches and the like.


#8

But now something very interesting comes up:

Knowing a matchup well gives you an advantage. If that is correct, then the fact that some characters are used less would provide for an advantage to offset the technical handicap at least to some extent, right?

Once again, if the game was as balanced as the tiers indicate, then we should expect better results from lower ranked characters because nobody plays them and thus there is little information about their potential. Once again going to Dan ( as I play Dan), if most of his matchups are 6-4 and yet is one of the least used characters in the game, the lack of knowledge in the matchup should allow for the top Dans to excel… and yet we don’t see it. You know what I mean?

But what we see is dominance by characters which are very well known and studied throughout. This would indicate that even with the element of surprise, some matchups are a bit more skewed than the tiers would indicate.

However, we also have to point out that some characters DID break out of obscurity as their potential was revealed over time. Hakan recently comes to mind, I’m sure you can think of others. But this is an exception, not a rule…

My whole point is that* perhaps* the tiers we see here are a tad too conservative, that’s all. Given what we see among the top players, I feel that 7-3 and 8-2 should be put around a bit more liberally. Actually, 9-1 also.

Between to players of equal skill and equal knowledge, is it really too much to expect Sagat to beat Zangief 8 or 9 times out of 10?


#9

What I have a gripe with is how the numbers are added.

Being 6-5 against Cammy should add much more value than being 6-5 against, say, Oni. Some characters are inflated by this (like Seth) and make them seem higher than they really are.


#10

I think numbers are more likely to be on the conserative side, precisely because there are so many other factors besides character strength that can contribute to a win or loss.


#11

guys tiers are subjective and will change a lot more as the game ages. conservative numbers are ok for a game thats only one year old.

in another year maybe gief and sagat players will be more comfortable calling it worse than 7-3.

the numbers available now are a starting point for future discussions