Feb 222010

(Continued from Part 1)

Part 2: Variety, Viability and Balance

Street Fighter IV is often said to have really good balance between characters. It’s current tier listing shows it isn’t too bad at least as fighting games go – 53 points between top and bottom, 25 characters (2.12 per character). Although HDR’s current tier listing shows the differences there are probably even less – 26.5 points between top and bottom, 16 characters (1.66 per character), however if you remove Dan, SFIV also drops to 1.71. The other point people say is good about SFIV is that it has mostly 6-4 type of matchups rather than really bad ones. However what people fail to notice a lot of the time is the way in which SFIV has achieved a lot of this – by sacrificing variety.

By it’s nature a game where all characters are the same is balanced, and in SFIV the characters have become more homogenized than usual. Firstly the more common subsystems like Focus Attack you add, the more similar the characters will become, but SFIV goes further than this – almost all characters have very similar jumps now, and almost identical throws, with identical ranges (outside of specials). They all have almost identical dashes as well, and guess what, all the characters can still block just the same. So when you emphasise these abilities that all characters can do the same (more on this in Part 3), naturally the characters become more balanced, and that’s why you can feel you always have a chance win, even in a ‘bad’ character matchup (and why it’s only 3-7 on a matchup chart).

But despite this similarity between characters, SFIV still has a another problem with balance. It has too many clearly awful characters – for example Claw, Guile or Dan, who there is really no good reason to use at any level of play, other than for a handicap. At least every weak character in HDR has some reason to play them – either a good matchup, or a unique playstyle not offered elsewhere. In the top levels of play at HDR, almost all the characters do get used for one reason or another. Yet in SFIV it’s a different story, take a look at this. A staggering 70% of the top players use ‘shoto’ based characters (characters derived from Ryu & Ken). Also notable is that the only ‘charge’ character to survive at the top level is Boxer. For me this is the same problem as a game like Marvel Vs Capcom 2. A game might appear to offer lots of variety in terms of number of characters, but if masses of those characters are just at a disadvantage almost all of the time against the clearcut “upper tier” characters, then it’s actually got very low viable variety.

For example if Blozzard added a 4th race to Starcraft that was just like Terrans but all their units don’t do quite as much damage, have I really added any viable variety to the game? Similarly is Dan really adding any variety in Street Fighter IV? Whilst he’s a hilarious diversion and a fun handicap mode (he’s actually my 2nd most played character in Ranked Match in SFIV), in design space terms he’s quite literally a waste of precious resources.

Note that this debate often goes along in Super Turbo vs HD Remix comparisons too. Super Turbo preferring players argue that HDR has “less variety” because it has made the characters slightly more similar in play style. For example Dictator is slightly less about ‘all offense’ and Fei Long is far less about a ‘single lockdown corner trap’, and Old Sagat (being all about ‘full distance zone’) is gone completely! And I’d agree with them on this point. However to me, by making far more characters viable in tournament play, HDR has increased viable variety, which is a far more important concept to me. What viable variety means is a delicate confluence between variety and balance: ie: you are aiming for both! And this is an incredibly hard thing to do. I believe it’s practically impossible in a single iteration of a game, which is why games like Starcraft took many patches to get right, and a game like HDR needed to be made based solely on an older game. Sadly Super Street Fighter IV looks very unlikely to improve things as much as it will just change things around – reading many interviews with developers, despite the odd bit of lip service, it’s just clearly not a design goal of theirs to attempt to have balance and certainly not viable variety. If it happens, it appears it will be by pure accident.

Here’s my viewpoint in a ranking style between some different current fighting games:

Pure Variety (differences in and number of playstyles): 1. MvC2, 2. ST, 3. HDR, 4. SFIV

Pure Balance (closeness of tiers): 1. HDR, 2. SFIV, 3. ST, 4. MvC2

Viable Variety (aka: Both): 1. HDR, 2. ST, 3. SFIV, 4. MvC2

To summarise, I’m explaining here how SFIV manages to have relatively low variety, and relatively low balance at the same time.

To be Continued in Part 3

  11 Responses to “Fourts and Chaos Part 2: Street Fighter IV and Variety, Viability and Balance”

  1. I used to think Marvel was a terrible game for balance, but when you think about it’s actually pretty balanced. I mean, a pretty small fraction of the characters are viable on a high level, 12ish out of 56, but having 12 viable characters is still pretty good. HDRemix has 16 viable characters (i.e all of them). That and you have to pick three, so each different combination of characters is almost like a different character in and of itself with the assists and such. Mind you not every combination of characters and assists is equal, but there are still a good amount of viable combinations. There are also a million ways you can go about actually conducting the fight too, so there’s depth there too. The game is incredibly complex, it’s just most people don’t get to see it because of the high (think Mt. Everest) execution barrier. I’m TERRIBLE at the game, but I can see by watching videos of the top players that there’s a lot going on in the game, all the different factors you have to think about when playing it.

  2. Thanks for the great comment Jeff. I like to hear counter-points and I’m definitely not a high enough level at MvC2 to completely judge, this is just the way I see it from my perspective – I could have my opinion changed in future of course. 🙂

    I feel that the choice of 3 characters actually reduces the viable variety as there are a few characters that show up in virtually every viable team (I’ve not looked at the game ‘seriously’ in years, but I recall Sentinel was one), so every team just felt like subtle variations on a theme to me. And like you say, no matter what you pick there’s insane execution requirements – so, well, see part 1 for why it’s really not the game for me to be able to play at a high level!

  3. The god tier characters are Sentinel, Storm, Magneto, and Cable. From what I can tell these characters are all the best because:

    Sentinel: Controls a huge amount of space with lasers and limbs, super armor, DHC (delayed hypercombo) with Storm that kills most characters, two amazing assists

    Storm: Best runaway in the game, builds meter FAST, decent rushdown, amazing supers, DHC with Sentinel, has an infinite on Sentinel, two great assists.

    Magneto: Best offense in the game, with mixups, resets, trijumping, infinites, good projectile assist,

    Cable: Best super in the game with Hyper viper, uses up meter well, good anti air assist, seems to be the best zoner, in terms of how much space he controls and how much damage he can give off if they get caught in it.

    In terms of overall contribution to the team these characters are unquestionably the best, but there are other characters and teams that are used. For example, Captain Commando, Cyclops, Dr. Doom, Psylocke, Cammy, Tron Bonne are all characters that are used solely because of their amazing assists. There are also other lower tiered characters that are viable with higher tiered characters to help them, such as Spiral, Blackheart, Ironman (though apparently some argue he’s pretty close to god tier), and apparently even Rogue. Plus there are the low tier teams like Juggernaut/Tron/Doom that people like Mike Z bust out, though to be honest you have to be amazingly creative and skilled to pull wins with a team like that.

    Basically what I’m saying is that yes the god tiers are the best, but there are plenty of viable combinations between them and the other characters that aren’t quite as good/have good assists.

    Also, you said that the different teams were only minor variations on each other. I’m not disagreeing (this is mostly to generate discussion), but how is that different than SOME characters in HDR? Cammy and Fei Long seem similar to me. They both are pressure/mixup characters. Fei Long has his high/low with chicken wing and crouching fierce, while Cammy has her poke into cannon drill pressure and mixes it up with hooligan throws. They both have dragon punches too. You could also argue Balrog is similar, in that he’s a pressure/mixup character? Or what about Dee Jay and Guile (since I know you’re familiar with Dee Jay; as am I)? They’ve got similar move sets, and can both zone and be offensive depending on the matchup (though to be honest Dee Jay seems more suited for offense and Guile defense, but I’ve seen both go both ways).

    Not hating on HDR, btw. I LOVE that game.

  4. 🙂 Don’t worry, I’m more than happy for people to disagree with me on anything, in fact, I’d encourage it, as long as they have good reasoning as to why!

    I have to admit I’m not fully qualified to have a discussion on this level about MvC2, it’s really SFIV and HDR that I know a decent amount of current information about, and have played a good amount of recently.

    That said I think part of my problem is that I internally emphasise the “all characters must be useful” thing. I look at % of playable characters as an important stat, because, as I’m pointing out with SFIV – the other characters are actually a waste of design space and resources; I’d rather they spent the time balancing just 12 characters than doing an extra 44 useless ones. But I can also see your point that it’s the raw number of viable characters that can matter as well. I also am historically aggrieved since all the characters I used to play on older Vs games, are actually all near useless in MvC2 (although some of them were probably always semi-useless), so it hurts to have to ‘lose’ your favourite characters and playstyles (in psychology terms – deprival reaction syndrome). I also don’t think that a character just being “viable as an assist or combination” means enough. You’re still playing as Team Sentinel.

    In terms of playstyles, yes some characters in HDR are indeed similar, I would rank it pretty low in terms of pure variety for this reason. This is all a matter of opinion though really; it’s hard to define what constitutes a truly ‘different’ playstyle in a fighting game, especially when comparing very different games. But in more similar games (SFIV, HDR, ST), it gets easier to define, and once you eliminate all the non-viable or pointless (as they are usually 4-6 vs all the upper tier) characters I find that HDR offers a lot more interesting variations than SFIV. It isn’t all shotos!

    Aside: As I’ve mained both, Guile and DeeJay actually play very differently at times in certain matchups in HDR! If they were that similar I would’ve had no reason to switch. Whilst their projectiles are pretty close, Flash Kick and JackKnife Kick are very different moves … and as ever with ST/HDR its really all about the normals, which have very different properties and uses.

  5. Another good way I find to look at a tier/matchup charts is this: What reasons (ie: characters) are there to NOT play the top character, since everything else stems from that. In HDR, Ryu and Boxer are currently seen as joint top, yet 4 other characters are seen as having an advantage vs Ryu, and 2 other characters are seen as having an advantage vs Boxer. So suddenly you’ve got 7 viable characters (not double counting Boxer). Then you’ve got characters that can beat the characters that beat the top tier… and so, you see why almost every character at least has a “niche” in HDR.

    In SFIV, no-one beats Sagat. So instantly you see why many characters are really quite pointless.

    I guess MvC2 would be similar that no-one beats the “god tier” but of course the team based nature makes it much harder to analyse on a pure matchup chart level. Heh. I probably shouldn’t have put MvC2 anywhere near this discussion really.. but there it is now, so feel free to tear it to pieces 😉

  6. I also ought to point out, that this is far from being my biggest problem with SFIV. I’m going to go in many more points in this series about SFIV as you can imagine, and these aren’t intended to be ranked in any order of importance – just things I feel are worth mentioning. I felt this one was especially relevant by comparison. When earlier titles in a series do something better, it’s really a shame. But as is obvious, they really don’t care too much about this stuff – they’re perfectly happy making an unbalanced game. Nor do they see it really as a series, they just reboot and start again.

  7. I agree with you that Street Fighter IV isn’t a good game. Not very much variety, hence the ‘balance’. It really feels to me like they were afraid of giving anyone any real powerful stuff (except Sagat. WTF were they thinking…) so while it’s true no one really soars, no one’s fun to play either. Sure it’s ‘balanced’ but it’s boring and shallow.

    I agree with you about Dee Jay and Guile, I know they’re acutally really different characters. Hence why I’m a decent Dee Jay (won’t say I’m good since I’ve only played online) and a terrible Guile. I was just generating discussion.

    Also, MvC2 was a fluke that it turned out the way it did. The game is played in a far different way than the designers probably intended. The forty odd characters that aren’t viable probably weren’t designed to be worthless (except maybe Roll and Servbot), that’s just the way the game evolved. A game like MvC2 that’s sort of balanced and very deep and complex through it’s brokeness won’t happen again (probably).

    Your comment about matchups is also very true. For example, in Blazblue Calamity Trigger if you look at the matchup chart Rachel is technically the best character, because she has better win ratios against more characters than Nu-13, the real best character in the game. Nu is actually better because Nu’s matchup against Rachel is HEAVILY in Nu’s favor in addition to having even to favorable matchups with everyone else.

    ST Honda’s another character sort of like that. He had quite favorable matchups against the non fireball characters, but had a real tough time against the ones with fireballs. Since the fireball characters were generally popular, that made Honda less viable and the nonfireball characters more viable with the absence of Honda.

  8. SFIV is very deep and ‘good’ if you like doing difficult combos! 😉 eg:

    A thought to ponder – “ST (and HDR) would be a better game without Honda”.

    I don’t completely agree, but his design is unfortunately divisive in the manner you’ve described.

  9. I still don’t think SFIV would be too deep even if the difficult combos were easier/didn’t exist. The game would still be slow and reward defense too much, I think. Plus you’d still have the awful character variety.

    I honestly am not sure whether I agree with your statement about Honda. On one hand I hate banning stuff, and only favor it in the most extreme circumstances (Akuma in ST/HDR, for example). Honda is not in anyway overpowered (this is coming from a Cammy player lol). Sure he shuts down like half the cast, but he gets shut down by a lot of popular characters. His very existence though really discourages people from playing characters that have bad matchups with him. You COULD say that it’d increase variety, but another anti no fireball character could rise and take his place, so I dunno.

    I’d be interested to see HDR/ST without him, and see what other characters would rise in popularity because of it. Maybe some non fireball character would get real popular and make more people play fireball characters to try to counter it. It’d be too complicated to predict.

  10. Another great data point about the real lack of balance in SF4. The table really speaks for itself:


Leave a Reply