Sponsored By

Q&A: Backbone's Sirlin Talks Remixing Street Fighter II

Answering questions on Super Street Fighter II Turbo HD Remix, Backbone lead designer David Sirlin talks in-depth with Gamasutra about rebalancing the classic one-on-one fighter - suggesting that design accessibility is possible without damaging hi

Brandon Sheffield, Contributor

April 16, 2008

15 Min Read

There are many notable questions to be asked on Super Street Fighter II Turbo HD Remix, one of Japanese-headquartered publisher Capcom's most high profile developments right now. The game - which is a hi-def re-imagining of the classic Street Fighter II with both original and remixed gameplay modes - will launch for digital download on Xbox Live Arcade and PlayStation Network in the near future. In this interview, Backbone Entertainment lead designer and competitive SFII player David Sirlin talks with Gamasutra about the design decisions he has implemented in working with Capcom on the game and why - arguing that accessibility is possible without damaging high-level play. [NOTE: This interview goes quite some way into specifics. If you want to know something about more advanced tactics for the franchise to orient yourself, then Sirlin's tutorial videos from the earlier Capcom Classics Collection 2 are a useful starting point. In addition, he has written a series of articles for Capcom.com about tuning individual characters for the upcoming Remix title.] You’re doing a lot of tuning with Super Street Fighter II Turbo HD Remix, and I know you’ve written articles about it, but still: why simplify it? David Sirlin: There’s been a lot of reporting that the game was dumbed down or something, but I think people are off base there. The changes that make the moves easier to do, do not really affect high-level tournament play at all. There’s a couple of cases where the move changes really do affect it, and I’ve changed the properties of the moves to compensate, but in most of the cases, it doesn’t really affect the balance. People maybe think it does, but it’s making no change at the high-level play, and at low-level, it’s giving people a chance to experience more of the game. If you’re stuck at the level of, “I can’t do Cammy’s Hooligan Throw,” then you’re not really playing Cammy and not really playing the game or feeling the strategy. I want to get you past the beginner phase into the intermediate phase, where you get the strategy and the fun. So that’s the idea -- don’t mess up the high-level play, but get the beginners to the intermediate stage faster. I remember when I first played Street Fighter II; there was definitely a bar that you had to pass before you could even figure out what was happening in the game. There was some complexity that kept people out: how can you really communicate this to people? DS: It’s really pretty straightforward actually. With Cammy, say, you give any player Cammy and her new Hooligan Throw command, where she flies through the air and throws you, is a quarter-triple towards, and then punch, like a fireball, they do it all the time. It’s just immediately obvious that they’re able to do all her moves when they couldn’t before. Even I couldn’t before. I didn’t really play Cammy before HD Remix, so that’s one of the reasons I probably steered clear of her, even the experts can’t do the original motion every time. It doesn’t seem to me that lowering the barrier of the move complexity is changing the strategy too much, because really, you’re just using the right move at the right time; it’s all about timing and strategy rather than complexity of input, right? DS: That’s right. There’s still a lot of nuances of timing. One developer I worked with during the project, at one point said, that Cammy’s Dragon Punch move -- call it a thrust kick or cannon spike depending on who you ask -- is too good and has too high a priority. That was the claim. I was playing as Cammy at the time, and he was playing as Fei Long, and later we switched characters so I was playing as Fei Long, and he said that Fei Long’s Flame Kick has too much priority, it’s like his Dragon Punch. The thing is, the priority on these moves is actually about the same; I don’t have the frame stats off the top of my head, but I have a feeling that both of them are invulnerable for a certain amount of time, and both of them become vulnerable later. It’s really about, did you do yours a little bit before mine or a little bit after? It’s all about those nuances in timing, and we’ll deliver there just as much as ever. How do you balance the priority of these moves? Excel spreadsheets? DS: [Laughs] Well, maybe it’s not as you imagined. Of course I have access to all of the hit box data which is tables and tables of numbers, and all the formulas and equations of how the trajectory is computed, so you could say I have access -- they’re not in spreadsheets -- but I have access to all this data and I must use some kind of analysis and some formula to come up with what to do, but it’s really not to do with that. My biggest secret is that, even though I have a math degree from MIT, it’s not about math at all. If I was going to make a fighting game from scratch, starting with nothing, there’d be a tiny bit of math to make sure that, if you make them block a move, you can’t block it again to prevent infinite loops, so there’s already a little bit of maths done for me in Super Street Fighter, but when it comes to things like, should this guy’s priority be a little bit better or not? It’s just this holistic approach. We know the results from all these tournaments from all these years, and it’s just having an intuition of what a tweak is going to do. So for Fei Long, for example, he’s invulnerable for a certain amount of time, and then he’s vulnerable. Some people suggested he should be invulnerable for a greater amount of time. Now, I don’t need a spreadsheet to tell me that’s a good idea or not. I just know that his Dragon Punch is pretty much effective as it is, and that’s not the reason he loses matches. So, when I look at expert players try to play Fei Long and try to be successful with him, it’s just that they cannot get in, can’t get close enough, and once they do, he has a lot of options and a lot of ways to deal damage. There needs to be a trade-off, it needs to be kind of hard, but not as hard as it currently is. So I have to look at things and ask, what can I do to let him get in a little more easily? One example is that he has a flying kick where he flies through the air as one of his special moves. It’s really hard to do before, we made it much easier to do. I made the small version of that -- there’s three versions, small, medium, and large -- I made the small version be able to go through fireballs at the beginning. So that’s one more option he needs to get in. The opponent can counter it by just backing up and sweeping, but he didn’t have any really good options. So it’s a feel, it’s intuition. Have you talked to any of the original team to see what they think? DS: I have not. Most of them do not work at Capcom. I did say hello to [Yoshiki] Okamoto. I spoke to him briefly, but he’s really moved on from Street Fighter and didn’t have anything to say. Have you had the ability to put anything extra in? Such as dashing? DS: What you have to understand is that it’s technically very hard to make a lot of these changes, so a lot of the changes are bound by what we can reasonably do in a certain amount of time. Changes to sizes of hit box, which affects the priority of moves -- they take me a while to do, but I can definitely do them. Anywhere that’s needed, I can put in the time to get it done, without relying on other programmers or anything. Things that have to do with the trajectory of how things fly through the air, I can actually affect those too myself. Changing frame-stats, like how quickly a move starts out or how it recovers at the end -- this is actually a strange thing, but if I want to change it, there’s a certain threshold where if I want to change it too much, everything breaks and I need a programmer, but if I want to change it a fair amount, I can do it myself. But you’re asking about really new moves, and Ryu’s Fake Fireball is an example of a new move -- of course, it uses the graphics of the old move -- but I needed a programmer to implement that because it’s actually pretty complicated assembly code. And we gave Bison a new fake slide, actually, which is similar in that it required a programmer to code that. So, most of the changes I’ve been able to do involve those first things I talked about; starting, recoveries or trajectories. That gives a lot of the old moves -- they feel like new moves. For example Blanka’s Rainbow Roll, where he hops back and then rolls diagonally forward, it was so slow it was completely worthless, and now it’s very good, so it feels like you have a new move in your repertoire even though it’s using the same graphics. It sounds to me like a lot of the changes are in order to combat the fact that for some characters there is no way to get close fast. DS: Fireballs are really powerful in this game -- more so than other Street Fighters -- and I actually think that’s a really good feature, one of the things that’s a characteristic of what makes the Street Fighter II series what it is, so I really don’t want people to think I want to nerf fireballs and have them be worthless. It’s really that they’re too much: they’re a ten out of ten and really they should be an eight out of ten, and they should be dialled-back a little bit. So, yeah, Fei Long was an example where his short Flying Kick can now go through fireballs. Cammy’s Spinning Backfist? That move could go through fireballs in Super Street Fighter II, but not in Super Turbo. But I gave it back that ability to go through fireballs. But it’s kind of slow, though, it’s not really overpowering, so it’s just an option instead of no options. Is there any major thing you would have added if you had the time or the resources? DS: Off the top of my head, it’s hard to answer this, but there were a lot of times where we felt a character needed something to balance them, and if only we could think of a new move. T-Hawk was one of those. But we were able to find solutions with what we had. So, yeah, it would have been nice and maybe easier to add more moves, or to add another character, but that’s technically very hard and was not scheduled for our project. So nothing like a dash, or counters? DS: I guess I never really considered that. I guess I thought it would be too radical of a change. If someone said, hey, make some new Street Fighter, then maybe, yeah, I would consider that, but this is supposed to be based on Super Turbo. How often did you feel, “Oh, I can’t change this because it will break people’s ideas about SFII”? DS: There were a few of these things -- just trying to remember what they were. Balrog’s one of the best characters in the game, and my balancing tricks here don’t work too well on him, because a lot of my tricks are: take a move that is not really used or doesn’t do much, and then really improve it so it feels like you have a new move. So Balrog already is really good and all of his moves are really good, so what am I supposed to do? One solution is to add a new move that is just for fun and doesn’t do a lot, and we spent a lot of time thinking about that and maybe trying to do it. Another solution is to just leave him, because he is what he is. Another solution is to change him around a lot, you could say for the sake of it. But I mean, I could change the commands of his moves, make his headbutt not go through fireballs, make the turn-punch go through fireballs, and the sum total of these changes give you a really different Balrog experience. It might be more fun, but that’s one case where I was just too afraid to do it. He’s just a really dominant character in that game, everybody knows him and has had to fight against him forever. If I really changed him around, people would not understand: Why am I doing that? Why am I making these changes? So that’s a case where I think I’ll back off and go to how the game was originally. One more thing is -- a big issue is the length of time you have to do a reversal attack. So you get knocked down and then the opponent is sticking his move out as you get up, and you want to do a Dragon Punch right when you get up, but you have a really short amount of time to do that Dragon Punch -- I think it’s one frame, I haven't actually looked, but I’m sure it’s 1/60th of a second. So people suggested we increase that time to two or three frames -- double or triple the length of the window. On the one hand I’m in favor of that because it’s sort of artificially difficult. If you look at Dead Or Alive or any of those other games, if you get knocked down, you have a full second to do your rising kick as you wake up, but in Street Fighter, it’s super-hard. But, if I were to make that change, it has all sorts of ripple effects throughout the entire game -- it would change everything, a lot of it would be bad, and even though I would like the game to be easier in that way, it would have required so many counter changes, the whole thing would maybe not be Street Fighter II anymore. It would be a ton of work and risky, so that’s a point where it’s probably best to just be left alone on this project. It sounds small, but the reason it has such an effect so large is because if you made it really easy to do those wake-up attacks, it would shift the whole balance of how good is attacking against a guy who’s knocked down. And in Street Fighter II, the answer is, really, really good. If you can knock a guy down, even if the knock down didn’t give you any damage, you’d still do it because it gives you a positional advantage, and you can play guessing games with a knockdown guy and they’re all in your favor because it’s so hard to reverse. And for better or worse, it’s balanced around that, so if we made this change, it would really improve defense -- but it’s better to have a good offense. The game’s more fun on offense. Characters look a little thicker -- has that affected the hit boxes of the characters? DS: The hit boxes are the same; I’m not changing the hit boxes to match the art. If you perceive that the new art looks kind of different, a lot of that is coming from you playing the widescreen mode, and the widescreen mode zooms in, but it doesn’t change the aspect ratio, and it does mess with your head a little bit. We could actually show you the widescreen mode with the old art, and then you’d see it’s actually not that far off. It’s pretty close. Yes, some characters are more beefy -- Sagat especially, was very skinny in Street Fighter II and is much thicker here. I guess the jury’s still out until we get all of Sagat’s artwork in the game in a final state, but so far it seems okay. It doesn’t really throw me off. Does it still look like moves are connecting in the right spots where they did in the past? DS: It does. I think we’re saved here by, if you analyse what the original game was like, it’s much worse than you think. You think it was pretty good, but there’s all kinds of cases where moves are intersecting pretty far before they hit, or even more cases the other way around, where you think they’re not even close to hitting but they hit -- and you get tricked by the hit spark. We actually had one version of the game that had no hit sparks, and I thought the entire game was broken or something: how come people are hitting from so far away, and nothing looks right? It’s just that those hit sparks solve all your problems for free. There’s a much bigger margin of error than you might realize. I think we’re within the margin.

About the Author(s)

Brandon Sheffield


Brandon Sheffield is creative director of Necrosoft Games, former editor of Game Developer magazine and gamasutra.com, and advisor for GDC, DICE, and other conferences. He frequently participates in game charity bundles and events.

Daily news, dev blogs, and stories from Game Developer straight to your inbox

You May Also Like