Home > Technical > Combo Design Pathways

Combo Design Pathways

The initial spark is always accidental. You fire up Training Mode, you pick a character, you mess around, you try something on a whim, and it happens to work. Now your real creative task is surrounding that discovery with the right pieces to refine it into something special. Choosing the right direction is the single most important step to creating a good combo. Here’s a basic roadmap of the most prominent routes.

    Build Around an Element
The direct approach is starting with one substantial breakthrough, then expanding the combo in both directions. Let’s say you notice SF4 Akuma’s c.MK xx HK Tatsumaki Zankukyaku landing in a strange manner that allows him to follow up. What’s the most impressive combo you can construct atop that foundation? Since Akuma requires lots of meter for extended juggles, how about arranging a five-bar combo and saving most of it for the end?

    Connect Multiple Elements
A slightly more complicated method is taking two or more such pieces and tying them together. The more you add, the more challenging it becomes – and thereby more rigid. Sometimes you’re left with only one viable path from one point to the next. Of course, it depends on the game engine. For instance the Versus series is considerably more accommodating than the oldschool SF2 series in this regard.

    Follow Through on a Unique Setup
Once in a while, an innovative setup can be intriguing enough on its own to warrant a combo demonstration. It’s usually impossible to add anything extra prior to such elaborate provisions, so you can only build forward in these cases. The distinct benefits provided by each individual setup tend to dictate their resultant combos. Try to utilize every advantage gained.

All three methods above evolve organically from a predetermined core. What if you want to select an arbitrary character and start from scratch instead? There are ways to do that too.

    Build for Damage
Most combo video makers have gradually begun ignoring damage unless specifically searching for touch-of-death combos. Maximizing damage output demands a straightforward, methodical approach which produces relatively dry material. Even after developing a solid grasp of any game engine’s reduction scale, there’s still no shortcut around the massive amount of trial and error required for these types of combos.

    Build for Hit Count
Seeking the highest number of hits leads into the same linear brute force mentality. You simply stack up as many attacks as possible without worrying about what the combo looks like, under the assumption that it’ll add up to a noteworthy accomplishment. In fact, any combo built around the question “how many?” falls into this category. How many Air Slashers can Dee Jay perform in one uninterrupted block? How many fireballs can Sagat throw in a row? How many walking jabs can Ken link under certain circumstances? Whether or not these queries will lead to interesting combos is essentially a toss-up.

    Build for Variety
The goal here is to inject as much attack diversity as possible into one combo, making this perhaps the most dynamic character-centric approach. It still requires a fair amount of trial and error, but quickly transforms into a conceptual puzzle. Even deciding on a good starting point can be a tricky challenge – producing consistently stylish results.

Well, that covers the main categories. Keep in mind, these are only general guidelines. Don’t be surprised to find that some combos defy classification and don’t quite fit any generic mold.

Categories: Technical Tags:
  1. July 7th, 2010 at 04:11 | #1

    I’d been meaning to write this for a while, but i haven’t gotten much sleep so i’m not sure if it’s clear enough. This is a follow-up on a combo construction article i wrote a while ago, except now with more specifics and examples. Hopefully they help illustrate the points.

  2. CPS2
    July 7th, 2010 at 05:40 | #2

    I’m just wondering, do you figure out links, juggles (and trades?) and other game specific stuff (e.g. combo into EX level 2 focus attack) first, and how long does that take you? Or do you kinda skip any of those steps and have shortcuts for interesting ideas, then build outward if you have to (i.e. if it works, then see what works from there). I know that for instance finding 2 hit links can be done pretty quickly, but if you were to write down all possible 3 hit links and test them, it’d take a REALLY long time and isn’t worth doing, so that’s more of an experimentation thing (rather than meticulously testing). Is it like that for any other elements? (not sure if this question even makes much sense haha)

    The first sentence “The initial spark is always accidental,” I get what you mean but it’s kinda hard to define right? You gotta know something before you can mess around and find these interesting things by accident, so what do you tend to look for when first starting out with a character?

  3. July 7th, 2010 at 10:51 | #3

    It’s not always accidental stuff in training mode. There are times when you can figure out something cool in your head.
    For instance in this combo http://www.youtube.com/watch?v=o22DePR7H4M#t=0m49s
    seeing how maj abused the ability for rose to get a bunch of fireballs on the screen with Rose’s reflect. Thinking about how else it might be abused, it made sense to try it vs rose as she has the slowest projectile and a good dash. So I decided to try a reflect, reflect fadc, reflect fadc, reflect combo before I even looked at the game. Still there is an eureka moment where an idea happens

  4. eredef
    July 7th, 2010 at 12:09 | #4

    Excuse me Maj, I know this is not a Q&A though this article has gotten me quite interrested and i have a little question:
    Do you know any programing software i could use on sfIV pc to give combo creation a try?

  5. July 7th, 2010 at 12:25 | #5

    @eredef
    I would be the one to ask about that, try AutoMacro http://code.google.com/p/macrolua/ . Ask me if you need some more help or examples. Also if you want to just try raw auto it, I wrote a tutorial here http://sonichurricane.com/?p=1493. Maj needs to post something about automacro after evo anyway if he gets a chance.

  6. eredef
    July 7th, 2010 at 12:40 | #6

    hum never mind I’ve just checked error1’s video and there’s a link to scripting tools, sorry
    Ty anyway

  7. eredef
    July 7th, 2010 at 12:42 | #7

    and rewoops sorry for the semi-triple post but ty error :)

  8. July 7th, 2010 at 12:59 | #8

    CPS2: Here’s the way i look at it. When SF4 came out, i could’ve spent the first month recreating SF2 combos in 3D. I’d have no trouble filling up a 10min video with recycled material. But did anyone really want to see j.HK, c.MP, c.MP, c.MK xx fireball in 2008? I certainly don’t.

    So what we all did was go into Training Mode and try random stuff – new moves, new chains, old links, old tricks, old juggles, etc. until something unexpected happened. For example, i got Ryu’s j.MP to connect once against an airborne opponent then pressed a random button on the ground and it connected. That was the first thing to surprise me. So i focused on it and gradually developed it into something more by testing its limits, finding an ideal setup, trying different finishers, and so on.

    Of course the development process is much easier if you’ve been playing a game for a long time, because familiarity speeds everything along. When i stumble onto something cool in a game i’ve never played before, it takes me much longer to refine it into something presentable.

    But the bottom line is, combo making is about introducing a new element into the field to shake up conventional knowledge. The first step is luck because it has to be. It can’t be new if everyone knew it was there (although familiarity helps narrow down your search by keeping you focused on promising areas). The second step is transforming that raw element into a finished product by pairing it with the right ingredients. That’s how you go from “weird” to “wow!”

  9. July 7th, 2010 at 13:19 | #9

    error1: Yeah but my video put that discovery in your lap. How did you know all those fireballs would stay active? That’s luck – just like seeing something crazy happen in a match when both players were mashing. It was luck when i happened to find it too. You can’t take credit for inspiration. But you do have control over turning it into something unique and optimized. It’s up to you whether you accomplish that through rigorous testing of every possibility or whether you simply plot it out in your imagination. I totally agree with you that those “eureka moments” are more satisfying, but that’s not where the original pieces come from. You can’t connect the dots on a blank sheet.

    eredef: I definitely want to write a bunch of articles about tool-assistance methods, but it’ll have to wait until after Evo. The thing is, i want to do a good job of it, which requires time that i haven’t been able to free up.

  10. July 7th, 2010 at 14:33 | #10

    @Maj
    certainly that combo wasn’t anything new and you get full credits for figuring out how to abuse Rose’s reflect, but then how many combos don’t build off of other combos.
    A better example might be
    http://www.youtube.com/watch?v=NadAipFN6_A#t=0m27s
    I got the idea for the combo just by thinking gouken could probably trade an ex rush punch with a limb and knock the opponent into a fireball he threw earlier. I don’t make nearly as many combos as you do and they are certainly not as constantly excellent, but normally I get the idea for a combo in my head first and then go into training mode to test it. Sometimes the crazy unexpected happens
    http://www.youtube.com/watch?v=zPmlYVQltGI
    but normally it builds of the original idea I had

  11. July 7th, 2010 at 15:48 | #11

    No i’m not trying to take credit for it. I only happened to stumble onto that spark. But you built as much out of it as i did. I guess it’s misleading to say, “That starting point was discovered entirely by luck,” because i wasn’t just mashing. And you certainly didn’t arrive at your Gouken combo by trading moves completely at random.

    However, you did have prior knowledge which you got from outside sources. You knew that Gouken’s EX palm launches really high, you knew that it creates a juggle state, you knew that airborne stretchy limbs make for good elevation setups, you knew that Gouken’s EX demonflip autotracks from full screen, and you knew his fireball would provide a free juggle state. Knowing all this, you did the mental work of combining them into something elegant, then you tested a bunch of stuff to find the best finisher. That’s cool.

    The point i’m trying to make is, that mental work wasn’t present in discovering most of those pieces in the first place. And whichever one happened to trigger the domino effect was kind of given to you as an inspiration.

    I guess my stance is that we’re generally not “creating” but rather “refining” or “processing.” I feel the same way about the bulk of human creativity. And honestly, most combos definitely feel like they were built primarily via established channels. It’s really, really rare to see a combo so completely foreign that you’re like, “Where did THAT come from??”

    That said, the end result is still progress. It doesn’t matter whether it happens in reasonable, predictable steps or in massive, unforseen leaps. Personally i’m content as long as we’re pushing the boundary.

  1. No trackbacks yet.
You must be logged in to post a comment.