Touck KO: Tech Review

Ding, Ding, Ding! Two combatants emerge from their corners, they jab, throw cross hooks, they bob, weave as the crowds’ volume swells to a deafening decibel as they connect with blow after powerful blow.
In essence, the preceding sentence only scratches the surface of what boxing is all about, it can be raw, visceral and most importantly entertaining – well unless you’re the King, then it’s all about the money.
Technically speaking, boxing is one of the oldest sports on earth, dating back well into the ancient history of the Greeks, who were credited in formalizing boxing as a “sport”. And if you think about it, why wouldn’t boxing be one of the oldest sports? When faced with confrontation, fist-a-cuffs usually comes naturally and even in this raw form, it draws crowds who chant, provoke and cheer the two combatants on. Even though boxing requires only two fists (and sometimes a headbutt or bite) there’s an immense amount of “sweet science” behind the sport. From the Philly shell to the cross hook to the uppercut, there’s a lot to be said about the sport of boxing – and that’s not even mentioning the footwork. So how do you whittle such an experience down to the iPhone platform? A platform with no tactile feedback and a game that requires very complex controls. To satisfy my inquisitive mind, I spoke with the Mechtley brothers, developers of Touch KO.

Tale of the tape

Published by Chillingo, at first glance the cleverly named Touch KO or “TKO” appears to be a heavily inspired Fight Night re-imagination for the iPhone/iPod Touch platforms, presentation wise the hip-hop track/s, outfitting your selected character with stat improving equipment are all things reminiscent to the comparable title. And with really no fault of their own, the brothers would be against the ropes from the very beginning, choosing such a well establish genre that has been now carved into the gaming psyche of many, it would be near impossible for the brothers to live up to gamers expectations – which is clearly evident in respect to some reviews.
Taking on a monumental task of providing an identical experience to that of Fight Night would not only be incomprehensible, it wouldn’t be the smartest way to spend your development time. To be honest, I’m not sure how many devs worked on the Fight Night series but I can guarantee you, it was more than two; and to take on the daunting task of developing a “fully playable” realized sport such as boxing would no doubt require a full scale team or does it?

“Pretty early on, we said that we were “inspired by” Fight Night, among other things. Some people got the misconception that we were trying to port Fight Night onto the iPhone, which is not only daunting for two people, but fairly unrealistic for a host of other reasons. We simply set out to make a 3D boxing game for the iPhone.”

Boxing is all about control, whether its in the ring or in a game, at this point EA has “practically” mastered the twin stick combinations to produce a well-rounded result, so it only makes sense that the brothers make that their initial starting point.

“The initial spark for the game was an idea for a control scheme, so that’s what the first prototype was. Once we got it to that point, we started looking at other boxing games for ideas of what to do with the game play and presentation. There were some things we liked and some things we disliked about Fight Night, and some things that simply did not make sense to transport onto the iPhone.”

{click image for larger view}

From concept to completion TKO began development in October of 08′, the first prototype was established at that time. Unfortunately the title took a brief hiatus to around January 09′, the brothers worked on the title part-time for roughly 6 months.

Real estate, real estate, one of the biggest issues for developers on iDevice platforms, with the lack of any tactile feedback controls wise, developers have to be rather judicious in their use of onscreen components. Too many digital controls your thumbs obscure the playing view, too little and it becomes frustrating to the end user, this is why I tip my hat to the brothers as they handled this particular area rather well, and with boxing being such an agile sport with complex movements, I was pleasantly surprised at how intuitive and responsive the controls felt.

“The core of the controls was the first thing done for the game, and they didn’t change too much throughout the course of development. The only thing that did change was that the first prototype used the accelerometer not for dodging but for leaning. The idea was you could hold your device in a tilted position to lean and still take shots at your opponent. We eventually scrapped this because it required huge amounts of animation production, and most people don’t like having to hold their device a certain way to play anyway — tons of gamers apparently hate using the accelerometer at all.”

Count me in as one of those gamers, the iPhone’s accelerometer controls when incorporated correctly in games designed specifically for its use work well, but as stated in previous articles, forcing this technology rarely leads to enjoyable results. Adam and Matt however opted to reduce the accelerometer implementation to sudden jerks left or right to simulate dodging. Even though this works well, I found often I had to remind myself to use it as this sort of movement doesn’t come naturally, so some refinement in this area is definitely needed. As an idea incorporating a gentle 15 degree rocking motion using the accelerometer may feel more natural all the while not skewing the view to the extent of breaking your game play.

In addition, movement was handled entirely by the CPU for both player character and your opponent. In all honesty I found this very jarring at first, since footwork and the ability to move freely around your opponent is key for many strategic boxers. But again as stated before, striking a balance between features implementation and usability especially on the iPhone platform can create a conundrum for many developers, and to be fair even the big boys (namely EA) haven’t totally nailed this mechanic themselves. Although I found the tandem movement awkward at first, after a few bouts it does become invisible, however a way to move your character back and fourth would be a welcome addition.

“There are actually plenty of reasons for not allowing free player movement. Part of it is real estate, but another part is just having intelligent controls.

One option is to use the accelerometer for movement, but as we mentioned, tons of people hate using the accelerometer, especially for precise movement.

Another option is to leave the controls basically as-is and include an analog stick to move. That’s fine, but it encroaches substantially on your left hand control area (especially if we want to add body blows) and is also pretty awkward to switch between attacking and moving on the left side.

The third option is to just let your left thumb stay on a movement stick and your right thumb handles attacks. This means you either have a huge number of buttons on the right side or you try to do a dual-analog approach like Fight Night. Without the tactile feedback of an actual joystick, trying to execute gesture-based moves on a right analog stick is actually incredibly frustrating.

That being said, we are not totally satisfied with how the character movement looks right now and want to continue to find ways to improve it in the future.”

{click image for larger view}

With that said, one of the first things I noticed while into the throws of my first bout, was the realization that there currently was no way to counter punch or more importantly a way to pulverize the body of my opponent. Even though I’m not an avid fan of boxing in general, if there’s anything Rocky IV has thought me, no matter how tough Drago was, taking away the body brings any giant down to your size. And yes I realize the example is fiction, but it holds true in real life boxing, body shots are essential in taking away your opponents legs and draining their stamina. So when asked about the omission of body shots Adam had this to say;

“Ultimately, there were a couple of problems with getting body shots in for release. The first issue is control responsiveness. When we first started working on the game, input processing from the iPhone OS was simply not robust enough to accommodate the quick flicks we needed to rely on. Consequently, dividing the screen up too much wouldn’t have given us the amount of space needed to really execute some moves (e.g. uppercut to the face versus uppercut to the solar plexus). Since both the OS and the Unity engine have been substantially improved since then, we can now go back and add body shots. We are currently working on them and they will be included in our second update, very shortly after the first.”

As long as technology keeps improving it will always be our limiter when trying to explore our creative juices, and if it’s not tech it will be time, the one thing we can’t change or alter. Consequently time was indeed the ultimate hindrance that came between the implementation of body shots in the initial release.
As a side note, a term coined by my production manager in-house was “feature creep” it was a perfect term which describes features not originally planned in your initial scope that creep in towards the end of production which can be quite costly in many respects.

One of the largest time-sinks in production is that of character creation, it’s one of the many things I’ve learned to respect on a day to day basics, in short creating characters take quite a bit of work to produce. From concepts, to sculpting the model in Zbrush or Mudbox, to creating a low resolution mesh and projecting the detail in the form of normal maps to finally texturing, rigging and animating, you quickly begin to see how daunting this task can be. Stating this, it would be an extremely lengthy procedure for one man to create a full roster of characters within the time allocated alone. As a dev team with limited art resources, one of the many solutions you can use to drastically reduce production time would be to simply reuse as many assets as possible, this includes rigs and animations that can be use on a wide variety of similar sized characters. However if you’re fortunate enough to use MotionBuilder in your pipeline, transferring animations between characters quickly becomes a non issue, but the caveat becomes apparent when those animation have to be paired between two characters at any given time.

{click image for larger view}

“We use the same rig but not exactly the same mesh. Obviously, if all of your characters are roughly the same size and you use the same rig, then it is easier to share animations across them, since you know footsteps and punches and things will line up accordingly.
We have five different character models right now, each with two different texture variants. The bodies are generally the same, but have slightly different head and face shapes. Since each model requires a high-res and low-res version for different contexts, we could not afford to really do tons of custom model work for now. When we add licensed boxers, this will of course change though.”

One thing to note is that Touch KO is a great looking title, and having been a rabid fan of the PSP version of Fight Night Rd 3, it’s safe to say with a bit more development time, Touch KO could surpass the aforementioned title with ease. Each fighter base model was created with roughly a 2,000 vertex count which was then divided up into separate pieces including the gloves, trunks, shoes and body. Using an industry standard body double trick, the mesh used in the slo-motion closeup sequences omits the lower regions to allow for higher detail and mesh deformation for added impact. The lower resolution is used for the actual in-game mesh since the camera is pulled back far enough to not notice any major differences. The technical term for this is call LOD (Level of Detail) where models are built with varying degrees of detail depending on the situation. In addition, the background characters were modelled and animated within Maya, which were then rendered into a 16 frame animated sprite sequence.

{click image for larger view}

Pages: 1 2

Pages: 1 2