I've always thought of my games as stepping stones. Each one learns from its predecessor, and then tries to bring something new to the table. That philosophy is brought to life with the new greatest hits CD: Evolution of the MPire! To celebrate the release of this landmark album, here's a summary of the breakthroughs that made it all possible...

Give Piece A Chance
It might not look much, but the graphical system from Federation Booker is arguably my greatest achievement. The big catch with 2D games, and the reason I no longer make them, is that it's very much like making a Disney animation - I.e. more trouble than it's worth! 3D models are like actors that you can hang clothes on, but 2D characters are pure work - nothing exists unless you draw every damn pixel. So to bring a large army of characters to my wrestling games, I had to do some serious thinking. The result was a system whereby each character was made up of 3 separate pieces: the head, body, and legs. It sounds innocent enough, but this was a far cry from what I had done in previous games. It literally meant I was dealing with 3 times as many graphics - plus I had to worry about keeping them all synchronized. Fortunately, I managed to pull it off - but not without wrestling with some mind-boggling mathematics! The results were worth every single headache though, because it finally meant that my 2D games were customizable. The player could mix and match each body part, squeezing thousands of possibilities out of just dozens of costumes. Without this breakthrough, Federation Booker would have had a fraction of the many characters that it ended up with. In fact, it's no exaggeration to say that the game probably wouldn't have existed at all. More information about this technique can be found here...

Spare Parts
The struggle that went into a 2D editor is made all the more frustrating by the fact that a 3D version had always been so easy! In fact, my very first 3D game boasted such a feature. Although it was as clumsy as you'd expect, it was the pride of the original Big BumpZ game of 2001. As I mentioned earlier, the beauty of 3D models is that they're the gaming equivalent of "actors" - and can be relied upon to bring much more to the table. I designed the graphics for Big BumpZ with that in mind, and the result was a character that consisted of several distinct limbs (a preliminary sketch can be found here). The idea was that you could decorate each limb independently, instead of painting a whole preset costume to the skeleton. The effect was somewhat ugly, but like the 2D breakthrough, it was a triumph of efficiency. It once again meant that there were thousands of possibilities - and those possibilities were in the player's hands as much as mine! The fact that an increasingly better version of the editor has appeared in every 3D game I've made since is testament to what an important breakthrough it was...

Model Behaviour
By the time the editor reached The MDickie Show, it was taken one step further with the addition of "accessories". This technique was almost a breakthrough in its own right! I had gotten so comfortable with manipulating the models that I realised I could create all kinds of illusions. The premise behind this technique was that each model was littered with hairstyles, headwear, and eyewear of every possible kind. By showing the ones that were needed, and hiding the ones that weren't, any combination could be portrayed! For instance, when you change hairstyles you're just hiding one part of the model and showing another. The illusion even extended to the actual game, whereby you could "lose" accessories during the fight. Once again, they were simply made invisible to the human eye. As fanciful as it sounds, this was actually a very important development. Up until this point, characters with accessories such as hats needed a model of their own. This way, each basic model could have numerous variations...

Figure Hugging Fashions
With such sophisticated models, it was always a shame that the texturing didn't do them justice. When Wrestling MPire came around, I was adamant that progress would finally be made in that department. Not only were the textures better drawn, but they were also better applied. Due to my inexperience, I had never gotten too adventurous with the concept of "mapping". I simply threw the textures on there with the most simple method, and left it at that. After a couple of years, my standards surely needed to improve - so I toiled away at 3D Studio MAX to learn some new tricks. The result was yet another breakthrough in modelling. Instead of separate textures, the skeleton could now be fed an entire piece of clothing and spread it out automatically. From one image of a pair of legs, the model could now portray everything from the hips down to the feet. Not only did this mean that the limbs flowed more consistently, but it also meant that changing costumes was a breeze! Just like in the 2D version, all the player had to do was choose a head, body, and legs - and the rest wrote itself. This was a far cry from the earlier attempts, which required each individual limb to be decorated. The technique proved to be so accessible that homemade costumes became more popular than ever - and helped to seal Wrestling MPire's place as the number one game...

Skeletons In The Closet
In amongst all these cosmetic breakthroughs, there was a programming revelation that helped to bring the models to life. By the time the all-new Big BumpZ rolled around, I had mastered a new technique whereby the same animations could be applied to any skeleton. Traditionally, a different shaped body would require its own set of files - but my technique could extract the same results from any model, whether it was fat or thin, and even male or female! This was nothing short of a revelation, because otherwise I would have had to re-animate everything several times over. As long as each skeleton had the same correctly-named body parts, they were happy to use the same animation data. The fact that the skeletons had to be so similar did have its drawbacks (especially for the females), but the pros far outweighed the cons. Like many others, this breakthrough has been honoured by time - appearing in every 3D game since its discovery. By the time it reached Wrestling MPire, things were made even more efficient. This time, even the animation sequences didn't need to be broken down! Instead of a file for each and every action, the game could now read from one long sequence of movements. This meant that the game packed up smaller and ran smoother, because there were far fewer files in the system...

Movers And Shakers
My biggest achievement in 3D game programming - and arguably the greatest of them all - is the move animating system from The MDickie Show (and later Wrestling MPire). Like many of the other features, I have no doubt that this has been taken for granted. Most people see a game in action and assume that it's nothing more than a series of animations - but nothing could be further than the truth! The animations are just the tip of the iceberg, as you must then tell the game how to deal with them. When it's wrestling moves, that involves bringing two separate animations together in perfect synchronicity. To pull it off, I created an elaborate scene in 3D Studio MAX - where I could animate two characters together and then extract them as separate files. That once again was just the tip of the iceberg. Another animating dilemma is that the game doesn't care what's happening in a sequence. If you animate something like a Suplex, the game doesn't know that the character has fallen backwards - and it certainly doesn't know what happened to the victim! Educating it on the matter required the most mind-boggling mathematics I have ever had to program. It combined data from the animation file with data from my game world - which is the equivalent of translating something from Russian to Japanese! Miraculously, it worked out okay and the game could finally detect where a move was taking each model. You may notice that the effect isn't perfect (with the characters sometimes getting out of sync when pushed against a wall, etc), but you can rest assured that I put as much work into those moves as the wrestlers do in real life!

Collision Course
Long before I had to worry about moves flying through the air, I was pre-occupied with the simple premise that you shouldn't be able to walk through things! Believe it or not, the collision detection in my earlier games was so abysmal that you could do just that. The laws of physics didn't apply until Federation Booker's 2D gameplay hit the scene in 2002. I made every effort to ensure that the items felt solid for once. It's not as easy as you'd think, because blocking things has a frustrating "knock on effect". The secret to collision detection is that you identify when a character is invading an item's space, and then push them back to the point where they weren't (before the human eye notices!). In the wrong hands, this technique can result in "sticking" to the item or wall in question. After yet another session of mathematical gymnastics, I finally learnt how to make a character smoothly slide against a surface as they would in real life. The system certainly deserves to be called a "breakthrough". Not only did it polish off the 2D wrestling gameplay, but the exact same premise was transferred to 3D for all the games that followed! Actually, by The MDickie Show the 3D world demanded a little more fine-tuning. In addition to an item's location, I now had to assess how its angle affected things. This was yet another mind-bender - not unlike the animation problem above - but I managed to pull it off. The result was that my furniture-filled gameplay survived the leap from 2D to 3D, and even managed to become wilder than ever...

Sound Advice
Not all breakthroughs are up there to see on screen though. One of my most enduring techniques is a simple twist on the sound effects. Ever since the first Federation Online game in 2001, I've squeezed extra life out of sounds by altering their "pitch". That affects whether they're played fast or slow, and by mixing it up you can make an effect sound less repetitive. Try it with something like a metal weapon, and you should notice that the clang can either be sharp or dull. The same thing is done with practically every other impact sound in my games - and even things like the crowd. The results are not to be underestimated. You might think the sounds are bad now, but they'd give you an even bigger headache if they're were identical! With all those fists flying, it's a godsend that just a few samples can keep everything sounding different. The technique culminated in The MDickie Show, where it was even used to give the characters different voice tones. The same lines of speech could be made to sound deep and masculine or high and effeminate! It was instrumental in making each character (and conversation) as unique as possible. So although my sound work is notoriously awful, I can at least claim to be resourceful with it!

The Fun Comes First
There are countless other little achievements that keep the MPire strong - too many to mention, in fact. As different as they might seem, they all come under the banner of "the fun comes first". My work might be flawed, but it's dedicated to pushing things forward. Whether it's the programming from Federation Booker, or the low-poly modelling from Wrestling MPire, each game came out the other side with dozens of characters on screen! That's no coincidence. My games are designed to achieve breakthroughs like that. You drop the quality in one place to improve it somewhere else. Whilst the commercial developers exchange gameplay for graphics, you can rest assured that I'm doing it the other way round! Every single day, another breakthrough is made to keep it that way...