Applying physics to your games can mean more than upturning
tables and ragdolls.
Time was, adding physics to your game meant that barrels could be rolled, crates knocked, and enemies could be sent spiraling (or rather flopping) to their demise. But a new trend has started to emerge, in which games don’t ‘feature’ physics – they’re built entirely on top of it.
“We see far too much physics in games that just add that little more realism to the environment,” says Havok’s principle engineer Dave Gargan. “But when you add physics into the gameplay, that’s when it’s really exciting.”
“Look at Motorstorm 2, for example – their foliage system is basically sets of physically simulated plants, but what that lets you do is go right through the foliage if you’re in a big truck or something and potentially clear a path for people that are behind you on bikes. So, suddenly, physics becomes this tactical thing in the game, rather than just some environmental effect.”
Hanging parts of your mechanic on physics doesn’t just help come up with new concepts – it also makes your game easier to understand.
“There’s no country borders on physics; wherever you are in the world the laws of physics are pretty much the same,” says Takeyuki Ogura, chief software design engineer for Japanese middleware provider Prometech.
“So, in other words, it’s totally unconnected to where the game was developed, cultural differences, or even skill as a gamer – everybody can quickly grasp the rules. It doesn’t even matter if the game is unrealistic – it’s still hugely important that anyone can understand the rules quickly and get into the game straight away – but if those rules are based on physics that job is much easier.”
The biggest problem when making physics a large part of a concept is enabling the player to interact with the world in an intuitive manner; in a way that gives the player a sense of physicality in the environment and a sense of immediate control over the objects in the world. “When Havok started out, physics for us meant car racing games, simply because everybody knew how to interact with a physical environment with a car,” says Gargan. “It’s simple – you drove into stuff and you knocked it over.”
The next round introduced the concept of the player as a two-dimensional sphere – games such as Super Monkey Ball and Katamari Damacy – again because the concept of rolling is a simple one to grasp. When you try and get a step closer, though, things get more difficult.
“When you extend that to levels that have lots of dynamic objects you run into problems really quickly, because it’s hard to interact with something that’s got six degrees of freedom when you’ve only got thumbsticks to try and organise it.”
One of the things that’s helped with this problem, and has allowed for more physics-based games to flourish in three dimensions, is the Wii remote. Although it doesn’t give a full six degrees of freedom, the nature of the controller makes a much more intuitive (and in some cases more accurate) interaction paradigm than traditional input devices.
If you’ve not got the benefit of the Wii remote, though, another way around the problem is to reign in an aspect of the physics to ensure that the game is still fun. “If you look at something like LittleBigPlanet, for example, that game is basically constrained to 2D or 2.5D,” says Gargan.
“And that’s what makes the game fun, because you don’t end up with this frustration of, say, almost stacking two things on top of each other and then having it all fall down because it wasn’t perfectly balanced.”
A great example of a rapidly growing area of physics games is Prometech's OE-CAKE!, a piece of software available for PC and Mac. Initially conceived as little more than a demo application to showcase its 2D soft-body physics and fluid/gas simulation middleware OctaveEngine Casual, it managed to attract a sizable audience – not of middleware purchasers, but regular gamers.
The app lets users draw and interact with objects made out of different types of matter, from normal solid rigid bodies to elastic and brittle materials, viscous fluids, powder and fuel. The result isn’t really a game – although Prometech’s Ogura says “you can see that it has a hint of a game about it” – but rather an interactive physics sandbox, the opportunity to play with objects and matter that would be difficult – or, at least, messy – to do in real life.
Where it becomes a game is entirely in the hands of the user: taking subtle hints from a few of the pre-packaged scenes, and utilising the ability to save and share level files with other OE-CAKE! users, challenges soon emerge. Get the water from here to here, destroy this object using only powder, make a volcano.
The program has spread around forums and blogs, gaining a popularity – and showing creativity – that Prometech hadn’t anticipated.
“Users have uploaded a huge amount of their OE-CAKE creations to YouTube,” claims Ogura. “There’s been a big reaction to it, and there are already a huge number of creations that we weren’t expecting.”
OE-CAKE! isn’t the first of these physics sandbox apps to surface, of course – LineRider has enjoyed similar success, with over 11,000 videos shared online and over 16 million impressions to date, despite it too being an ‘aimless’ exercise in user-generated physics.
Havok’s Gargan doesn’t think that these applications are any less worthy of attention: “I think that these collaborative games like LineRider are where we’ll see physics games grow, rather than say really heavily physics-based puzzle games.”
So where does the future of physics in games lie? Will it be through this introduction of physics to core gameplay that will see the field really reach it’s maturity? “I think it still has a long way to go, and having physics integrated actually into gameplay is a step on that path,” explains Gargan. What Havok is seeing demand for in the short term essentially filter to two big things – real-time destructable objects and physically-based animation.
The former, which Havok is targeting with Havok Destruction, will again allow physics to interact with the game on a core level, not destroying objects for eye candy but to open up new routes in a level or take down enemies in an efficient manner. Of course, a completely open application of this technology would open up big problems for level designers, accustomed to using static geometry to shepherd players through particular routes of a level - and so, once again, the problem becomes one of constraining that ability to ensure the integrity of the experience.
The other aspect, being addressed not only by Havok but also companies like NaturalMotion, will help increase the physical integrity of actors in the world – and this is one of the things that Gargan thinks will herald physics’ maturity.
“Until I can actually interact with objects in a game with any sort of constancy that’s anything close to way I interact with them in real life – well, physics’ game applications haven’t reached maturity. We’re nowhere near that.
“While the rigid body simulation is basically a solved problem, I can’t even lift a phone receiver and make a phone call in a game without animators actually making that possibility. So we’ve still got a lot of work to do, and a lot of that is on the control side.”