Day dreaming designers required

Started by
8 comments, last by Limdul 16 years, 5 months ago
In the game we're working on, pet wars we have a pretty difficult design problem to solve. The kind of problem day dreaming designers relish and rarely actually get presented with... Basically pet wars is an rts, but instead of building buildings and generating units, you gather weapons and defenses from within the location and recruit units from within the location. We want players to be able to combine the items they collect in non trivial ways. We are powering the game with a physics engine which means we can do trivial combinations like putting a can of petrol in a trolley, or laying planks on top of one another to create a wall - but ideally we will allow players to combine a firework rocket and a pipe to create a makeshift bazooka, combine a plank with a log securely to create a sturdy catapult. Our dilemma is that if we start prescribing which combinations can occur, then we are breaking the consistency of the world the game takes place within, giving an awkward and irritating playing experience. If I can combine a pipe and a firework rocket to create a bazooka, why can't I combine the pipe and and this hairdryer to create a precision blower? Or why can't I combine the pipe and the rocket in a different way to produce a different hybrid object? On the other hand, only allowing trivial combinations might be a bit boring and a waste of an oppurtunity to allow for more sophisticated play - allowing the player to exercise his creativity and intelligence in combining things at the cost that all his ideas might not be supported? All thoughts welcome!
Advertisement
Wimps!
I'm not sure I understand what you're asking but I'll give it a shot. Here's what i see your question as: "We're stuck between two extremes - predefined behavior for items or completely free form creation. Which one is better, or how can we balance the two?"

You can explicitly define item combinations that the player can make, which will allow for more complex creations and interactions at the cost of denying the player the ability to come up with his/her own inventions beyond what you have already created. Once they've exhausted the predefined combinations there is nothing new to create. Getting better at the game consists of finding the best combinations that can be made and using them well. Strategy and knowledge are rewarded, not creativity. If you take a step back that's rather RTS-like.

You can also allow the player to do anything they want. But, this creates a set of problems where you're heavily reliant on your game engine to interpret and model the behavior that the player wants (NOT necessarily what will happen according to your physics model). For example: It's no fun if to spend a lot of time building a super firework cannon with 8 barrels that LOOKS like it should work, but because of either limitations in the game UI (I couldn't get the barrels lined up right) or engine (the physics model says my cannon is so heavy it should tip over) or my own engineering abilities (just because it has duct tape on it doesn't mean it will work!) it falls apart when I try to use it.

The best option, in my eyes, is a middle of the road approach. Think of objects in your game world like LEGO blocks. LEGO can be used to create just about anything, but at their base the blocks only fit together in a few different ways. Any one piece WILL go together with another. They are of a standard size and form factor. They are limiting enough to help ensure that what you build stays together, works, etc. But free form enough that your imagination is what limits you- not the way the blocks can be put together.

So, for each object define some number of attachment points- spots where any other object can be attached using its own attachment points. Give the objects a "sticky" behavior in the UI that tells the user "this'll fit here." You'll probably want to allow the player to rotate their objects about the attachment points.

Say you want to have a truck axle or two lying around the game- making vehicles could be fun. But it doesn't make a TON of sense to attach things to anywhere but the stationary part where the truck frame would be bolted.

Maybe it would be wise to give the player the option to disable attachment points and do things truly free form for some objects- say, a hotkey they can hold that lets them freely move and mount their creations anywhere they want. Then if the player wanted to add spikes to the wheels of their truck axle they could.

For your blower example, you could define your blow dryer nozzle as a "sticky" point, and the ends of each pipe as sticky points, and make both the same size, thus allowing them to be connected tightly without a ton of air leaking out. You could define edges and faces of plywood boards to be sticky. Let the player put them together to form a cube, stick the handle of the blow dryer to the top of that box, then sit the whole thing on a pair of axles - bam, you have an APC stuffed with adorable kittens or puppies! The player could also put the nozzle of the blow dryer up against the flat face of the plywood and have... nothing happen. The game engine doesn't have to know that blow dryer + pipe = air cannon, just that blower nozzle + pipe end or blower nozzle + float board = air tight seal.

In summary what I'm saying is give basic objects a little bit of predefined behavior to make the player's life easier, to limit them some- but not a lot. The basic behaviors that you define can be combined to create something more complex.
I think you may bit a bit ambitious.

If you want what you're describing you'll have to program the game to respond to situations as the real world would. for example you make shift rocket launcher.

From your description I'm guessing you create a rocket class, a pipe class and a rocket launcher class. instead remove the rocket launcher class there is no such thing as a rocket launcher only the two separate objects and how they interact. the rocket has a member function say light, which is called when the rocket is ignited the result of this is the rocket produces a force out the back with no solid object to propel off the rocket doesn't go very far. Now when you place the rocket in a close ended pipe there is now something to push off so the rocket flies. If you placed the rocket in an open ended pipe you would get the same reaction as with out a pipe.

Now lets look at your catapult it needs more a plank a log and something to act as a spring. The appearance of spring here brings me to the main point lets say you have a piece of rope to use as your spring as in the good old days. now just adding rope will not give you a catapult you need to twist the rope enough.

So what does it all boil down to forget combinations just describe the physical properties of every object in your world or at least the ones the player can interact with and if they are described in full such as material properties, forces, energy. if the object is fully described as it is in the real world it will react just as it does in the real world.

Another option is to go with predefined combinations but make them funny or unrealistic this will stop players think well if this was real I could do that why can't I here.
For the reactionary components of these combinations, I would suggest creating a list of attributes which might be present in any one part. These attributes, if active, could be set to react to other attributes to trigger certain abilities or allow certain actions.

For example, combining the "blow air" attribute in the hairdryer with the "precision" attribute of the pipe would generate a more precise blower experience. I suppose, more functionally, the pipe's "hard weapon" attribute might also combine with the dryer's "collateral damage" attribute to shower the victim with plastic pieces when hit. The dryer might also have a "breakable" attribute, meaning that it is destroyed after one attack. As far as how they combine, or stick together, Detroit above has some excellent ideas.

This method would require the creation of an extensive table detailing the triggers or actions from every attribute combination. You could also add a rating to each attribute for smaller and larger reactions. Sounds like fun.

I hope my daydream ramblings are helpful!

Cheers,
Tim
GameDeveloperTools.comA comprehensive library of game development resources.
### You have to FOCUS YOUR MAIN THEME OF YOUR GAME

- If it is the 'combination variety'
The combination process should be categorized.
If ye are manufacturing a weapon,
[attack medium] rocket (for rocket launcher) | stone (for catapult)
[momentum source] rocket igniter | elastic stick and rope
[aim medium] pipe | none
[mobility medium] none | a 4 wheel body
etc etc

- If it is the 'test of player ingenuity'
Then we are talking about something like Incredible Machine craft system.
There should be another 2D interface so that players may put gadgets and connect them one another, try out and release as a 3D model object.

I would give you more if you clearly tell me what your main concept of your game be.

[Edited by - Limdul on November 2, 2007 7:25:34 AM]
*agrees with Limdul* Focus your theme before going too far. You don't want to unbalance the player's time + effort spent in one section when you intend the game to be something else.

That said...

Check out Gary's Mod for HL2 to see a great "construction set" sandbox. There are no objectives in the game, but you can create a lot of stuff using the Havok physics of HL2. They've gone a long way with the interface and mechanics of creating complex stuff from simple pieces. You'll quickly see some of the limitations though.

http://www.garrysmod.com/about/



Detroit rock your suggestion might be spot on for what we need, each of our objects has points that our animals grab onto to drag the object - these could also be points where objects are combined...

Ramearess, I think you could be right that with a high enough quality of humour associated with the combinations, players might not mind only being able to use a few. It does change what the game is then however - it becomes a logic puzzle rts! I Doubt thats an existing genre. And when all the combinations have been seen it will decend into a regular rts with very practised strategies.

Tim thanks for that too - a good solution but might be too difficult to extend beyond a few items, and we are leaning toward something we can apply to all items...

I don't really understand why you guys think that area needs focussing -
And theres no easy way to explain the concept other than its a real time battle game where the environment is highly interactive.

I'll check that mod out, thanks :)

EDIT : A better description of the concept : remember the name : petwars

[Edited by - thelovegoose on November 3, 2007 7:40:58 PM]
I'd suggest splitting it into 3 different rule sets: weaponry / vehicles and buildings.

Weaponry should be able to link into vehicles and buildings - to give them mounted weaponry.

The weaponry category could include anything that has the ability to be 'used' (either by a unit, timed or scripted etc) that will then cause an 'effect' (damage, fire rocket, heal, blow away, give buff, stun, nuke etc).

The use and effect attributes of any object are defined by its components.
'Use' Examples:
A trigger would suggest the object is used by a unit (killer puppies or whatever). Another handle +trigger would suggest its handled with 2 hands.
An alarm clock or timing mechinism (egg timer) would suggest that the object will do its effect after a set time.
A radar / motion sensor would suggest the object is scripted to cause its effect when something comes near. A trip wire or pressure plate could be used to cause the effect when a unit/vehicle etc passes through certain area.

'Effect' Examples:
Would range much more wildly. From fireworks, ammunition, medicine, fan, drug/poison, etc.
These would define the type and amount of damage, which buff to give to whom or any other effect.

Some pre-defined objects would already set up - for example the hair dryer which is a combination of a trigger, fan and outter casing.

As for actually building the weapon, I would suggest a couple of basic 'building blocks' which can be joined together. For example a plank of wood could be used to create a stock for the weapon, where as a couple of planks of wood could be joined together to create a tri-pod. Both of these ideas will add accuracy to the weapon.
A lead pipe would work as a barrel, the more lead pipes, the more barrels so the rate of fire goes up. However a lead pipe and a sheet of metal could be combied together to create a sealed hollow tube for a pipe-bomb, or sealed at one end for a mortar launcher.

The interaction of building blocks (in this idea) would be pre-defined, so you basically end up giving the player the 'type' of weapon - pistol, rifle, shoulder mounted, grenade, mine, potion. When then placed together using something like detriot's ideas, with 'use' and 'effects' attributes you have a defined weapon.
This doesn't make use of physics per-say (doesn't determine if a rocket can travel down the barrel or whatever, just knows what is fired when its used, doesn't figure out if the amount of force you have stored in the rope/spring is enough to launch a boulder).


Example Weapon:
Building Blocks: 1 tube (barrel), 1 cardboard box (ammo box), 1 piece of string (ammo belt), 3 planks of wood (tri-pod). These objects and their uses are predefined by you.
'Use': 1 forward facing motion sensor + 1 'special' trigger (here, the special trigger takes orders from the motion sensor, not a unit).
'Effect': Bullets

Result: A tri-pod mounted sentry gun that fires bullets at targets that are in a 45 degree arc in front of it ("its field of view").

A combination of ammo box + ammo belt + bullets could be used to calculate amount of ammo / clip, reload time.
A combination of motion bullets + sensor + tri-pod + barrel could be used to calculate the accuracy of the weapon.
The bullets determins its damage, speed of projectile, penetration etc.
Roger that, lets run like hell!
The overall idea is good.

However, we will be stuck in some situations whether to maintain some game component while giving up the other things since we cannot embody all of our happy imaginations, finally in need of prioritizing things to implement.

Thus, we need to focus on some core game components - or in other words, main theme? - so that when we are dealing with practical/potential feasibility, we could protect the backbone of the game.

Good luck :)

This topic is closed to new replies.

Advertisement