Hello, I haven't written on the GameDev.net forum in quite a while, I got a bit caught up in making a game, but soon realized, yes harder then it looks. So I'm going to stick to ideas, hopefully some nice team will pick me up, but until that time, here's my article. I've been reading and watching a lot of hype about "Spore" from the creators of the Sims, Maxis. I have to say that I'm personally quite excited at the idea of being able to create nearly everything in a game, which of course, got the ball rolling for me. In an interview Will Wright said that everything is worked out by the computer, he said that no matter how many legs you put on it, or how few for that matter, the computer will work out how this creature moves and reacts, the computer works out all the shading also. Spore has many editors; it has a creature creator, a building creator, vehicle creator, etc. So I was thinking, why not make this editor business part of a serious real time strategy. The idea would be to have a simple RTS lay out of collecting a resource, building a base, building an army, completing a goal, But and it is a big but, why not have an editor for buildings and units, using lots of parts and colour schemes to create an army how you want it. You would start off with so many points you could spend on your first base and unit. Obviously this would be your first base, adding different things, for example a "giant pump" would allow the building to extract resource from a resource truck, or a "silo" part would allow the storage of more resource, etc. You would have a set number of parts you could put on per building/unit. Units could be three types unit, armor or air. you could simply add on different types of weapons, make them look how you want with plates of armor, stretching and rotating things like you would on spore's creature creator. As you work your way through the campaign you can complete objectives to get more points to spend on more base/unit designs. Other objectives could unlock new parts, etc. The editor would be separate from the In-Game meaning that before going on a mission you would have to build your base/army plans. This could lead to players specializing in certain areas i.e. defense, production, etc. It would also make for interesting multiplayer play. The down sides I could see would be balancing of units and bases, the Spore engine is very advance, also the multiplayer could take a long time to send as models and textures must be sent, though spore handles this quite well. Given the right theme/plot that could limit it down to organic aliens or robots or something of the like, so not to make the game to big. Of course I have just brushed over this idea, somebody has probably thought of it, or maybe even posted it. Ha. But that’s my idea of the day, I have it all pretty much mapped out, hypothetical questions anybody? Cheers.

Impossible Creatures did this on the unit side - you were able to customize and create your own units by combining 2 animal species together into monstrosities of war and mayhem. You can choose whether your unit should have a Chameleon's back for camouflage (stealth) or have that back be of a Porcupine's for a Quill Burst attack. You could select whether your unit is to have a head of a Giraffe for increased sight radius, or the head of a Shark for a nasty bite attack. Etc.

I like having a deck of custom units. Maybe have a SP like Total War (except with a 4x battle engine) where you can design your units in the turn based section. Also, maybe you would start off a skirmish with a base, then you would have to level up the base to be able to build your special units. The way to deal with balancing is having the upgrades have a good and a bad side. Heavier armour = slower movement and so on.

Problem is that Spore is by design a sandlot game. Its a collection of toys to play with. There isn't a set goal of beating other players, you play it however you want.

But by decreeing a overrideing goal like in a RTS. You are going to have to design the combinations in such a way to promote gameplay balance. Which essentialy means that the larger the number of potential units, the more alike they need to be inorder to have balance.

You don't want certain units to be so powerful over everything else that they become the defacto choice among players. You don't want one gamplay stratigy to overrule them all. So there needs to be a built in form of balanceing the game. Unfortunetly when you increase the choices a player makes, inorder to maintain balance, you basicly remove the value of the choice. It can make for a prety game haveing so many different units, but if they all act nearly the same it becomes a less interesting game to play.

Glad you liked it but also could see the design faults.

A very Rock-Paper-Scissors method would have to be instilled in the game.

"Are there basic, universal units, like a main building and a builder/resource gatherer? If not, how do you know what is the user's initial inventory?"

The initial thought was to give them "Types" they have to fulfill, like a check list, saying "resource building" "resource unit" etc. The user is given the basic parts for a basic functional army.

"How would resource costs be calculated? Is it just a matter of adding chassis+mods costs?"

Resources would be used in game, gather to create buildings and units. But points which would be given at the end of the mission/skirmish would be used to buy parts the player has unlocked; each part has a set cost. Each part also takes up a number of slots or space on the unit/building, which means smaller units for example couldn’t carry to much stuff, meaning that super troopers couldn’t be created.

"Would multiplayer have "leveling" to gradually unlock new designs?"

That is actually an amazing idea, this stops newer units getting designs they don't understand or deserve. This means people will play the multiplayer and the game's life span would be increased.

Thank you for all those who gave examples, I’ll look into it.

Any other comments?

Configurable units in an RTS is one of those things that sounds great on paper, but personally, I've never seen an implementation of it that I liked, for a couple of reasons.

Balance is certainly one problem; it's going to be incredibly hard to balance a game with such a huge range of possible unit configurations. You'll need to be able to twiddle more than just damage/defense/speed to make for interesting unit interactions, and the balancing difficulty will increase exponentially with the number of different options. I don't think there is a really good way to solve this; just take a very careful approach to component design, and playtest for about ten years.

Probably the biggest and most obvious problem for the player is the unit editor itself. It's fine to ask the player to muck about in a complex editor in offline sandbox games like Spore, or turn based games like Alpha Centauri or MOO, because time is unimportant. In an RTS, the player's attention is already in heavy demand, and expecting him to waste more time on an editor is only adding to the micromanagement burden, which makes games like Warzone 2010 virtually unplayable IMHO. The alternatives are to run the editor offline, entirely outside the main game, as per ruby-lang's suggestion, or to build the editor into the gameplay itself, e.g by breaking factories up into pipeline components that each build a different piece of unit; chaining these together in different ways gives you different units at the output.

Personally, I think everybody in this thread has the wrong idea about the idea of a unit builder.

The awesome thing about spore isn't that you can put extra parts on your creature to give it new capabilities, but that you can easily and quickly create a model of a creature that will be automatically animated to do things.

What you have is a really, really fancy modeling program that makes modeling (within constraints) simple enough that an average person can make something interesting and yet unique.

The use of this should be obvious - allow players to customize the appearance of their units in an "army editor" of some kind that is used completely separate from playing the game.

Creating the units themselves (that is, the parts that are mechanically important, such as damage, abilities, etc) is a completely separate challenge and should not be completely entangled with an appearance editor. I think having a free-form editor for mechanics (meaning selecting individual abilities and being able to tweak the numbers for each instance of the ability) would be really nice and would make an awesome combination with the appearance editor, but it is important to keep the two ideas conceptually separated.