Hello everyone,

I want to create a RTS game with RPG elements. Here is a brief description of my planned game:
-Medieval fantasy setting.

-Traditional bird's eye view.

-Multiplayer supported.

-The main game scene will be a world map. From there, player can access to different interface tab such as his character tab, town economy,

-Town based mechanics like Civilization where players only have to build and upgrade their town. An example of RTS game following this approach is Supreme Ruler.

-Players play as a character instead of a nation/faction. Each character will have their stats, an avatar and unique traits that affect AI's behavior. Characters can marriage and have offspring. If a player's character died, he can continue playing as his character's offspring.

-Each army is represented by one single 3D model, and combat will be resolve via formula. There will be no combat scene.

-I want terrain to have important impact in this game. Calvary army will receive penalty when combat in rough terrain. Tree can be chop down to build siege engine, but forest will regrow after a while. Armies move in desert will receive attrition, etc...

The graphics aspect of this project should be light-weighted as possible. I'm more into creating gameplay mechanics than designing graphics. However, I want to create a detail and unique portraits for each character. Each character should carry an unique ADN string. When player access character's tab, the game engine will extract the ADN string to draw a unique 3D portrait.

That is a quick view of my project. But I still cannot decide where to start. Is there an available engine that saves me from all the ground work, especially the character portrait part? I knew UDK and CryEngine have mesh deformation system that would really help creating custom avatars, but do they unnecessary increase my game's hardware requirement? Using an engine also take away the ability to write low-level code, does it have a big impact on the performance of this kind of game?

Which tool should I use to speed up the developing process? Or is it better to have a frest start from a graphic API? I have very little knowledge in graphic programming though.

Any advice is welcome.

Which tool should I use to speed up the developing process? Or is it better to have a frest start from a graphic API? I have very little knowledge in graphic programming though.

I feel this question answers itself... If you're wanting to speed something along (not necessarily a great idea in itself, but if you're determined to do so... then using a library vs. an engine that handles most of that work for you, is kind of counter-productive.

I am learning Unity3D and am fond of it so far. Once I've got a good grasp of the game development process and feel I'm suited to take on larger tasks, I'll jump into coding with an API. One step at a time, though.

Which tool should I use to speed up the developing process? Or is it better to have a frest start from a graphic API? I have very little knowledge in graphic programming though.

I feel this question answers itself... If you're wanting to speed something along (not necessarily a great idea in itself, but if you're determined to do so... then using a library vs. an engine that handles most of that work for you, is kind of counter-productive.

I am learning Unity3D and am fond of it so far. Once I've got a good grasp of the game development process and feel I'm suited to take on larger tasks, I'll jump into coding with an API. One step at a time, though.

I'm pretty familar with Unity, but for my project, Unity seems does not fit for this kind of project. Many important fuctions like LOD and mesh deformation require either commercial plug-in or pro version.

I doesn't mind to learn from scratch. But if there is already a solution out there, I rather take the shortcut. I hope people read the description of my project and recommend several tools to start with.

Many important fuctions like LOD and mesh deformation

I don't think those sound like "important functions". They sound like something that would be great additions to your game for performance and eye candy, but you could still make a game with Unity and purchase a plugin or the pro version later. You can deform meshes with code, rather then some plug in. So the only thing left out is possibly LOD. Couldn't you make the bulk of your game first, then implement LOD later?

Maybe that is bad advice, I have never worried about LOD because of the scale of my projects doesnt require it. It just seems you want AAA results without having AAA skills. My advice is scope your project down and use the tools in your budget to fit that. i.e. if you budget is $0, then use something like Unity.

Good luck, it sounds like a cool project!

well - you definitely want to use an engine (unity is a good choice) if you want this project to take less than years to complete - honestly though I think you have too many features in mind right now.. This scope of project could take a team of professionals a couple years to complete

I'm not saying you can't do it I'm just saying - sometimes the excitement of being able to put whatever we want in to a game makes us a bit over zealous!

well - you definitely want to use an engine (unity is a good choice) if you want this project to take less than years to complete - honestly though I think you have too many features in mind right now.. This scope of project could take a team of professionals a couple years to complete

I'm not saying you can't do it I'm just saying - sometimes the excitement of being able to put whatever we want in to a game makes us a bit over zealous!

Indeed, I admit that my plan seem to be overwhelming for a hobbyist. My main goal is to set up a gameplay frame for this game, other asset such as Graphics and Audio can be added later.

In your opinion, how should I reduce the scope of this project? The core gameplay of this game is a text base RPG with dynamic AI (AI programming is my major in college) combining with a very simple RTS elements. Other feature can be withheld until later, athough I really want to include a customable 3D portrait in my game.

If I were you - I would do what some call a "vertical slice" - take some of the most basic elements you want and create a single town(if your doing the RTS style I'm guessing you will be using the world map system that selects different towns) that is playable with all of these systems - then expand... I would also hold off on the multi-player ideas until later

for example - you want to have each army represented as a single mesh, terrain that gives certain bonuses depending on terrain type, buildings that build certain armies, and a governing AI that competes against you

Build one small town that has all of these elements - start with simple versions and make them more complex as you go

from there you will have a much better idea on how far you can take the game - or maybe if you would like some others to join you at that point