A few of you have looked at my previous game engine, which i have put on hold for several reasons. The largest of which was finding the style of game i was working to build has been done to death.
Firstly ill touch on the confusion part.
There are an awful lot of tutorials and guides out there that are very helpful, however a large amount of them seem to be based on 2.0 rather then 3.0 or 3.1 which doesnt sound too much of a problem until you get down to the brass tacks. As most of them refer to coding and what the code is doing, rather then the concepts. I know that i didnt explain that very well, but im sure i am not the only person who ever have tripped over the compatability issues between the versions. and im sure that the more familiar with XNA i become, the more i will be able to compensate, but the guides im looking at are mostly beginners guides.
The reason i made this post though is that i am looking for some input. My original idea was a third person shooter, on an open world setting with specific missions based around a plot that evolves as you move through. Yes i know not very original but then its pretty hard in this day and age to do something that nobody has ever done.
Im now a little stuck on deciding what genre to aim for now. i like the idea of rpg but to be fair that coding would be a touch too long winded for my tastes at present.
not to mention i am hampered by a lack of modelling skill or any good sources of animated models. please before anyone starts spamming me with turbo squid adverts and other such websites, ive been on most of them, and their .fbx and .x libraries are still pretty limited, and i dont fancy paying the prices on custom made models, particularly when the theme of the game hasnt been decided.

I do have a basic understanding of tS7.6 however nothing like the kind of ability that is really required.

so im looking for artistic input as much as i am conceptual.
If anyone has any animated fbx or x files then they are always appreciated, nothing used in anything that would be released unless there are any bored artists out there. purely stuff for testing.
but if anyone has any interesting game ideas in the meantime, im always open to suggestions.

I suggest making a game of a type you'd enjoy playing. If you don't have 3d modeling skills, use 2d graphics or just abstract shapes. If you really have no ideas, try looking at themes used for 48-hour game compos, and run with one or more of those.

Your all over the place, a game usually starts with an idea of a mechanic or vision of some sort. Your worrying about problems such as art direction to far in advance.

Some people travel halfway across the world only to find the inspiration they where looking for in their own back gardens.Try picking up a book, looking at art, following tutorials.

Are you more interested in plot and narrative then game play? Then try and make more of an interactive story to begin with.

As for a theme suggestion...try thinking of black and white?

I'm in agreement with Agent Oil.

What you should do is come up with a game mechanism you think would be enjoyable, describing it in detail
Make it into a prototype, see if the idea works and can be fun.
THEN make a game idea around the mechanism, planning it out in detail before laying down code, look over it as you do and make sure you don't overburden yourself (this is my prime error! Many a project scrapped due to my biting off more than I can chew).
Create a supportive code to implement the game mechanism along with actual game.
LOOP BEGIN
Create Content (this is probably where you'll spend the majority of your time).
Test Content
LOOP FOR HOW MUCH CONTENT YOU ARE MAKING
Overall tests
Beta test among some population of testers
Polish until you can sense a slight glow.
And then release the damn thing already.

Although perhaps I'm not the one to give advice on this, I have plenty of projects that got to the create content and test loop but never quite got though it without running into issues with the code I had written, or ideas of what was fun and not fun, etc... but there are my two cents, for whatever they are worth.

Graphic content is a major problem. If you don't have the graphics, you are pretty well stuffed these days.

However that does not have to be the end of your coding.

Spend half a day looking around at special effects. Have a look at shaders, particle systems, water shaders, all that sort of code.

View it from the point of view of "could I make a game out of that?"

It is quite possibly to make a game only using textured spheres. If the game play is good enough, people will love it. The gameplay has to be superb though.

There is one game out with NO graphics, it's all sound based.

I remember a game that only used a linedraw, you had to build bridges and test them.

Jeff Minter released a game that was just an interactive music visualiser, and it was awesome.

Graphics are the not THE limiting factor, the final limiting factor is your imagination.

As most things has been said, i still give my 2 cents ;-)

Everything begins with an idea or concept. You are the creator, so you must think up an idea which you want to have visualized as a game.
For example Your idea can be:
- an innovative gameplay concept
- has an original story background
- A game which yourself always wanted to play
- A game with artisic graphics
- etc..etc..

The first step is the concept and idea, think up all the things the player should be able to do in your game, and what the global layout of your game would be
(2D Horizontal platformer, with a jumping running plumber who want to save a princess across levels and picking up coins along the way etc..)

the second step normally is concept art and character design. So you have a visual feel of the style
(A plumber with a moustache wearing a red overall and hat, running through colored worlds etc..)

Parrallel with step two is the game engine and structure design.
(2D tile engine for world visualization, with a horizontal scrolling background, patrolling enemies and a start and an end point etc..)

Step 3 is the actual programming of the game
(Bring all your ideas together in some sort or engine, framework, or just start to code; step by step to bring your ideas to life)

As said by other repliers, adjust your concept and ideas to your own capabilities and expectancies.
So if you don't have outstanding 3D modelling skills at your disposal, think again about your game concept if you wanted to create the next gen 3D First Person Shooter.
Except if you calculated a whole bunch of extra months of development time to learn the needed skills  (3D modelling skills, 3D programming techniques etc)

And there I mentioned development time, many people have great ideas and big concepts, and some have also the required skills to get the job done, but their ideas are so grand and contains so much functionality, that it ends up as a never ending project, and eventually in a game which never sees the daylight..

That's why you should combine all these game development areas and come up with an idea which matches your level of expertise and still contains the parts which you want to see back in a game.

You can ask here if you have issue with tutorials that are 2.0 and don't convert well to 3.0 or 3.1... personally the only real issues I've seen is trying to convert older tutorials from pre-2.0 version... 3.0 wasn't a very drastic change.

While you're just learning XNA, I would keep the scope of your game small... a wide open first person shooter is going to be hard to make while learning. It might be easier to start with a 2D game as it's one less dimension to worry about and you don't have to get into anything reguarding modeling/ect. Pick a core concept you enjoy playing and make that. Keep it small enough that you can reasonably finish your game. If that goes well, it will give you an idea of the scope of games you can feesibly create and give you a better idea for when you're really ready to make your masterpiece.

I appreciate all the replies guys and girls. I will take a lot of it on board. My initial problem was that i wanted to make a tps/rpg based on a storyline that i had in my head. It was going well, until i played a lot of xbox games and realised that if i wanted to be original, i was going to struggle to tell my story without being accused of plagerism. Anyone who has downloaded my "Engine" for the game, which i didnt finish before i came to my crisis of creativity will notice that i have no problem with coding per-se so much as im terrible at module based programming. I tend to get carried away coding and end up with one massive code file instead of lots of smaller modules, and then i cant figure out how to split it up in to more maneagable chunks. But this is not really my OP.
I still plan to use my plot idea, im just having to rethink the way in which i tell it. I want to try and steer clear of 2D stuff as really my coding abilities arent the issue. I am still learning but i like to think ive moved past the 2d level and am capable of coding the 3d stuff. Besides most of it is just using vector 3 instead of vector 2 for placing things in the game world. I have tinkered with some basic physics of my own, as i dislike using other peoples work, more because i like to know what every line of code does and what i meant to use it for when i wrote it.
my main issue really is my lack of modelling skill, i can sit with a piece of paper and sketch away my ideas, just really bad at doing it with a mouse or tablet.
i think i need to spend less time coding and more time practising creating skinned models, and adding motion captures to them in a way that doesnt make them walk like a wierd retarded fairy.

A game must be designed with it's presentation in mind. Alot of great games do not translate between 2D and 3D. Alot of games were made in one, but designed for the other, such that they suffered. Making a game in 2D is just as much an accomplishment as making one in 3D. There are strengths and weaknesses with developing for both. Don't so lightly toss 2D aside because you are "past" 2D. 2D is not adolescence. It's not just a phase in game development careers that is to be quickly moved past.

All that said, I honestly don't think you are ready for 3D. Forget character models, you have to develop lots of content to fill your space, tens or hundreds of models. And then if your game play itself is 3D you'll have to make sure that they interact with the player right across all dimensions, and you'll have more beta testing woes as a result of the untold number of test cases.

If you haven't noticed, I am a 2D advocate, and have had my share of cuts and bruises trying to make 3D games. 3D is nice, but 2D is more fun. That and hey you can always mix technologies. I loved how Gish has 2d game play and graphics, but they implement 3d lighting and other effects to polish their look.

I feel i have to apologise for making light of 2d games. It wasnt my intention to say that 2d was in any way less fun or demanding. But basically, 3d is more difficult to code as a general rule, 2d is still tricky, for example AI still needs to be effective in 2d, and if you are coding a 2d rpg then yes its still going to take a lot of work. But i am more referring to a graphical perspective. I can sit with a pencil and pad and sketch as well as the next man, but when it comes to using a mouse or tablet im not better at 2d then 3d. AS far as codeing goes im quite ready, im more then capable of using hundreds of models in one scene and manipulating each one as required. The only things im working on getting better at are working with skinned models, as documentation on the subject is relatively limited. By which i mean that the skinning sample on here is fine, but i was attempting to move across to a custom content library that handles more then one animation per model and found that the documentation was very vague. Im now reading up on making my own custom content handler so i can work with my own methods in mind.

THe only thing that is holding me back is slight inaccuracies in tutorials, 99% of which are just typos that get fixed quickly enough, and my lack of ability in modelling. As much as i appreciate your perspective i have no intention of throwing in the towel and giving up on 3d because i suck at making models. There are plenty of people out there who have suffered the same problems and just learned to model. What i was asking for was if anyone had any little skinned models that i could use in the mean time, so that i can still code, and just replace the models as a last stage. I understand that my OP was sort of lost in translation, mainly because my original game concept became a bit of a problem once i realised that i was going to face the firing squad, by which i mean people who revel in being able to point out similarities between one story or game and another.

Either way, thanks for the cunstructive advise guys

Here are a few tips I'll just throw out there, some you didn't quite ask for others you did.

Model Animation:
The key to animation is understanding the file format that you are looking at, since xna provides little out of the box animation loading. Specifically most FBX files out there, have all their animations put into a single "Take" which is a stream of transform information for the models bones. The 3d programs add specialized tags in the FBX Take to parse out individual animations, you'd either have to tap into that knowledge know how to look for thoes tags and parse the animation, or get a animation program that when it outputs to an FBX you know it generated multiple takes. The animation information itself consists of a list of transforms each having a time attribute for each bone of the model. Elsewhere in the file, and I believe XNA does this for you, is a list of the bones in the model, each bone keeps track of it's children bones, and what vertexes it's transforms affect as well as the weights of that affect. Animation is simply cycling through bones, cycling through their vertexes, applying the bones transform multiplies by the bone's vertex's weight to the vertex's orientation relative to the bones origin. I could go on and on about this, but understandably it's a rather complex ordeal to keep track and apply all this information. The skinning example uses its shader to apply the actual animation to the model. This should be all the information you need to get you started. Of course, you'll need to do some digging elsewhere to get the FULL picture (like exactly how to get the animations and bone information and such), but this is a draft.

Reguarding multiple identical models (regardless of their animation state):
implement a way to group them in instances, and develop your shaders used to draw models to batch draw multiple instances of the model at multiple locations on the screen preforming the transforms on in your animation data for that draw. This will increase your frame rate greatly. A very basic implementation is available on ziggyware if I remember correctly, however it is only good for unanimated props (they used cubes). However the same methodology can be used to display multiple identical bodies with animations however depending on the shader model and the complexity of the animations you may reach your limit of passing matrix transforms rather quickly.

Also just another tip that I don't know if you're doing or not,
The garbage collector is a fickle cpu hog. do youself a favor and implement a GarbageManager, what you can do is when you want to release a load of objects, send them to the garbagemanager, and the garbagemanager should be programed such that over time is slowly releases objects in such a way that the garbage collector doesn't overwhelm your frame rates. Also there is the reuse of garbage, if you need a new resource, and it can be reinstanced from an old one in the garbage manager, do that. You'll save the time it takes to make the new object and the time it takes to clean up the old object.

Tons of more ideas, though I'm not sure if I'm assuming too little of you or not, I'll just stop there, as these are semi relavent to managing large data sets like animation information and such.

Thanks for the useful info. although the animation stuff sounds like more work then previously anticipated so i might stick to one of the available libraries but the large data set stuff will come in handy. :)

Glad to help you make an informed decision even if it was to give up. XD

Seriously though, the programming isn't the hard part of any of these undertakings; programming is very straight forward it just takes time. The only real challenge is understanding the subject matter which you are trying to programming.

Taking on a library however, isn't a cop out. I think if there is a good pre-built free to use solution that meets your needs, there is no reason to rebuild the wheel. Good luck.

I know you've probably already decided on using a library, however if you are interested in how to animate multiple instances of a model get the DxSDK installed on your computer, there is a tutorial project called "MultiAnimation" which goes over an approach similar to what I outlined for having animated instances of a boned model.