Because I am an incredible organized person (Stop laughing) . I have made a deployment plan. There are three tracks Programming, Art, and Sound. I want to do most of the programming my self. I am capable of doing the art, but I know TIGSource is has much better artiest than I. Some of whom could lend a hand maybe? I also will not turn away offers to help with the sound.

Update 11/15/12: The very talented artist Sam (aka sonjool) is lending his talents to the cause. Look forward to seeing his work in the game.

Programming tasks:
Camera tracking (done)
Controllable bike (done)
Player control of bike (done)
Bike to ball interaction (done)
Bike physics bounce off walls (done)
GUI indicator of ball (done)
AI control of bike (done)
GUI indicator of other bikes(done)
Bike on bike interaction (done)
Lance/riding animation (deprecated)
Player control of lance (done)
Lance pickup ball (done)
Lance throw ball (done)
Lance kick ball (done)
Lance kick bike (done)
Ragdoll death activation (done)
dead biker reset (done)
Ball to goal interaction and Score tracking GUI (done)
Post goal game reset (done)
AI control of lance (deprecated)
AI react to ball carrier (deprecated)
End game conditions (done)
Main menu (done)
End game statistics (done)
AI mech goalie (deprecated)
Goalie interaction with ball (deprecated)
Goalie interaction with Enemy Jet bike. (deprecated)
Split screen 2 player VS (deprecated)
Split screen 2 player Coop (deprecated)

Art assets needed:
Jet bike mesh  (done by Son Joo)
Jet bike texture Blue team (done by Son Joo)
Jet bike texture Red team (done by Son Joo)
Arena mesh (done by Son Joo)
Arena texture Blue team (deprecated)
Arena texture Red team (deprecated)
Driver Mesh (done)
Driver Rigged (done)
Driver pose Driving, Pull Back, and Strike (code will tween)
Driver texture Blue team (done by Son Joo)
Driver texture Red team (done by Son Joo)
Main Menu Logo graphic (done by Son Joo)
Main Menu buttons for: single player, split screen VS, split screen Coop, quit.
Game over screen graphic (done by Son Joo)
Winner graphic (done by Son Joo)
Looser graphic (done by Son Joo)
Icons for Goal, Kill, and Death stats. (deprecated)
Goalie Mech mesh (deprecated)
Goalie Mech Rigged (deprecated)
Goalie Mech texture Blue team (deprecated)
Goalie Mech texture Red team (deprecated)
Goalie Mech animation: shuffle left, shuffle right, punch, grab and throw ball. (deprecated)
Other assets that would liven up the arena (artists discretion)

Sound assets:
Jet bike engine
Lance charging (deprecated)
Lance shooting (deprecated)
Ball bounce
Lance on ball strike
Lance on player strike
Spectator ambient
Spectator cheering
Announcer voice for events: Intro, Player Kill, Goal, Idle banter  (deprecated)
Mech move sounds (deprecated)
Mech Catch sound (deprecated)
Mech Punch sound (deprecated)
Additional Ambiance

Contest Version Here:
http://www.monkeydev.com/unity/lance/index.html

This Game was made with the assistance of:
Sam (Joo) Son: 2D and 3D Art
Beto Cabral: Audio

Against my better judgment I am throwing my battered helmet into the ring. My game idea is reinventing the sport of polo with jet bikes and power lances. Oh and carnage, lots and lots of rag doll style carnage. The current game design looks like this:
 
There are two teams of jet bike riders which spawn at opposite ends of the arena with a ball in the center.
The jet bike riders have power lances which they can use to punch whatever is in front of them. Power lances must be charged for half a second to two seconds in order to deliver their punch. Charging the lance prevents the bike from accelerating (still will coast).
If the ball is hit with a power lance it flies away at greater speed than if it were just hit with the bike. If jet bike rider is hit with a power lance the rider is killed. Both rider and the bike go rag doll for a few seconds before being put back in their spawn point. Aside from time lost and relocation there is no penalty for death.
If the ball is put into a goal there will be much fan fair and the both teams will have seconds to inflict closing brutalities on each other before the arena is reset.
Winner is first team to 5 or when 10 minute timer runs out.
There may be a lumbering battle mech acting as goalie time permitting.

The player will control one jet bike rider from a 3rd person POV. A top bar radar will inform them where the ball and other players are relative to them. Controls will be left/right, accelerate/break, and charge lance(on release fires lance charge). Input can be from keyboard or game pad. There may be multi-player split screen time permitting time permitting.

Here the latest screen shot of what I have done so far:


Here is a link to the unity web player of my progress:
http://www.monkeydev.com/unity/lance/index.html

I am open to collaboration from 2d/3d artists. I am keeping the place holder graphics intentionally basic for now. I can generate my own art, but would rather spend my time coding. Reply to this thread or PM me if you are interested. Also see the list of art assets.

Old Screen Shot:

I think the term you are looking for is "Relative" vs "Absolute" movement.

Relative movement is movement relative to the camera. When the user presses left the avatar will move towards the left of the screen regardless of which way the character is oriented.

Absolute movement is movement based on the character's current orientation. Here pressing left will rotate the avatar counter clockwise. Pressing up will move the character forward along the direction they are currently facing.

I have two ways I like to do random movement:
1) Pick a direction and move in that direction until a timer elapses, then pick another direction and reset the timer.
2) Pick a point in space a random offset from the agent. Move towards that point until the distance is less than a preset value. Then choose another point. Sometimes I combine this with a timer to prevent the agent from getting stuck on terrain.

It looks like you are doing the former. Your code looks like it should work. Except

direction = ( (int)rand() % 4 + 0);

will give you a value between 0 and 3. you should start your direction to zero or add 1.

direction = ( (int)rand() % 4 + 1);

Will you implement a supply chain?
One planet consumes basic goods and produces raw materials at a low price,
One planet consumes raw materials and produces basic and luxury goods at a low price,
One planet consumes luxury goods at a high price.

Even if it is crap, finishing a game is very important. There are things you learn as you wrap up a project that you don't learn from abandoning it.
Stuff like:
The value of outside play testing
How your impression of difficulty differs from that of the average player
What coding practices lead to the most headaches in late development
What bugs are OK to ship with
Which features to cut if the deadline is too close
When to stop adding features and start polishing

Are you going to implement a supply chain and illegal vs legal cargo?

basic food stuff
drinkable water
luxury food stuff
basic alcohol
luxury alcohol
medicine
illegal drugs
livestock
exotic animals
radioactive materials
raw ore
processed metals
construction material
vehicles
luxury vehicles
robots
weapons
fuel
premium fuel

Game development is not for everybody. The good news is there are many people out there making games. One day one of them will create the game you wanted to see, just be nice and pay them for their effort.

As for your constipated friend Jack, I am glad you can help him.

I just remembered Cortex Commander does the 2d ant farm thing quite well. Of course each soldier has a jet pack which makes navigating vertical tunnels easier. You can still do ant farm style with humanoids if you have ladders or stairs as part of the background where the level can be navigated vertically. You will have the same problem in 3d, finding a way to depict your dudes traversing vertical shafts.

I never played dwarf fortress, but I did enjoy Dungeon Keeper quite a bit. Strangely enough I was also reminded of Sim Ant from way long ago. 3d might be an interesting approach for this design, but it may be too much for the average player's mind to fully take advantage of.

You could do this 2d side view like an ant farm.

I had a different experience with the multi player Super Mario which lead me to conclude that platforming is something best left to one player avatar. There is plenty of room for asymmetric game play where one person is doing the platforming and the other is assisting in some way. Shooting enemies with stars in Mario galaxy is an example.

As for making a coop experience were players rely on each other, Left 4 Dead set the high water mark in that area. Many of its systems are explicitly set up to make helping your fellow survivor easier than helping yourself.

Oh crap, I guess I'm missing out.  Time to go play Farmville and fill my library with wiggley waggley Wii party games!

Try playing Europa Universalis to unwind. There is good to be had in simple and complex games. But it is easier to mess up the design for complex ones.

Richard Kain's post got me thinking:

Wouldn't it be cool if League of Legends (or similar arena based ftp/pay for skins game) made custom skins from its top tier pro players? Something that would reflect that player's personality, and be based on input from the player. Fans of that player could buy the skin and use it in their games much like football fans wear their favorite player's jersey.

It is probably too much work on the development team's side to make this a reality, but wouldn't it be cool?

I do use Unity, it is awesome. I still retain practices from when I worked on my own engines. The I make a static play balance object where I put the world defining variables. That way I don't have to hunt though all the objects in my scene to find the major game play variable that needs tweaking.

I am all for iterative development but some features will only make sense when tested in relation to other features. For example you don't know how long a player should be invulnerable after being hit by an enemy until after the enemies are implemented. Yet the movement rate, attack speed, and stun lock all interact with this value and may be implemented at different stages in development.

As stated in my previous post I put something in this field to make the game playable, but I hold off on finalizing the value until all the systems that interact with it are also working.

Play balancing variables is one of those things that needs to be done regardless of how good or bad the core rules of the game are. Often you won't know for sure if the core rules are any good until the variables have had some tweaking. It always struck me as a waste of resources making the programmer do this tweaking, but that could be me wanting to skip the boring part of game development in favor of the more fun parts.

I like the right better, in part because the icons aren't obscured with buttons. I am worried that users would be put off by putting a point into something without knowing what it does. The one on the right does a better job showing the options available.

I like buhbuhcuh's explanation of the need for math in game a designer.

I think that higher level math is useful if not required for game design because it helps you understand the relationship between numbers. For example a small change in the value for air resistance can have a larger effect on game play than a big change on initial jump velocity. I can't say I have ever worked with a good designer on a game, but I have suffered many idea guys.

When I am programming a system for a game I find there are numbers that affect how that system preforms. A change to any of these numbers will alter the "feel" of the game. I used to hard code these numbers and fiddle with them in the code until it produced the "feel" I wanted. Now a days I assign them to a variable that is loaded from a separate file at load time. I make sure the variable has a default value that will keep the game from crashing, but leave the calibration until I have the entire system programed.

Since I solo most projects I also do the calibration of these values. I have tried handing this off to a "designer" in the past, but the results were less than satisfactory. I explained what each value modified and even gave some suggestions on which variables to modify to get a certain effect.

Is it unreasonable to hand a designer a working game demo and a file with 20+ variables, then tell him to "Play around with these until it feels right"?

Have you tried fiddling with the material settings on the object once it is in Unity? I have had blender to Unity imports go wonkey in a different way before. It sorted itself out after I applied a new material over the default one generated by the blender import.

They posted on some of the other game development pages I frequent. Viope solutions is some sort of web training company. I guess they are doing this for the publicity. I read their contest fine print and did not see anything too draconian about it. Still I don't think I will participate in this one. I do game jams for the community experience, and I don't see much of a community around this one.

Your artists know their stuff. I really like the colonized Jupiter in the background.