Title:
Kung Fu Tamagotchi

Tagline:
Mr Myagi Evolved

Genre:
RPG & Fighter.

Mechanics:
Tamagotchi meets Soul Calibur

Setting:
Medieval meets Futuristic

Target Audience:
“soccer mom ”, Japanese teenager

Goal:
Create the ultimate fighting machine.

Description:
You get to train a young budding fighter how to fight. You get to make your own,
original PHYSICS BASED MOVES. You get to pit your Kung Fu Tamagotchi
against your friend’s to show them who the real Mr. Myagi is. The game makes use
of ragdoll physics so the fighters act as marionettes with pull points that
the user can select and move accordingly to his/her desired movement.

There two parts to the game:
Move Editor:
This is where the user can create and modify the moves for each player.
To create your own moves:
-Select pull points with the mouse
-Use the sliders to change their position(x,y,z)
-Add, Delete and Play moves
-For each move, Add, Delete or Change poses
-Load and Save moves to player

Fighting/Training Mode:
This portionof the game is similar to other fighting games in that you have two
players, the user's custom player and the opponent. The purpose of this mode
is to teach your marionette player to fight an opponent who learns from your
actions. The marionette performs the moves your created and the opponent performs
moves that were also custom and specific to that player. The user may call out moves
such as "BACKHAND" or "KICK" or "PUNCH" into a microphone or use the keyboard
to tell the marionette what to do. The controls are as follows:
Game Controls:
-At the beginning to you load two players
-List of available moves is on the right side of the screen
-Speak into the microphone to select a move
-After each round, press SPACEBAR or say "Next Round" to continue
-After each match, press ENTER or say "New Match" to start a new game

Artifical Intelligence:
The Artificial Intelligence used in Kung Foo Tamagotchi is a reinforcement learning
based algorithm. It is highly configurable in its ability to take different types of
game state information on which to learn on. One typical problem with seen with standard
reinforcement algorithms is the unmanageable size of the Q-Table which is partially avoided
in this implementation through the use of a Q-Tree. The Q-Tree is a decision tree where the
decision variables correspond to different game state variables. So we can construct a Q-Tree
by first examining the `Distance to the Opponent`, and then branching on the `Opponent's current
move` to arrive at a Q-State. The benefit of this approach is that we can store Q-States as we
see them (saving memory), and have the possibility of dynamically aggregating similar states together.

In terms of the AI learning process, reinforcement is given both by the environment as well
as by the user. The AI is positively rewarded for successfully landing an attack on its opponent,
and negatively reinforced for being hit automatically. Therefore in theory, the AI can learn on its
own without user interaction. The problem with this unguided approach is that Q-Learning takes a long
time to fully and accurately explore the state/action space. The user therefore comes in to shortcut
the exploration process by showing the AI what do to. The user can take two routes in this, the first
is to actually dictate the AI's next move, and the second is to reinforce the AI's last action. With
both automatic environment learning and user guided learning, we think our AI will provide new and
interesting results.



It is innovative because...
It’s a physics based fighting simulation where your trainee learns from you and
from experience in fights. You get to make your own moves which can be modified
and perfected. There is also the possibility of having the trainee make up his/her
own moves.

In the most recent versions of popular fighting games like Soul Calibur, Tekken, and Dead or Alive
you can select the types of weapons a player can use, the stage, and costumes. This type of customization
and personalization provides users with the ability to change the look and feel of players and fights but
only on a very high and imprecise level.

While these versions of recent fighting games give users a great deal of freedom, users want more control of
the actual game play and want to be able to customize the moves and skills of a player instead of using the
same pre-packaged animations over and over. This has inspired the creation of fighting games like Virtual
Fighter with different modes where you can train your own players by sparring to endow them with new moves,
and by judging their abilities in replay.

Our game makes use of real-time physical simulation with ragdoll physics which is a type of procedural animation
that has been traditionally used to replace static death animations prepackaged in most video games. The
Training/Fighting Mode is also unique and our AI provides the ability for players to play against the opponent
on their own and/or receive reinforcement from the user and create better winning strategies while performing
unique moves custom built for each player.