Mechmania Python Starter Pack

Here's all the code you need to get started with making a bot for Mechmania in Python. Just do these steps:

• Pre-Setup -- install Docker, Node, and the mm command line tools
• Setup -- Clone this repository and start running your bot!

Windows Pre-Setup

1. First, install Node. To do this, go here and download the Windows Installer.

2. Next, install Docker Toolbox for Windows.

   • Go to this site and click the button that says "Get Docker Toolbox for Windows".
   • Open the installer and follow the instructions to install Docker
   • If you don't already have Oracle VM VirtualBox, you will also go through a series of prompts to install VirtualBox.
   • Once Docker is installed, open the "Docker Quickstart Terminal" application (It will take a few minutes to get set up the first time you start it. This process may involve more Windows prompts for permissions.)

3. Within Docker, run npm install -g mechmania. This gets the mm command line tools, which are used to test and submit bots for the tournament.

Mac Pre-Setup

1. First, install Node. To do this, go here and download the macOS Installer.

2. Next, install Docker for Mac.

   • You can find the installer here.
   • Open the installer and follow the instructions to install Docker

3. Run npm install -g mechmania. This gets the mm command line tools, which are used to test and submit bots for the tournament.

Setup

1. Clone this repo (or fork it or download it somewhere as a ZIP)
2. Modify the script at MyBot.py.
   • Write your code after the YOUR CODE BELOW comment, making sure to call game.submit_decision() once per turn.
   • You may also add other files or dependencies, but if you do, make sure to update the .dockerignore and Dockerfiles accordingly. If you have any questions about this, we're here to help!
3. Run mm play .
   • This will build the bot in the given directory (.) and then starts a game in which your bot fights against itself.
4. To run two different bots against each other, run mm play bot1_directory bot2_directory.
   • You must have valid .dockerignore and Dockerfile files in each directory. We reccomend cloning this repository multiple times to achieve this.

Use mm help for more information!

Game_API Information

Within the file Game_API.py, there are a variety of helpful functions for writing your bot. You can use the game object in the MyBot.py script to run any of these functions (e.g. game.get_turn_num()).

IMPORTANT: Do NOT use print() for debugging, as this will make your bot fail when the game runs. Instead, use the game.log() function in Game_API.py.

Within Game_API, three other classes are defined -- Player, Monster, and DeathEffects. We recommend only accessing the fields in these classes, since changing them might cause the game object to not properly represent the current state of the game.

The following fields are in the Player class:

• player_num : int (either 1 or 2)
• name : str (either "Player1" or "Player2")
• stance : str -- string representation of the player's stance -- "Rock", "Paper", or "Scissors" (Note: on the first turn, each player's stance will be "Invalid Stance", since neither player has yet chosen a stance)
• health : int -- the player's health. If a player's health reaches 0, they die.
• speed : int -- the player's speed.
• movement_counter : int -- the movement counter will go down by 1 each turn. Once a player's movement counter is equal to their speed, they will move to their destination.
• location : int
• destination : int (Note: on the first turn, the player's destination will be -1, since the player hasn't yet set a destination)
• dead : bool (This should always be False, since once a player dies, the game is over)
• rock : int -- the player's Rock attack stat
• paper : int -- the player's Paper attack stat
• scissors : int -- the player's Scissors attack stat

The following fields are in the Monster class:

• name : str -- represents the class of this monster. All monsters with the same name should have the same base stats (health, attack, respawn_rate, death_effects, and attack)
• stance : str
• health : int
• respawn_rate : int -- number of turns for the monster to respawn after dying
• respawn_counter : int -- turns until this monster will respawn (Note: if dead=False, then you should ignore the value of this field.)
• location : int
• dead : bool
• death_effects : DeathEffects -- Gives information on the buffs given to the player when this monster dies on the same node as them.
• attack : int -- the amount of damage the monster deals per turn
• base_health : int -- the health that the monster will have after respawning

The following fields are in the DeathEffects class.

Each field corresponds to the buff provided to a player's stat from defeating a monster:

• rock : int
• paper : int
• scissors : int
• health : int
• speed : int

The following are all functions in Game_API

(Note: For any function that returns Player, Monster, or DeathEffects structs, the data in the structs may change between one turn and another, so you should make sure to only use structs returned during the current turn)

log(strn) Logs the string str to stderr for debugging.

get_duel_turn_num() Returns the turn number when the duel will occur

get_adjacent_nodes(node) Takes an int node and returns a list of ints representing the nodes adjacent to node.

get_all_monsters() Returns a list of all the monsters in the game.

get_self() Returns a Player object representing the player you are controlling.

get_opponent() Returns a Player object representing your opponent. (Note: the opponent's destination will always appear to be -1. This is by design -- neither player can see the other's destination)

submit_decision(destination, stance) Takes an int destination and a string stance and sends this decision to the game engine. You should call this method exactly once per turn.

shortest_paths(start, end) Returns a list of all shortest paths between start and end nodes (passed in as ints). Each element of the returned list will be a list of integers, representing the steps required to follow the path to end. These lists will include end, but will not include start unless start==end.

So, as a hypothetical example (that doesn't necessarily match the game map), shortest_paths(1,5) could return the following:

[[2, 3, 4, 5]
[2, 6, 7, 5]
[9, 8, 7, 5]]

In this example, there are 3 paths, each of length 4, that can get from node 1 to node 5.

has_monster(node) Returns a bool indicating whether there is a monster at node node.

get_monster(node) Returns a Monster struct for the monster located at node. It is reccomended to check if there is a monster at node first, using has_monster(node). If you call get_monster for a node without a monster, an invalid monster will be returned.

nearest_monsters(node, search_type) Returns a list of Monster structs for all monsters nearest to node (including on node itself), only considering monsters according to the search_mode parameter:

• search_mode = 0: Searches all monsters
• search_mode = 1: Only searches for live monsters
• search_mode = 2: Only searches for dead monsters

nearest_monsters_with_name(node, name, search_type) Same as nearest_monsters, but only considers monsters with name name. (Note: since, on our map, there is only one monster with any given name, the list returned by this method will contain a maximum of 1 monster)