Fine. Snake 🐍

team : Are You Okay?

• Catherine Meng
• Chloe Yu
• Licht Liu
• Olivia Zhu

Our strat:

The Strategy of our Fine snake is based on DFS searching algorithm. We want to make sure our snake could always find a safe path from its head to tail on its way to the food. It turns out that the Fine snake performs outstanding when it runs alone on the board (achieved 3000+ turns on a 11x11 board), or during 1v1 battle with one other snake. During the BattleSnake 2019 event, we won 3 Bounty Snakes in total (snakes from Giftbit, TELMEDIQ, and Semaphore).

Basically, based on the prediction of longer snakes' next steps, we have 3 DFS searching mode in totall. Mode C is the danger mode, in which we consider all the cells without considering other snakes' next steps. Mode B is the normal mode, in this case we consider longer snakes' next one step ahead, we'll try to avoid those cells when we calculate the path. Mode A is the safe mode, where we consider longer snakes' next two steps ahead.

The DFS is used for three purposes: 1. find the path from head to tail. 2. find the path from head to food. 3. find the path from food to tail. In order to keep our snake safe, we have to assure it can always find a path to its tail. For instance, when chasing a food, we run DFS from the head to that food, if the path exists, then run DFS to find the path from food to tail, if both paths exist, then go chase that food...

The pseudo code look like this:

# if num_of_snakes > 4:
#     if health > 90:
#         	chase tail (mode B)
#		if no path to tail
#			chase food (mode B)
#			if no path to food
#				hover
#     else if health < 30:
#         	chase food (mode C)
#		if no path to food
#			chase tail (mode C)
#			if no path to tail 
#				hover
#     else: 
#		chase food (mode B)
#		if no path to food
#			chase tail (mode B)
#			if no path to tail
#				hover
# else:
#     if health > 80:
#         	chase tail
#		....
#     else if len(food)/len(snake) < 1/2 and health < 40:
#         	chase food (mode C)
#		...
#     else:
#     	chase food 
#		...

Things that need to be improved:

We are good at solo and 1v1 game, but not with 7 other snakes on the board. At the very begining, since all of the snakes are in the same length, our algorithm would skip too many cells when we run mode A or B. And if we run mode C instead, we all listen to the fate, as there's no prediction at all. As most of our tests were on localhost with a tester snake, we didn't realize this problem until the last day, it could be a lot better if we can add some strategies to deal with this situation. :)

heroku url: https://ccilosnake.herokuapp.com/

This is: A Battlesnake AI written in Python.

• Docs and APIs
• 

---

Prerequests

• a working Python 2.7 development environment (getting started guide)
• pip to install Python dependencies

---

Running the Snake Locally:womans_hat:

1. Fork this repo.

2. Clone repo to your development environment:

git clone [email protected]:<your github username>/starter-snake-python.git

3. Install dependencies using pip:

pip install -r requirements.txt

4. Run local server:

python app/main.py

5. Test your snake by sending a curl to the running snake

curl -XPOST -H 'Content-Type: application/json' -d '{ "hello": "world"}' http://localhost:8080/start

Deploying to Heroku:sushi:

1. Create a new Heroku app:

heroku create [APP_NAME]

2. Deploy code to Heroku servers:

git push heroku master

3. Open Heroku app in browser:

heroku open

or visit http://APP_NAME.herokuapp.com.

4. View server logs with the heroku logs command:

heroku logs --tail

Start the engine💖

MAC OS Example:

1. Install Brew

    /usr/bin/ruby -e "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/master/install)"
   

2. Install Git

    brew install git
   

3. Install Jq

    brew install jq
   

4. Install Wget

    brew install wget
   

5. Download the Engine

    mkdir battlesnake-engine
    cd battlesnake-engine
    wget -qO- `curl -s https://api.github.com/repos/battlesnakeio/engine/releases/latest \
        | jq -r ".assets[] | select(.name) | .browser_download_url" | grep Darwin | grep 64` \
        | bsdtar -xvf-
   

6. Run the engine in dev mode

    ./engine dev
   

7. Open a browser and go to http://localhost:3010/