Looking into how a person makes a decision in high stakes situations has always inspired me to create something that could think of the best way to react to such a decision

What it does

It uses the MSTank API to create a tank and navigate it through the stadium. It kills other tanks and retrieves points by moving to the goal posts

How I built it

I thought of the situations the tank would be in and give it instructions on what would be the best answer and allowed it to generalise these solutions for the whole game

Challenges I ran into

Trying to get meaningful data from the environment. Most of the data given was rudimentary and it took a lot of time to create a substancial understanding of the surrounding of my bot at a given time

Accomplishments that I'm proud of

Being able to create a hard-coded AI which can take on multiple opponents with more information and still be able to come on top

What I learned

How to slowly build a decision tree using if statements and parse meaningful information from primitive inputs

What's next for ThinkTank

Trying to account for more difficult situations and have a broader understanding of the landscape it interacts with on a holistic basis

Built With

Share this project: