Posted: August 16, 2016, 5:07pm
I’m a nerd through and through. I really can’t help it. I’ve tried. There’s lots of reasons that I’m a nerd, but the one I’m thinking about right now is that I found a hobby that is the perfect representation of my work. What I mean is that my hobby and my work are practically indistinguishable.
My PhD thesis topic is math modeling of cancer progression and metastasis. In order to model the growth of tumors, we use a math modeling technique called Game Theory. This is the study of competition: during a game between two (or more) players we can model the interaction between them using the “rules” of the game. For example, I can model Rock Paper Scissors using game theory but assigning payoffs for wins and costs for losses. Player 1 receives a dollar if they play Rock at the same time that Player 2 plays Scissors. Everyone knows the rules of that game, but we can also model cancer cells as “players” in a game (a competition) with healthy cells. Sometimes the healthy cells win, sometimes the cancer cells win, and sometimes the doctors try to change the rules of the game with medicines, radiotherapy, or chemotherapy.
I realize that my research focuses on two of everyone’s least favorite subjects: math and cancer. But game theory is a very intuitive field in applied math that has many applications. One of the first to use it in a biological application was John Conway with his now-famous “Game of Life.”
In order to apply my knowledge to something less “useful” than cancer research but definitely more fun and light-hearted, I updated the Game Life to allow two players to compete under the same rules in a game that resembles a (more complicated) version of checkers. Here’s some screenshots of the app. I just submitted the app for review to be uploaded on the App Store (Apple) so I’ll edit this post when it becomes available to the public – hopefully soon!
Funny Artificial Intelligence
One of the funny parts of the game was building single-player mode. The game is designed to be a two-player competition so building out a single-player mode requires an Artificial Intelligence of some sort to play for the Computer Player. AI’s are difficult to build because they shouldn’t be too easy to beat – but also not too difficult. There’s a Goldilocks middle ground of difficulty that we shoot for- just right.
For a first start I built the AI to play a completely random move at each turn. The hilarious part is that this “dumb” AI is really hard to beat.
Published on July 25th, 2017
Last updated on August 10th, 2017