Featured Game: Trinity Overdrive Still my most involved project allowing me to direct, design, program, and balance the boss battles. Since then my skill has improved, but this game holds up as one of the few I was given the opportunity to make my own vision.
After receiving the DeLeonic Award for my work at Video Game Development club, I continue to work on projects and support the growing Nationally Acclaimed VGDev club format. I help complete up to 10 finished games a year while attending to school, my research, and my personal projects.
2 year work anniversary: My work at Georgia Tech’s IMTC Research lab continues. My position as faculty gives me on the job experience with video game R&D research projects. Currently I pursuing a thesis on Artificial Intelligence interacting with player behavior and preferences in real time.
My current and most ambitious project is titled Yomi Domini. I coordinate a team of artists for game assets while acting as Game Director, Producer, Programmer, Marketing, and Management. I am on schedule to finish the game by Spring 2016 for a PC release. The initial public demo is available here on this site.
Development Time: 2 Months
Chris Tansey: Project Lead, Lead Programmer, Technical Art
Jordan Hobgood: Music, Audio Programming, Testing
Kartik Kini: GUI Design, GUI Programming
Anna Hedden: Character Sprite Animation
Paul Loebs: Environment Sprites
Jamal Rhodes: 3D Modeler, 3D Textures, 3D Animations
Garrett Leach: Sound Design
Thomas Sulkoske: Sound Design
Cheng Hann: Level Design, Testing
Jacob Watson: Level Design
Peter Aquila: Testing
Matthew Ielusic: Programming
Yomi Domini is an ongoing project of mine. Since its public beta my team is working on a public release to expand the content within. The scheduled released is slated for year 2016.
Bison Bot is a functional Street Fighter 4 Bot that can offer a competetive agent in the game controlling M.Bison.
The bot employs a series of different features to not overwhelm the opponent with superhuman capabilities. Instead it is designed to mimick a competetive player's physical and mental capabilities. To accomplish this, the bot uses a series of different cutting-edge experimental methods from both Knowledge-Based AI and Machine Learning. Currently Bison Bot's 3 main features include a Cognitive Vector AI, a Lattice Network Neural Net, several implementations of Metacognition.
The Lattice Network is a new ANN designed to suit Bison Bot's interaction with his opponent and create inferences. The ANN resembles a Hopfield-Tank ANN that works with the game states being handled on the edges of the neurons instead of the neurons themselves. Giving the ANN a greater capability to form combos while deciding which response is correct for the right time.
The Cognitive Vector AI is a new type of Expert System that represents possible Classifications of actions made by the opponent. The vectors of each Classification work as a human brain would as if divided among different layers of attention demanding action.
Finally the Metacognition is designed to handle outside elements. Mistakes made by the inference method or the ANN can be corrected based on a series of different observations made by the bot. While observing the actions made by the opponent and observing adaptations to the fight, Bison Bot seeks to correct itself and adapt to the opponent's adaptation.
In the command line inside Moba Hero AI input: python runherocompetition.py MyHero BaselineHero3
I placed head of my graduate class in the final competetition of our MOBA AIs
A set of Artificial intelligent agents made in a MOBA environment. The MOBA engine is made from scratch in python. The agents minions are designed to form squad based strategy based on several game plans and states of the game meta (group up, attack base, defend, spread out, body block). The minions are made to form squads and squad actions to efficiently defeat the enemy squad in sight. The hero AI aims to conservatively level up, heal at base, challenge on good terms, bullet dodging, aoe, or fire gun. These AI do their best to destroy the enemy base and turrets while conserving their lives and navigating the map via a Dynamic A* Path Planner.
--May Require Mongoose web server to view--
1) Open the index.html file
This data visualization is an experimental visualization that allows a user to explore a set of Street Fighter games. Insight can be gained on characters, match ups, common win/loss conditions, and an overall examination of the metagame. The purpose is ultimately is the examination of a player's weaknesses, and also how the game develops when managing the balance of the game.
Christopher M. Tansey
I look forward to pushing myself past higher standards when working on games to introduce the game in a way that fits the player. I work a little longer and a little harder each day when thinking of new ways to create AI that can suit the needs and adapt to the playstyle of any project lead or intended audience while still involving a general audience. I want to involve design and programming without discouraging those less skilled at video games, and still challenge those who look towards something greater. I dedicate my life to this balance.