Singapore-MIT GAMBIT Game Lab spacer
spacer 2012 2011 2010 2009 2008 2007 spacer
Load Game
Summer 2009 Prototype | Action Puzzle | Adobe Flash (Windows / Macintosh / Linux)  

Dearth is an exciting co-operative action-puzzler. The Tribal Lands have been suffering through the worst drought in many lifetimes. Plant-life is withering away, and people fear the approach of a great famine. Rumors spread of the awakening of monstrous creatures who the ancestors warned would one day rise to drain the land and its people of their water. As the thirsty beasts emerge, worried villagers turn to the Tribal Lands’ two great shamans and ask them to restore the water to the land. 

Play as the tribal shamans. Force the mysterious water-sucking creatures to smash into each other, allowing stolen water to gush from their engorged bodies and be returned to the land. Plan movements with your partner carefully or be ready to make split-second decisions if things don’t go according to plan. The future of the Tribal Lands will depend on how well you work together!

2-Player mode offers lively, engaging, brain-cracking co-operative play. This mode allows a team to travel through the many different stages of the Tribal Lands and outsmart the beasts who have invaded. Stages offer a wide variety of puzzles and challenges which facilitates dynamic co-operative gameplay. Additionally, players are given the freedom to customize the landscape as they find puzzle solutions. Choose to concentrate water restoration in one area, or spread water throughout the entire stage.  Pick off enemies a few at a time or gather them all together and go for the Big Smash! Or, abandon the partner idea and try to complete stages controlling two different characters at the same time.

1-Player mode features extra stages which are played by one human player and one AI player.  This mode is meant to showcase the basic functionality of an experimental algorithm for a Markov Decision Process (MDP).  Try to understand what the AI player is thinking, and the AI player will attempt to understand you.


Go to Website

Research Question

Dearth is a research tool designed to help us understand the strengths and weaknesses of Markov Decision Problem solvers for implementing game artificial intelligence, specifically for side-kick characters. The goal is to develop character behaviors automatically, based on a description of the game rules, instead of a programmer-generated set of behaviors. The human-human game mode serves as a testbed for future research, presenting increasingly difficult challenges to the MDP solver.


Dearth requires Adobe Flash Player 10. The game has been tested on the following minimum hardware configurations:


  • Intel® Pentium® 4 1.6GHz, AMD Athlon 64™ 1GHz or faster processor (or equivalent)
  • 1 GB of RAM
  • Intel Core™ Duo 1.33GHz or faster processor
  • 1 GB of RAM
SplashScreen1 SplashScreen2 MainMenu OptionsMenu
AITutorialLevel 2 Player InGame Level20 Credits

Game Trailer

Research Video (from Game of the Week Podcast, 2010)
Poster 1 Poster 3 Poster 2  

Leslie Pack Kaelbling
Product Owner

Tomas Lozano-Perez
Product Owner

Lee Wee Sun
Product Owner

Andrew Grant
Embedded Staff

Tan Pin Xun
Producer / Scrummaster

Jeffrey Gan
QA Lead

Nick Ristuccia

Cheong Wei Yang, Aaron
Audio Designer

Benjamin Chee Siew Meng

Victoria Y. Chastan

Alec Thomson

Benjamin Chua

Mushfique Junayed Khurshid

Blog Posts


GUI Sketch

Bug Study

Dearth GDD

Press Mentions