Predicting Cognitive Performance in Open-ended Dynamic Tasks
A Modeling Comparison Challenge
Organized by: Christian Lebiere, Cleotilde Gonzalez, & Walter Warwick


:. Home .:
:. Results .:
:. Registration .:
:. The DSF Task .:
:. Method .:
:. Human Performance Data .:
:. References .:
:. FAQ .:
:. Contact Us .:

Dear Colleague:

We write to invite you to participate in a Modeling Comparison Challenge: "Predicting Cognitive Performance in Open-ended Dynamic tasks". The goal of the challenge is to advance the state of the art in cognitive modeling by evaluating alternative cognitive models against human performance in a common dynamic task. The task captures in an integrated manner some of the key aspects of human cognition including the detection of temporal regularities and the ability to control dynamic systems.

The description of the challenge and the information for participation (including data, task environment, references and documentation) can be found on the challenge web site:

Please send an E-mail to: to register as soon as possible. You must be registered before submitting your model. The most important dates are:

The Modeling Comparison Challenge will be the focus of a symposium at the Behavior Representation In Modeling and Simulation (BRIMS-2009) on April 1, 2009 in Sundance, Utah.

Deadline for model submissions is May 15, 2009.

Announcement of the challenge results is June 1, 2009.

The results and winner(s) will be presented at a symposium at the International Conference on Cognitive Modeling (ICCM 2009) in Manchester, UK on July 24th-26th, 2009. One representative for each of the 3 best models will be invited to present at the ICCM symposium. Their expenses will be covered for up to US$2000.

Thank you for your attention and we look forward to receiving your models!

Christian Lebiere, Coty Gonzalez, Walter Warwick.

Organized in part by the Dynamic Decision Making Laboratory, a part of the Social and Decision Sciences Department at Carnegie Mellon University. For updates and comments regarding this website, please email