Collecting gameplay data related to matches in which there is the participation of cheaters for research purposes is not an easy task. First of all, it is not an easy task to gather and monitor cheaters, since this kind of players are usually disguised and not willing to be marked as cheaters. A second concern is related to the way the data are acquired during the gameplay: there are not open systems able to collect realtime data from a match in an FPS and moreover is difficult to define which could be a meaningful set of features able to describe in a complete way the status of the game and of its players. In order to be able to analyze real data using machine learning techniques to distinguish between a legit player or a cheater, we have created a mutator in Unreal Tournament 3 able to collect the data previously described, and we have played several matches with our Nemezis Framework, which has performance comparable with the one provided by commercial grade cheating systems. The data collected during these experiments are being distributed as part of this dataset.
The original link was: http://www.aimbot-labs.com/datasets/nemezis.html
LicenseThe authors of the research paper are the copyright holders of all the traces included in the dataset.
References and Citation
If you use this dataset, please cite the following paper [GLC11].
GLC11: Galli, L. and Loiacono, D. and Cardamone, L. and Lanzi, P.L. “A cheating detection framework for Unreal Tournament III: A machine learning approach” IEEE Conference on Computational Intelligence and Games (CIG), August, 2011