This dataset contains information on player reconnaissance in over 500 professional-level Starcraft games. From the perspective of one player (the Terran), it contains information on how many enemy (Protoss) units the player has observed, can observe, has seen destroyed, etc., along with an overall measure of how much enemy territory the player can see. ## Acknowledgements This dataset was downloaded from [this webpage](http://web.engr.oregonstate.edu/~tgd/rts/scouting/). It was the basis for the following paper: Hostetler, J., Dereszynski, E., Dietterich, T., and Fern, A. (2012). [Inferring strategies from limited reconnaissance in real-time strategy games](http://www.auai.org/uai2012/papers/294.pdf). Proc. 28th Conference on Uncertainty in Artificial Intelligence (UAI 2012) (to appear). ## The Data Games are divided into 30 second chunks, with the first 7 minutes of each game being represented in this dataset. Values of variables at any given time cycle represent their values over the entire chunk that ends at that time. This dataset contains the following fields: - **game**: a unique identifier for the game being played - **cycle**: the cycle (in game frames, which are typically 24 fps) - **unit**: the enemy unit that this row gives info for - **losses**: how many of this enemy unit were lost during this time chunk? - **observable-units**: how many of this enemy unit could the player see during this time chunk? - **observed-losses**: how many of this enemy did the player *observe* being lost during this time chunk? - **production**: how many of this enemy unit became observable (i.e., was produced, or -- in the case of buildings -- was under construction) during this time chunk? - **scouting**: how many of this enemy unit did the player scout during this time chunk? - **vision**: what proportion of the total enemy territory could the player observe during this time chunk? -- NOTE that **vision** appears once per unit; however, the **vision** variable is not linked to any one unit. Its value spans the time chunk, and is identical in every row that represents a given time chunk