It's 10pm: do you know where your players are?

Lecturer: Darius Kazemi, Orbus Gameworks

In 2008, Orbus Gameworks carried out a study for IGN Entertainment where they investigated the metrics that IGN has collected on player behavior over hundreds of thousands of player hours in games such as Command & Conquer 3: Tiberium Wars and Unreal Tournament 3. This talk will cover some of their more interesting findings, and also go over best practices for gameplay metrics collection.

Orbus and IGN found some data that was truly surprising, and some that reinforced long-held beliefs about player behavior. We looked at ratios of units killed to units lost per player over time in C&C3:TW, and what that tells us about player engagement. Using these metrics, we even learned a little bit about the types of players who prefer different factions. We found some surprising data concerning the correlation (or lack thereof) between time spent in game and player skill.

In Unreal Tournament 3, we uncovered data about the interconnectedness of IGN's social network, and how those relate to player skill and time spent in game.

Some metrics are recommended no matter what kind of game you have. We'll discuss some best practices for metrics collection, including techniques like intelligent population segmentation, tracking event density, recording almost-orthogonal data, and a few algorithms you can run on a social graph to turn it into an essential tool for community managers.

Intended Audience: Designers and anyone who wants to collect metrics.

Prerequisites: None, although some math background will helpful.

Format: Lecture

Track: Design

Date/Time: Tuesday, May 12, 10:30 AM

Room: Harbor