GDC 2013 Sessions: Blizzard Entertainment Speakers

Description: StarCraft II: Heart of the Swarm embraces cinematics as a high-level storytelling tool, a tactic that has been under fire in the games industry in recent years.

This panel will examine the strengths and weaknesses of that approach by tracing the story development of one particularly difficult cinematic. By showing multiple versions of the animatic, we will track the differing speed of cinematic production vs. game design, the feedback we received, the choices we made, and how they impacted the story.

We will cover the obstacles that we overcame, those we didn't, and the lessons we learned.

Description: Game physics engines are used to create games like Angry Birds, Half-Life, Tomb Raider, Red Faction, MotorStorm, Limbo, and Diablo 3. Games would be far less compelling without the realistic physics simulation that engages the player's intuition and stimulates their motion awareness.

Physics has become a staple of the modern gaming tradition as we try to re-create and re-interpret the world around us. The Physics for Programmers tutorial brings together speakers from Blizzard, Dtecta, Havok, NVIDIA, Sony, and Valve. Topics include collision detection, continuous simulation, constraint solvers, and demos.

The focus is on rigid body physics and real-time simulation in games. There will be a mix of introductory topics, recent algorithms, and practical tips. Time is reserved for Q&A.

Description: As with all Blizzard games, Diablo III went through many iterations before the final product was released. Wyatt will describe different game systems that were attempted, explored, and then ultimately removed. We'll go into detail about the pros and cons of each system, what worked and didn't work, and how the team used each failure to bring them closer to success.

1. Health Recovery: From Regenerative Damage Shields to potion dilution, where potions become less effective as you drink them, many different health recovery mechanisms were tried. What were the major issues with each of these systems and how did it lead to health globes in the end?

2. Controls. At Blizzard, control is king. Although simple on the surface, a number of iterations were required to polish the Diablo III interface, in order to provide the player with tight and responsive controls.

3. Skill System: The skill system underwent multiple revisions. Early systems involved different variations of point spending, an evolution from Diablo II. Later systems had runes, with multiple rune ranks that dropped as items, which could be used to modify skills.

Description: Online games need robust, easy to use network APIs. No one should still be struggling with how to get their data from one place to another. JAM is World of Warcraft's inter-server serialization and routing layer.

This talk describes how JAM came to be, and how it is used today. Real world sample code from WoW and other Blizzard projects are used to illustrate key concepts, such as machine generated code for serialization, and dynamic protocol negotiation. Comparisons to similar open-source projects will be made.

Description: This session will present how the new cutscene editor and its underlying engine were built in StarCraft II: Heart of the Swarm. The talk will review the design process, and how having a user-focused approach shaped both the tool and engine alike.

It will then provide an overview of the technical design and architecture of these systems. Finally, the talk will discuss the emergent workflow that resulted in having the artists so heavily involved in the development process of their tool.

Description: As game designs lean on AI more and more (finally!), there is an increasing need for a tighter interface between AI programmers and game designers. While much of this relationship is about communication, there are other important considerations that AI programmers can address to make their life with designers easier.

The first part of this lecture will offer suggestions on how querying designers can correctly help the programmer to design an architecture that will scale up to handle the inevitable feature expansions that spill out of design throughout production. The second portion of the session will present methods to gracefully scale down AI systems as needs change.

This includes "killing the baby," which is when a programmer is forced to cut out their wonderful, clever, elegant, amazing, best AI feature ever because it doesn't fit the overall design, or simply isn't fun.

Description: Items in Diablo III can have one of 400,000 possible names. How do you translate and test this huge pool of randomized content? By developing flexible systems up front that work for every language - including English.

This presentation outlines the development pipelines that allow localization to support Diablo III's high level of replay value and randomization. Special focus is given to conversation implementation, and the randomly constructed item and monster names.

Blizzard's internally developed localization tool will also be shown. Stop translating whatever the developers hand you, and start thinking of English as another localized language!

Description: Game sizes at Blizzard and other game companies have grown significantly over the years. This has led to a situation where most game build systems are I/O bound, making it difficult to maintain fast build times.

This session will present Blizzard's in-house distributed, multi-platform build system, which was designed from the ground up to minimize disk and network IO, and to maximize scalability. Blizzard used this framework to reduce Starcraft II's build time from 6 hours to 30 minutes.

This session will cover the technical architecture of the build framework, and present a number of the scheduling algorithms involved.

Description: When building any game, but especially when it's a sequel to the beloved Diablo series, it's critical that you have a series of core design goals, or pillars, that you can use as a guide to making decisions and defining what you want the final vision of the game to accomplish.

In this postmortem, we'll explore the pillars that guided Diablo III's development and how well we felt we accomplished each of them. We'll focus specifically on the ones we feel we struggled with and the game design lessons we learned during those struggles.

Examples of specific challenges of living up to these values will be provided, prioritizing them against one another, and evaluating how well they were accomplished after Diablo III was released.