GDC is about game developers getting together and sharing their philosophies on game development, and Rob Pardo took part in that in a big way today with a keynote address called "Rules of Engagement: Blizzard's Approach to Multiplayer Design."

This wasn't a bragging session, though -- Pardo fully admits that not everything Blizzard has done with World of Warcraft was perfect -- and they've had to fix those issues over time. He used Alterac Valley as a strong example of something that the company realizes turned out totally different than they'd planned, once players were able to access it in mid-2005.

It was a fairly long address, with several great quotes from Pardo on what he instills in Blizzard's design teams, as well as his own philosophies for making a good game.

The topic also wasn't just about World of Warcraft; it covered everything Blizzard has done in the past, and their next project -- Starcraft II.

"Starcraft is a twitch RTS, and that's intended" he said while covering the core mechanics you'll need to mull over and consider when designing PvP for a game.

When speaking about Avatar Improvement -- such as obtaining items and new spells, another factor in balancing a PvP game -- he noted "We never really intended World of Warcraft to be a heavy E-Sports game," touching on the fact that initially, they had no thoughts of WoW becoming more than a casual PvP game with mini-games such as battlegrounds, and of course, world PvP.

Things changed, and Pardo made sure to stress that whether or not you want your game to be involved in E-Sports should be considered before releasing the game.

He then talked a bit about Balance, and the never-ending struggle of "Math-vs.-Fun." Blizzard deals with it all the time, and Pardo tells his designers, "Don't use the math to balance the game into mediocrity."

"You need to play the game" he said, when discussing understanding each and every nuance of a game; this is crucial to a designer actually being capable of contributing to a fun, engaging experience.

When he moved on to the User Interface, and how it affects balance, one of the striking points was UI mods themselves: "mods are bad when they play the game for you," he said. A short pause occurred and he then finished, "outside of that, all mods are good."

The next slide was titled "It Never Ends" -- targeting massively multiplayer online games specifically -- but also touching on Blizzard's history of continually patching their games -- even 8-10 years after release.

Pardo advised the crowd to plan patches, but always allow time to be reactive to new bugs that spring up during a game's live state. Still, you also can't panic. "If you change things too often, or change things all the time, players will start using you as a crutch."

Next up was "Player Psychology" and how that affects game design. "If the players don't believe it [the game] is, it doesn't matter." when speaking about a player's perception of fairness.

On the topic of Maps and their design, it's important to decide what is right for your game: random or pre-made maps, the latter of which Blizzard has chosen to stick with through all of their RTS games. "Size does matter on maps, bigger is not better in this particular case," he echoed when speaking about the size of maps, and how you typically want to keep them small if you can.

It was an eye-opening keynote address, because it gave the impression that even Blizzard, with their 10-million-strong juggernaut, has been humbled by the success of World of Warcraft. While the game may be the giant of online gaming currently, they recognize it has its flaws, and they're continually working to fix them.

Below is a list of the notes during his presentation:

PvP

Skill Differentiation

• Twitch/Micro-management
• Multitasking
• Strategic Thinking
• Economy Dominance
• Knowledge and Experience
• Avatar Improvement

General Game Balance

World of Warcraft

• Solo to max level
• Have a role in groups
• Have a role in raids
• Be competitive in group PvP
• Be fun!

Starcraft II

• Races totally different
• Better players can win games fast
• Offense over defense
• Creative Strategies
• Every unit has a counter
• Be fun!

Balance - Math-vs.-Fun

• Math is the foundation
• Everything should feel overpowered
• Must understand nuances
• Balance for all skill levels
• No super weapons!
• Use your beta

UI Affects Balance

• Unlimited Selection
• Multiple Building Selection
• Sub Groups
• Click to move
• UI Mods

It Never Ends

• Change is always bad to players
• Maintain the game to keep it relevant
• Ban the cheaters
• Plan patches, but leave time to be reactive
• Don't panic!

Player Psychology

• Perception of Fairness
• Players hate losing
• How do they climb the ladder?
• Right amount of complexity
• Inspect message was creepy
• The rest system -- it's not a penalty -- it's a bonus!
• Warcraft 3: "thumbs"

Visual Clarity

• Really important process
• Does a model visually suggest it's function and power?
• Special [spell] effects can cause confusion
• Team differentiation
• Avoid hidden modifiers
• Who runs faster? Who can take more hits?
• Harder in Sci Fi
• Very hard to do in an MMO, good reason for class-based armor

Maps

• Random vs. pre-made
• Number of maps -- more is not better!
• Black shroud vs. dark fog of war
• Size matters

Matchmaking

• Less buckets is better
• Don't give players too many choices
• Feedback on the system is critical. Players want to know it is working.
• Momentum can work against you. If players think the system is broken, it is...

E-Sports?

• Replays
• Decide upfront if you want an E-Sports game
• Spectator
• Referee Controls
• Web Support
• Map editor, modding ability