With the news that Rob Pardo is leaving Blizzard, it reminded me I collected his "top ten" design rules from Blizzard at a talk he gave at GDC 2010 and never posted them. There are probably more detailed write-ups elsewhere, but here they are in bullet form:
- Gameplay First
- Concentrate on the fun
- Art, design, programming all in the service of fun
- Design isn’t more important than others
- Easy to Learn, Difficult to Master
- SImple mechanics to begin with
- Easly play
- Actually: easy to learn, almost impossible to master
- Provide lots of depth
- Trying to build depth first - harder and takes longers
- Do multiplayer first, then singleplayer
- Accessibility second
- What is the Fantasy?
- What is the expectation?
- Where is the fun in that?
- Make Everything Overpowered
- Take everything to 11
- Every unit, class should feel unstoppable
- Doesn’t cost anything to make something epic
- Applies to story and world too
- Avoid balancing to mediocrity
- Celebrate big differences in the game
- Concentrated Coolness
- Make each feature the coolest, most concentrated expression of gameplay
- Limited amount of complexity a player can process
- eg vehicles ruined class system since you could use clas
- StarCraft 2 has 16 units per race
- Killed old units to introduce new units
- Play Don’t Tell
- Play as much of the story as possible
- Use text/voiceover/movies to enhance the story
- eg 512 character limit on WoW quest descriptions
- bad eg 2+ minute monologues in Diablo 2
- Make It A Bonus
- Players respond better to incentives
- Path of least resistance should be most fun
- Don’t fight player psychology
- eg “Inspect player message”
- Randomness & progressive percentages build up (designed randomness)
- Control is King
- Controls should be as responsive as possible
- Sacrifice “cool” for better control
- Players unlikely to complain about slightly bad controls
- Done well becomes a key skill differentiator for players
- Tune It Up
- Easy to do, hard to do well
- Plan for tuning hooks
- Know who you are tuning for and why
- Avoid the Grand Reveal
- Check in often and get lots of feedback
- Difficult to tell if an idea will work
- Must be set up to fail
- Iteration is critical to Blizzard dev process
- “If you can’t build that culture, it won’t work”
- “Want people to say ‘can you help me make this better’”
- Culture of Polish
- Polish doesn’t happen at the end
- Team gets to make their favourite game better
- Make sure team loves the game
- Cross section from other teams and preview players to feed back into the game
- Every voice matters - Don’t ship until it’s ready
On a related note, Pardo also mentioned something called "Strike Teams", which were also mentioned by Dustin Browder in this interview about Heroes of the Storm:
"We have a process here at Blizzard called a strike team where we get a bunch of people who are not on the game teams, well they're not on the game team making the game," Browder explained. Different people come into projects they're not a direct part of, play through the content that has been created, and give their feedback.
"And it's the most brutally honest experience you could possibly have in your life. They'll just look at you point blank and say, 'Well, you're not gonna release this cause it sucks, right?' I mean they'll just tell you whatever."
To me, that all sounds a bit like Pixar's Brain Trust system. (Read Ed Catmull's Creativity, Inc. if you haven't already.)