One of the things I used to do as we were coming into the home stretch on releasing a new RTS game was play test the levels.

It was fun and hard at the same time. I had to put myself in the mindset of pretending to see the level for the first time ever, and try to react as a new player would react to events in the world.

The hard part was trying to take notes and feedback during each play session. Nothing like having to stop ever 2 -3 mins to take notes as a way to ruin your game experience.

Play testing C&C: Generals, I had the idea that if we added code to track the game activities I performed, we could create simple graphs with the information and I wouldn’t have to write any more notes!

Worked like a charm. I could simply play a level and then use the graphs to tune each level based on the graphs.

Some simple patterns emerged. If the graphs showed the simply huge spikes in a repeated pattern of “collect money”, “build lots of tanks”, “attack enemy”, then I knew I was making unsuccessful attempt to crack a base defense or take down an enemy force. From experience, I knew that doing that pattern 2 or 3 times was ok, but when it got to be 4, 5, and 6 attempts, the game would start to get boring so we would use the graphs to go back and tune the level.

That experience taught me the power of using real statistics to help tune game play, rather than simply rely on a sense of “feel”. It’s also one of the reason I enjoy working on Social Games. We can use real stats and numbers to tune and improve the game experience for our players each day if necessary. There’s a sense of real freedom doing things this way. If you’re making a game and haven’t tried it yet, I strongly suggest it.

One more bit of info that bears highlighting from the October 2007 article about Yves Béhar in Fast Company.

His 7th design axiom:

“Never ask the consumer about the future. You can ask them what their aspirations are, but you will not get an answer about what you should do. Design will bring those stories to life.”

I’ve seen so many focus group testing waste money because they centered around asking players what features they wanted in a game rather than their “aspirations” or the “fantasy they wanted to live out”. I think that game players are better able to articulate what they want to do, be, and see than they are at telling game developers exact features they would like to see in a game.

I mentioned “Aldus FreeHand” in a previous post. I actually first saw this design axiom used in practice there by the designers there. Similar to game players, graphic artists were much more able to articulate what they wanted to do rather than give the team a design or feature to accomplish that same goal.

In “Command & Conquer: Red Alert 2“, we focused on the “fantasy” first started in “C&C: Red Alert” about time travel and alternative history and science. We didn’t go to focus groups to ask them what type of weapons, units and buildings we should add to the game. I find it hard to believe that a player would have come up with a “giant mind controlled squid” that our designer Todd Owens did.