Iteration makes great games – duh
Random point of interest – Blizzard and Valve are two studios known for their ability to do no wrong in game design. Just recently, Valve’s Chuck Faliszek and Blizzard’s Kevin Martens have given the same reason for why: Iteration.
Basically, both profess to build a playable model of the game as soon as possible, and then test, revise or even throw out entire ideas if they prove unworkable. The main difference seems to be that Martens tells Gamasutra they do this inhouse (which is explains why Blizzard has so little leakage on projects – though how they shut up their employees, gamers have long dreaded to speculate). Faliszek on the other hand told TVG that Valve brings in outsiders to do the playtesting, to see how actual gamers will act – because how they play is how the game will be played. That could explain why Valve has blazed through Left 4 Dead and Left 4 Dead 2 development at what seems like lightspeed? Playtesters uncontaminated by the designers’ idea of how a feature should work may well end up using it completely differently from how it was envisioned, and so encounter or solve problems that won’t come up with inhouse people, who have a rough idea of intended usage.
Speculation aside, fun to see top designers from different corners of the field agreeing that what makes ‘em great is plain work and correcting themselves, not some mythical X-factor.
Exactly the process the NetDevil folks have been using for Jumpgate Evolution. Using agile software development you can burn through iterations of functionality really fast, iron out any potential problem areas and respond to feedback from testers – be they internal or external. NetDevil have used both – they iterate and iterate and iterate internally then grab people from nearby game shops/off the street and ask them to playtest. And film them doing so. Then use the feedback to iterate some more.
I’m torn as to the pros and cons of this approach. Speaking as a long-standing community member over at the Jumpgate Evolution forums, one of the biggest complaints I see is that this “random Joe off the street” approach means you’re getting feedback off people that don’t necessarily have the drive or passion to provide meaningful critique of the game. On the other hand, if you want to make a game that’s easy to get into, accessible but ultimately complex to master (which I hope JGE will be…) then it’s a really good approach to take. But where do you draw the line? How much feedback can you get from someone that might not even have heard of your game or that doesn’t even play MMORPGs, let alone after only a couple of hours of gameplay? Maybe that’s where the “complex to master” element comes in; your random, focus testing public are the perfect testers to tell you what’s wrong with your user interface and to identify usability issues. Hermann Peterscheck (Lead Producer for JGE) has often been cited as touting that “first fifteen minutes” as being critical to capturing a potential subscriber in an MMORPG. Can you get the accessibility and polish early on in a project’s lifecycle (which is kind of the idea of agile, iterative development) with just internal focus testing? Is it crazyness to keep things internal (all use of cranial explosive devices aside)?
I’m interested to see how this pans out with a smaller developer like NetDevil. The game isn’t in beta yet but has been “Friends and Family” testing for some time so I’m guessing they’re listening to their testers and…iterating their iterations. Time will tell…