Next, on playtesting...

Next, on playtesting...

Originally shared by Rob Donoghue

Ok, yesterday I floated something from twitter about Playtesting, and it got some good talk - https://plus.google.com/+RobDonoghue/posts/hVLeQivNEJR- enough that I want to unpack further.

First, I want to call out the implicit 4th bullet point, one I consider obvious but maybe needs to be stated outright:

- This is not because your testers are dumb. They are not. But they are (and should be) serving their own agenda.

That "and should be" is important. You want them to be focused on their problems, wants and needs. If they are thinking too much about testing your game, then that's going to skew the result in a way that may lose you valuable insights. It also means that if you assume they're smart, then you want to gather feedback in a way that takes advantage of that, which is not just a matter of listening.

The other implicit point, and one that probably requires a lot more talking is the "ok, therefore....?"

See, that list of the limitation of playtesters is not a summary, it's a starting point. It is the list of challenges you'll want to overcome in crafting your own playtest. If you think they're untrue, that's fine, and you can craft strategies according to the truth that you see. But if you think there's something to them, they serve as a starting point.

Which maybe demands that we step back and ask what we mean when we're talking about playtesting.

I don't want to step it back too much. There's no one right answer, and if there were one, it would depend on a lot of assumptions about resources that won't be true. The perfect test will have independent testers, lots of test subjects, copious tools for data gathering and analysis, and all the time necessary to do things like robust A/B testing.

Nobody has that. It's not just a limitation of gaming. Even big companies (who face similar challenges in testing software, media and so on) have to work within limitations. No one can ever test as much as would be ideal.

Despite that, there's a big gap between "We can't test perfectly" and "no testing" and you can make decisions and use tools that allow you to get more out of the tests you're performing.

There is a lot of knowledge out there about this. As noted, testing is not only the domain of gaming, and we're well served to learn the techniques, tips and tricks that let us maximize our bang for the buck. There have been some great suggestions in the previous discussion, but I suspect we've just scratched the surface of the lore out there.

So rather than leap into answering that, I want to ask you all - given that playtesting is almost always constrained (in resources, time and availability). what specific, concrete and actionable tips or suggestions for people running playtests that might help them maximize the return.

I'll take my own thoughts to comments as well.