Now you know why you're playtesting, it's time to use that knowledge to figure out some of the details of the tests you'll be performing.
In this post, I'll be looking at two aspects of this: where you should playtest, and who should be testing your game.
The first thing you should probably decide is whether you're testing online or offline. These two options each have their attractors and detractors.
Online testing gives you larger exposure and allows you to test more people with no added time taken in testing itself. However, you need to do a lot more in-game setup to get data, and working through all the data can be a pain. Offline testing allows you to see specific player reactions to your game and can make pinpointing specific problems very easy. Yet it requires you to get participants into the same room as you (and probably at least one other person), and really limits how many players you can use, based on how much time you have.
Ideally you will be using a combination of both of these, particularly later in development, where you want a lot of eyeballs looking at your game, but also want to maintain that high level of scrutiny that offline testing provides. For early testing, however, you can easily just do offline tests (see my post on how many playtesters you need to get a better idea of why this is).
In fact, you should read that post now anyway, as the next step is to work out who should be playtesting your game.
As well as sheer numbers, however, you need to decide on a demographic to test with. This specifically relates to the testing purpose you previously decided upon.
Some things that can be important are: age, sex, general gaming expertise, whether they've played similar games and whether they've seen your game before. 'Some' because there are probably other requirements specific to your game (you'll notice that if you sign up to playtest Particulars that I'll ask you about your physics education), and 'can' because you should only care about these things if your game requires you to (you'll note that we don't ask for gender in our sign-up form, and we want to know about age in terms of brackets, and this is actually for us to gauge player interest in different markets).
For each test, different aspects will be most important, so I'll be talking about gaming expertise both as an illustrative example and as something that you'll have to consider 90% of the time. If the main purpose of the test is to determine if the core gameplay is fun, you might want expert gamers to have a look at the game as they'll quickly pick up the controls and mechanics and get right into telling you if the gameplay is fun. Expert gamers are also great at finding imbalances in your game. Conversely, if you're interested in how well the game teaches, your best bet is to go for people who've never played a game in their life. If they can pick up your game without any problems, anyone can (this is why it's great to test new UI's with the elderly). Eventually, you'll really want to test with all types of people, even if each test only looks at a single type (for instance, you need to know if your awesome teaching system that can teach anyone is really boring to the expert gamer).
If you look at the purpose for your test, who and where you need to test should become fairly clear. If not, if you have any questions or have a story about the where or who of playtesting, please let us know in the comments.
That's it for now. Next time, we'll be looking at what you're measuring.