Why? Well, one of my patrons asked about playtesting. But when it comes to playtesting, there’s a lot to talk about. As a game designer, playtesting is how you spend most of your time. After three years, I have a few more insights on the art and science that is playtesting.
Ideally, a game designer uses playtesting as a tool to explore game space, much like an astronomer uses a telescope to explore outer space. A game designer approaches a playtest like an experiment, coming in with something to focus on and a hypothesis; keeping an objective, open mind about results; and taking careful notes with as much quantitative data as possible. Playtests allow a designer to explore what a game can be and illuminate areas that are weak or need work. Ideally, playtesting will provide a designer with the empirical evidence necessary to say with confidence that the game is as good as it can be, or a direction on how to improve it.
In reality, playtesting is messy and imperfect. We’d like to try every possible core system, but we can only test so many. We’d like to test every possible combination of special rules, but the number of combinations is just too big. We’d like to be able to say with certainty that the game is balanced, but we can never get enough data to prove it.
The reality is that a game has to change as you’re designing it. You need to make decisions with incomplete data. You can only playtest so many times, and you just can’t try everything. And even if you could, you’re playing with different players every test, skewing the results. And even if you’re playing with the same testers repeatedly, those testers are learning about the game and changing their play accordingly. A designer just can’t test it all.
Given that we can never live up to the ideal, should we just ignore it? Of course not. If nothing else, the ideal keeps us following good habits. The more data we collect, the more playtests we run, and the more focused those playtests are, the better the final game will be. But it’s important to realize that a lot of playtesting, and game design in general, comes more from the gut than from the brain.
It’s a little unsettling to know that you’ll have to make calls without knowing if those calls are correct or not, but here’s something to console you: you’re making games for people, not robots. It doesn’t matter if the game is precisely correct according to the numbers, as long as it feels right. Ultimately, you’ll never achieve perfect balance, but that’s ok; the important thing is that the game feels balanced. And of course, the opposite here is also true: even if a game is balanced according to the data, is it really balanced if the players don’t think it is?
Making the Most of Imperfection
So, if you’re not going to get perfection out of your playtests, what should you strive for? That changes over the course of development. As in previous articles, I’ll describe some stages below, but it’s important to remember that these stages are not discrete: you can be in more than one at once, and you might sometimes go back to older stages. Additionally, they can take vastly different lengths of time for different games.
Early on, you should be exploring different possible core mechanics and systems. Here you should be making wild changes between playtests. Don’t worry about details like how the game will end or whether anything is remotely balanced. In fact, if you do, you’re probably wasting your time, because the things those decisions depend on will probably change radically before the game is finished. Instead, tear out and replace entire systems between playtests. Your goal here is to explore as much design space as possible and find what seems most promising. The focus should be on what’s most fun and you shouldn’t worry about getting more than a rough idea of the game’s structure.
Once you have an idea about what makes the game fun, it’s time to start settling on some of the more specific game structures. The important thing here is to make sure that the structures all support the core fun of the game. Everything should be evaluated by asking the question: did this enhance the fun, or distract from it? Every game is different, and some games, especially larger ones, can benefit from slight distractions. But overall it’s a good idea to at least consider how everything relates to the core fun.
After the game’s core structures are in place, it’s a good time to start exploring unique game pieces and special rules. Many games have a lot of unique pieces (cards, buildings, enemies, special events, etc), and it’s important to make sure that these all complement and support the core fun of the game. Again, you should explore as many different possibilities at this stage as you can, determining special rules that support the core fun, are easy to understand, and create interesting situations. Don’t be afraid to experiment with special rules that seem weird or bad–if they don’t work, you can just throw them out. It’s often good to make game pieces intentionally unbalanced in this phase. If a game piece is overpowered, it will likely get attention from playtesters, who can tell you if the special rule is fun or not, which is what you want to discover.
Finally, after you’ve determined the game’s core fun, its main systems, and you have a good idea of what special rules you’ll include, it’s time to worry about details like balance. This is a combination of tweaking individual game pieces and seriously considering how game pieces fit together as a complete package. This is the first stage where you should show restraint when making changes. Instead of experimenting in this stage, your goal is to fine tune everything, which means you’ll often want to make small changes to individual numbers.
My patron originally asked how much to change between playtests. The short answer is that it depends on the stage of the game’s design. For most of design, you should feel free to make huge changes. Towards the end, in what’s usually called development, you’ll make small changes, fine tuning specific cards. But with the reality that you can’t afford to playtest a single game more than a few hundred times at most, you’ll often need to make big changes towards the end of design even if it’s not ideal. So the really short answer is: designers change a lot more between playtests than you’d expect.
I mentioned that you’ll playtest a game a few hundred times. For an hour long game, that’s 300 hours, and that’s not including finding playtesters, organizing playtests, and commuting… that’s a lot of time. There are ways of getting important data with less time (like solo playtesting and doing partial playtests), but any way you cut it, playtesting a new game takes up a huge chunk of time. Ultimately, I think that’s one reason that most publishers outsource design work to independent designers.
Unfortunately, many designers, especially newer ones, neglect a lot of important playtesting work. Many gamers might think they neglect the balancing portion of playtesting, since imbalances can easily stand out. But it’s the earlier stages of playtesting that often see almost no effort. A new designer gets excited about an idea, be it mechanical or thematic, and immediately dives into testing for balance, neglecting to explore what the game could be at a higher level. As a designer, it’s important to remember that your job is to find the fun for your players, not for yourself.
Alright, that’s all I have about playtests. (For now.) But it’s a big subject, and I’d love to hear your thoughts on playtesting. Do you have any strategies or considerations you use when playtesting? Any good stories from a particularly good or bad playtest? Share them in the comments!
I would like to thank my generous Patrons for the support and encouragement that made this article possible.