On playtesting, part 2

Written by Heikki Kuorelahti

In the first part we established why you should playtest your game, now I’ll explain how to do it (or at least how I do it). Even though all of my games have been multiplayer games and I talk about playtest groups, there’s no reason the same principles wouldn’t apply to single player games and playtests.

When

When it comes to playtesting, there is really no such thing as ‘too early’. The earlier you start playtesting, the better. Even if you’re making a video game, you should consider making a paper prototype of the game before writing any code. Draw maps on paper, use coins and bottle caps for marking the positions of characters/objects, roll dice for randomization. Obviously not all games make interesting paper prototypes, but surprisingly many do (e.g. turn-based games, tower defense games and heck, even Sims).

When you go digital, you shouldn’t work too long on your first prototype. Don’t polish it, don’t make ten minutes worth of content into it. Just try to nail the core idea of your game and build that (only). Not spending a lot of time on your first prototype serves two purposes. For one, it makes it easier to see the true quality of your core idea. If the game doesn’t look attractive and the sounds are crappy placeholders but the players still have fun playing, that’s a very good sign. The second reason not to spend too much time on your first prototype is you don’t want to get attached to anything at this point. If it took you half an hour of drawing, cutting and glueing to make a paper prototype of a dog racing game, it’ll be easy to throw it all away if your playtest shows you that no-one finds the game engaging. If, on the other hand, you’ve spent two months writing code, modeling the hounds and the track and recording dog sounds, you’re going to be clingy when your playtest shows that players just aren’t having fun playing. You might think “they just don’t get it” or “they’ll like it once they see the final art and when the music is done”, because you’ve spent a lot of time on the game and don’t want to feel like it’s been for nothing (btw it hasn’t, you learn more from a failure than from a success). So, to start off in the right direction from the get-go, playtest early.

Incidentally, this is exactly the way Gunnheim was kicked off. A prototype with the barest essentials was made in less than three weeks for BeatCon, where the game was publicly playtested with convention participants. The prototype was well received, and this in turn gave SIEIDI the confidence to continue developing the game further.

Who

Who you playtest with makes a big difference. Friends and family will be biased towards a game you made (especially if they know you’ve spent months creating it), and will generally like it more than the average person. Still, they have their benefits, mainly the fact that a playtest session with your flatmate or sibling is much easier to arrange than one with a stranger. The earlier you are in development, the more it’s ok to playtest with family and friends. Playtesting with strangers will yield the best results, as it most accurately simulates the way most people will come across your game once it’s published: knowing next to nothing about you or the game and having no predisposition towards either. Ideally, you want to playtest with members of your target audience for the most part (if you don’t have a target audience yet, you should establish one now!). You learn a lot by playtesting with people outside your target audience and you should definitely do it to some extent. Just know that the things they might dislike about the game could be the things that are the most appealing to the target players, so you want to prioritize the feedback.

How

As for the actual playtest, there are a few things to keep in mind. First of all, you should give clear instructions on how to get there. You don’t want to have the playtesters stressed about finding the place. Have everything ready for when they arrive: check that the game works, print out the feedback forms, take care of all the little stuff. It’s difficult to observe and take notes if you’re desperately trying to find a pen or hastily brewing the coffee you forgot to make. If you are going to record the playtest for further analysis, you should check that the playtesters are ok with being recorded.

Before the playtest, explain that you are not testing the players, you are testing the game. There should be no pressure on the players to do well, and they should feel free to critique the game openly without anyone getting hurt. During the playtest talk as little as you can. It’s very tempting to quickly help out the players when they get stuck or to start explaining a feature in the game. Doing so, you are missing one of the key benefits of playtesting: smoothing out the game experience. Use playtest feedback to make the game explain itself (so you don’t have to). Take notes and observe the players. Do they seem engaged? Bored? What frustrates them? What makes them smile or laugh? When do they quit (or seem like they want to quit)?

game-survey_jpeg

After they’re done playing, give the players a feedback form. This shouldn’t be very long. An individual feedback form is important, because that way you get at least some feedback from everybody. Quiet players can say very little in the group discussion. After filling out the form, discuss the game with the group. Ask open-ended questions (don’t ask “was the weapon too weak?”, instead ask “how did the weapon feel?”). Direct your questions at individual players so you get to hear from the less vocal players as well.

When the players critique something, don’t get defensive! The whole reason you are playtesting is so they can give their opinion on the game. If you start arguing about design decisions or explain all the cool stuff that will be in the game eventually, the playtesters will feel like their input isn’t valued. Even if the player isn’t a target audience player, openly accept all of their feedback and thank them for it. You can analyze and prioritize the data later.

Once you have all of the data, collate it and see how often a certain feature comes up. If almost all of the playtesters say that the player character moves too slowly, you’ll have a pretty good idea on at least one thing to change for your next build. If one of the playtesters says that the weapon sound should be changed, you should consider the priority of that task. The input of target audience players will obviously be of higher value than other feedback.

Refine the game based on feedback from playtesting, all the while creating new content. Playtest the next build, refine, add content, repeat. Iterate for as many times as necessary and you should end up with a polished game that your target audience enjoys playing.

As a final point, remember to always be thankful to the playtesters. They are an invaluable help to your game development and receive no compensation for their time. A sincere ‘thank you’ will go a long way in recruiting players for future playtests!