As we have already mentioned many times, we work on Caromble! only on Fridays and every spare hour in the evening and weekends. Mostly, we write these blogs about our part-time struggles of making Caromble!, but this time, I want to take a moment to talk about my other occupation. Not that of being a boyfriend, but my ‘real’ job at Motek Medical.
Motek Medical creates products for rehabilitation. They combine motion platforms, instrumented treadmills and motion capture systems with a virtual environment to provide motivating and engaging training exercises and precise assessments for rehabilitation:
I am primarily a C++ developer here, but lately I am also working on rehab applications (so called ‘serious’ or ‘health’ games) for patients, doing game design and implementation.
Making a health game is very different from making a game for entertainment. For Caromble! I make design choices using my gut feeling and my own experiences of the games I love. It has to become a game that I too could enjoy. When designing a game for a specific patient group, the approach is very different. More than in any other game, the player is the main focus in a health game and the desired outcome is different. The game should be motivating, challenging and captivating for the targeted patient, not only to entertain, but with the goal of improving her (let’s assume she is female) capabilities. A game designed for this purpose, could very well be a game that I wouldn’t like to play or look at. For me, this takes me a bit out of my comfort zone, where as a gamer I think to know which are ‘good’ game design choices and which are not; knowledge which I can use in the game design choices of e.g. Caromble!. In health games these ‘rules’ I have learned simply do not apply.
When designing a health game, the main question is: “Who is the audience?”, and to go a bit deeper, also: “What is the (clinical) goal of the health game? “, “What is the patient capable of (cognitive, physically and visually)?” In an ideal world, a health game is designed for one person, so you can make specific design choices for that person, taking into account her capabilities and interests. Unfortunately, this is often not the case. A health game has to be designed for a whole patient group, for example stroke patients. The problem is, that one stroke patient is very different from another; one can still move her arm, a second can move it only a little and third can’t move it at all. Or one has an easily overloaded cognition, while another can handle a lot of different cognitive signals easily. Besides these variations there are also the subjective preferences that patients can have about visuals, gameplay and sound. Making a game that is suited for a whole patient group is according to me, the biggest and most interesting challenge.
Before we start prototyping a new health game for a specific illness or physical/psychological problem, we sit around with therapists and patients to get to know the intended audience. With the retrieved insight we start prototyping both gameplay and visuals. We playtest the prototypes ourselves and with patients and with those new insights we often have to go back to the drawing board. It still amazes me how difficult it can be to fully let my own assumptions of a ‘good’ game go and think ‘with’ and ‘for’ the patient, about what is desired for her.
Once we have a suitable prototype for the project, we still have the challenge of making the game such that it is challenging and captivating for the whole range of patients. A scalable difficulty is one of the most important aspects to achieve this in my opinion. There are two ways to do this:
- Incorporate many settings that the therapist can adjust, such that the game can be tailored to the specific patient.
- Design the game in such a way that it uses adaptive difficulty.
I think the best solution is to use a combination of these two. Expose some settings to the therapist, but not too many or the game will be difficult/unclear to use. Also, find a way to let the game decide itself whether it should change its difficulty and if so, think of a good way to do this. Probably many games that I have played use some form of adaptive difficulty, but two examples that come to mind are FIFA Football and Mario Kart. FIFA did this in the worst way possible: after a few games I lost, the game suggested something like: “shall we set the difficulty to a more suitable level for you?” Well thank you very much EA, for pointing out that I’m a loser. Mario Kart on the other hand, gives losing players a higher probability of better powerups, improving their chances of winning. I think that these methods are not the best way to do this, but to hide this mechanic from the player, instead of what FIFA did, seems like a good choice.
Today, my colleague Coen and I went playtesting our current project Bloonies with a patient. The goal of the application is to train dual tasking: stroke patients can have difficulty with walking when executing a cognitive task. During this extra task they can slow down or even stop walking. The goal of Bloonies is to train this dual tasking. I have been reading some game design books (currently ‘A Book of Lenses’, love it!), but none of these books could give me the insight that I got from seeing this patient playing Bloonies and hearing his thoughts about it. Making health games requires a whole different kind of game design, one that I grow more fond of every day. I will share a video of Bloonies on this blog once it is finished and perhaps explain something about the way we achieved scalable difficulty in that project.
Thinking about the concept of adaptive difficulty, it makes me wonder if such a thing would be suitable for Caromble!. We are now too far in the development to incorporate such a thing, but it is an interesting thought experiment. I think that the difficulty in Caromble! is linked to two situations:
- Losing the ball; a higher difficulty makes the player lose the ball quicker and more often.
- Taking more time to complete a level; either to destroy all blocks or to solve a puzzle.
If we would incorporate an adaptive difficulty, we could influence both of these situations. Losing a ball can have multiple causes, but I think the main ones are a ball speed that is too high and blocks close to the paddle where the ball bounces off. If the player loses the ball too often or too quickly, the game could decrease the ball speed and destroy nearby blocks. A solution for a level that takes too long is to destroy some of the blocks after too much time has passed or to provide a hint system for the puzzles that offers these if too much time has passed. I think it is not smart to try and implement this into Caromble! now. These kind of features have to be incorporated early on in the development of a game; otherwise it can introduce balancing difficulties in terms of gameplay.
Also, the process of thinking with the patient, what I am learning through experience at Motek Medical, made me think of how this approach would work on Caromble!. Who exactly is our audience? In my opinion it are retro and indie gamers and perhaps (hopefully), some casual gamers can appreciate it (read: love it to the fullest) as well. Would the game be very different if we carried out an extensive study on the preferences and desires of these gamers? I think it can be very interesting to try this approach on a future entertainment game we make: to let go of our current assumptions and learn from the potential players.
Caromble! is a game that at least we enjoy playing a lot and we can only hope that many players feel the same. Since Caromble! is the first full game we create, I believe that is enough.