The Defining Cycle of Super Practica

Super Practica is defined by the process of its optimization rather than by any particular design or product. That is why it is both a set of games and a scientific research program. Both are integral parts of a single process.

In order to explain this process, it will help to look at the two processes it’s derived from, used in science and engineering.

The Cycle of Effective Engineering #

It’s common knowledge among effective engineers that, if you want to build something good, you should iterate. There’s room for disagreement about which steps should be considered essential in the iterative cycle, but I identify these: 1. Design, 2. Develop, 3. Test. Then redesign, redevelop, retest, and so on.

I’ll call this the cycle of effective engineering, because it seems to me that all effective engineering must follow this cycle, whether formally or informally. Otherwise engineers can only succeed by being very lucky or by copying an existing solution.

Development is of course necessary to get anything done, but why these other 2 steps?

To design means to get a plan for solving a problem in terms of a schematic or specification. It has much practical value in coordinating developers to work together following the same plan. But even more importantly, it includes taking a step back to ask what exactly the problem is and what exactly would constitute a solution to it. (This is sometimes called “design thinking”.)

Testing is necessary to check if you have indeed solved the problem you intended to solve. It also helps to obtain valuable information for continuing iterations. If your only goal is to sell a product, then you can do that by selling garbage to unsuspecting buyers, but the way to make sure that your product actually solves the problems you set out to solve is to test that it solves those problems.

If this iterative cycle is allowed to go on indefinitely, then its end becomes not just success, but optimization. What exactly gets optimized is determined by the selection of tests. So it is vital to select the right tests.

Testing to Inform Game Design #

Systematic and formal testing is less common in game development than most other kinds of engineering. Beyond bugtesting, how can you test the quality and success of a game? Are good games not just a matter of personal preference?

Saying that game quality is reducible only and exactly to personal preference amounts to saying that the reason (or reasons) anyone likes the games they do is necessarily mysterious and unknowable. But there’s no reason to believe that there is anything essentially mysterious in the experiences people have when they play games. In fact, if we couldn’t recognize patterns in how people commonly react to the various situations that games present, we could hardly design good games at all!

Disregarding confused intuitions and philosophies on this point, the history of science shows clearly enough that mystery amounts to ignorance and a lack of expressive language. (Consider how mysterious chemical processes must have seemed to early alchemists!) It is only for lack of effective analytic vocabulary and theory that we fail to make explicit and testable our feelings about games.

When game designers observe playtesters playing games, they aren’t waiting for mystical insight but are rather looking for information that could answer their questions or suprise them. Good game design follows from good information, and good information follows from good testing, hence good game design follows from good testing. This is true for informal testing and will remain true for formal and systematic testing.

While I argue that it’s possible to optimize games by systematic testing, I don’t mean to say that it’s easy. Attempted naively, the likely outcome is a lousy selection of tests, a game worse than could be developed by following intuitions, and a massive waste of time. It also requires the development of scientific theory.

The Cycle of Effective Science #

That ignorance and a lack of expressive language should be so keenly felt when attempting to inform game design with systematic testing, shows how much designers and engineers rely on scientists. It would hardly be possible to organize effective engineering work without scientific theories providing both precise vocabulary and reliable empirical expectations.

Empirical science proceeds by its own cycle, which I reduce to just 2 essential steps: 1. Theorize, 2. Experiment. Then expand or revise the theory to accomodate the results of the experiment, then predict and test experimentally new predictions of the revised theory, and so on.

Calling the structure of scientific progress a cycle implies a fallibilist understanding of science. The philosophical commitment is that nothing should be considered unrevisable bedrock to build science from. We can only know which principles are solid by seeing them continuously stand up to challenges against them.

Just like the cycle of effective engineering, the end of this cycle is optimization. We can understand the progress of science as the gradual optimization of our theories about the world.

The Development Cycle of Super Practica #

These 2 cycles can function well enough apart, but the goals of Super Practica will be served best by combining them together.

Developing scientific theory on games will be valuable to inform game design. But we can go further and base design entirely on theory using a logical method of design. Following this logical method has a few important benefits. First, it improves feedback from testing, since tests can be used to give information about general cases of design instead of only about the particular cases tested. Second, it allows explaining why the resulting designs are considered optimal, beyond them being the product of an optimizing process. Third, it improves designers’ ability to cooperate effectively and at scale, since coherent design thus follows from logical argument achieving consenus rather than from personal vision or popularity contests.

So the Design step benefits from being preceded by a Theorize step. Likewise, the Theorize step benefits by being inserted into the cycle of effective engineering, at least when using the logical method of design.—The same tests which test the design and the implementation can thus test the theory too, and good experiments for relevant theories are often expensive to run.

Thus we derive the 4-step development cycle of Super Practica:

The Selection of Tests #

Super Practica being defined by this development cycle makes it both a program of game development and a program of scientific research with the end of optimizing both. But what exactly does it optimize about them? That is determined by its selection of tests.

The empirical framework of Super Practica consists of 3 types of tests: Playability tests, playthruability tests, and transferability tests.

Playability and playthruability tests tell if players can reliably and efficiently play through a game from start to finish. Optimizing for both of these makes a game that is minimally frustrating and which a maximum number of people can complete in the shortest possible time.

And transferability tests tell if the goal of reproducing practical knowledge is achieved. So optimizing for all 3 makes an optimal method of reproducing practical knowledge, together with the requisite scientific theory to do so.