Tuesday, June 24, 2014
Sports, games, and simulation: Quiddich Professional League enters Alpha testing
I'm still alpha-testing my new boardgame, Quiddich: Professional League, which is part of my game series based on the writings of J.K. Rowling. Since this is part of a series, I felt it needed to keep at a complexity level close to the others. This, as it turned out, was a naive hope. Quiddich is just too nuanced, and needed more rules than I usually like to model it.
In the Harry Potter series, Quiddich is described in a number of places as being "very complex". Given the rules, however, it seems, while multi-faceted, to be simply layers of simplicity stacked atop each other. Given my love of designing abstract strategy games, this should have alerted me to the trouble ahead.
Known as emergent complexity, this is the idea that very complex results are often the result of a few simple rules systems interacting with each other. Complex and numerous rules, on the other hand, usually have the habit of constraining complexity. Using more and more rules to control the situation is the sign of weak design, in an abstract strategy game. Go is the gold standard for simple rules and emergent complexity.
Rowling's statement about the complexity of Quiddich proves true.
For a game like this, basically a simulation of a physical sport, I needed to model the movements and actions of fourteen players in four different roles, as well as the attributes and actions of four different balls. All moving in three dimensions, all at the same time. Each turn.
Keep it simple, I told myself, over and over. For instance, I wanted to keep it pure strategy by not using dice for action resolution. But that left me with the problem of players selecting from numerous possible actions by calculating the success of each and choosing only successful options. Talk about paralysis! The game threatened to get bogged down before I ever tested it. The solution: ignorance.
Not knowing a predetermined outcome is as good as randomness, so I hid a lot of the information from the opponent on cubes. You can see your information, but the other team cannot. Now there was no way to know if you'd succeed or not until you declared your action. You'd have to correctly predict your opponent's intentions, a gauge how much they wanted something to happen.
The 3D nature of flying players and balls I abstracted away, and I kept the field very few spaces, to force players into tight proximity. But the complexity of modelling a sport dogged me. Every element was modeled with the fewest rules I could manage and still perform its function.
Throwing a ball is a good example. There are two main types of throw in the game of Quiddich: throws form one player to another (a Pass), and throws through the golden hoops that score your team ten points (a Shot on Goal). It seems nit-picky, but it's essential to allow blocking and interceptions, as well as defending the rings by the Keeper. Oh, and then there are penalty shots. How do you manage strategy when there is only one shooter and one defender? Both players automatically have highest priority, so how can it be made strategic?
Taken together, this network of simple parts transforms into something else entirely. And that says nothing of the technical task of writing the rules to make this comprehensible to others. As prose, the rules are about fifteen pages long. Gak! I'm hoping when I do a full-art version, they'll boil down a lot.
Modeling a sport has really brought home how complex the rules of the average sport are. The reason sports are, from a game-designer's point of view, inelegant, is that there are just so many rules. But dealing with the real world in real time means you have to institute all sorts of controls. Even in the wizarding world, the game was sometimes hard to watch and understand without magical assistance, and I can only imagine how a Referee could do their job.
I wanted to allow basically every interaction that was shown in the book's Quiddich matches, and I believe I have done a good job of it, now that I've been alpha-testing for a week or so. The four types of players are all very useful, and each of the four balls in the game works completely differently. But, now that I have a working model of Quiddich, the true complexity of the shifting tactics has really come home for me. I wonder if J.K. Rowling realized how well her imaginary sport actually plays out, because, if my game is any sort of good simulation, the sport of Quiddich is a hugely deep, and very impressive piece of game design. I wonder what other games she might have designed, and just doesn't talk about much.
But I can't wait to start beta-testing, as offloading half the work onto another brain would be such a relief!