BrettSpiel is a blog about board game design, written by game designer Brett J. Gilbert.

2 Comments

Game Spaces: Why Everything Not Forbidden is Compulsory

In which I discuss the existential angst of the game designer, introduce the idea of the ‘game space’ and reveal what a central tenet of quantum mechanics has to tell us about game design.

I am very interested in the Universe. I am specialising in the Universe and all that surrounds it.

Peter Cook

“Most players are not like us.” That was Lewis Pulsipher’s recent observation about a group of playtesters who surprised him by entirely eschewing what he had considered the clear strategy of a new prototype. The statement reminded me of a post by Dave Dobson in which he discussed reactions to his game Yoggity by the judges of the Gamecon Memphis design contest: “One frustration — the testers apparently never traded anything, and trading is key to the game with more than two players.” He followed this up in later posts with some good discussion about the implications of the judges’ experience, and ended by saying “I wish they’d tried the trading, and I need to get people to want to.”

I think these comments reflect a realisation that all game designers come to sooner or later: that there is a limit to our control over the actions of players, and that our reach, in this sense, always exceed our grasp. Designing a game creates only the potential for a play experience that matches the designer’s intent. The designer builds a world, but can only ever give players a map of that world, not the world itself. And once you place your game in the hands of prospective players then you relinquish not only your ability to influence how it is played, but also your right to do so. That’s the deal: you give them the map, but they get to build the world anew, all on their own.

To better explain my meaning, the following diagram proposes an idealised view of this relationship. It also introduces a few concepts that will be useful later.

Game spaces: The game. The rules. The experience.
  • The game — A designer begins by imagining the game as an ideal set of actions and consequences that exist, at the beginning, only in the designer’s mind. This process involves the creation of a mental arena that I have labelled the ‘game space’; all that is part of the game lies within it, everything else lies without. And I mean ‘everything’ quite literally: I mean every conceivable action, consequence and idea that ever was, is or will be. By marking out the game space the designer claims a tiny yet unique territory of this infinite domain, and populates it with only those ideas needed to construct the game.
  • The rules — To communicate the game space’s contents and limits, the designer must create a map in the form of a set of rules. An ideal set would describe and delimit the game space perfectly, neither leaving any detail out nor including anything superfluous. They would be complete and exact; a precise circumscription of the game space that creates an ideal window through which players can view and comprehend the entirety of the game.
  • The experience — The game world built by the players is directed solely by the designer’s rules and is expressed in their experience of playing. This world is their ‘play space’ and if an ideal set of rules were to be interpreted perfectly, then that play space would precisely match the designer’s original intent, and every part of the game space — nothing more or less! — would be present.

From my descriptions you can perhaps tell how unlikely I think it is that any game, even the simplest, could be so perfectly designed, described or discovered. I have used the words ‘ideal’ and ‘perfect’ in the full and certain knowledge that ideals and perfection are, at least most of the time, impossible goals.

With this in mind, and to return to Lewis and Dave’s comments, I can wholly understand their surprise and disappointment at the mismatch between the game spaces they designed and the play spaces constructed and explored by their players. But I think that we all understand not only that ‘most players are not like us’ but that this is a truth not to be ignored, but to be enthusiastically embraced. The lesson for the game designer is to acknowledge this relationship between the designer and the player for what it is, and to build an understanding of its consequences into the game from the beginning.

And here, as they say, is the science bit…

It may be surprising that something called Gell-Mann's Totalitarian Principle of quantum mechanics has anything to teach us about game design, but take a look at the following explanation, which I quote without revision from Wikipedia:

The statement is in reference to a surprising feature of particle interactions: that any interaction which is not forbidden by a small number of simple conservation laws is not only allowed, but must be included in the sum over all “paths” which contribute to the outcome of the interaction. Hence if it isn’t forbidden, there is some probability amplitude for it to happen.

What this means is this: that in the obscure world of particle physics, the only way to calculate what’s likely to happen during any particle interaction is to properly take into account everything that could happen. Every conceivable outcome, barring a very limited set outlawed by ‘simple conservation laws’, not only can happen but must happen with some probability — but not only that! The bizarre nature of quantum mechanics tells us that not only must all these outcomes happen, but that they must all be considered to happen at the same time! It is only by averaging out all of these probabilities that we can accurately predict the behaviour of the world around us, which is something quantum mechanics does with startling accuracy.

And when physicist Murray Gell-Mann proposed this principle he summed it up by borrowing a wonderful phrase from T.H. White’s classic retelling of Arthurian lore The Once and Future King, a phrase used in the book as a form of warning to its protagonist about the dehumanizing nature of totalitarian regimes in which every instruction is followed without question: “Everything not forbidden is compulsory.”

I don’t think you need to look too hard at Wikipedia’s statement of the principle to see its application to game design. After all, it already talks about ‘interactions’, ‘laws’ and ‘outcomes’: three things intimately involved with games and their design. But just to be clear, my interpretation is this: that if you, the game designer, want to understand how players will play your game — if you want to accurately predict the behaviour of that world and your players’ interaction with it — then you need to look at every path that it is possible for players to take through that world, not just those paths that you would wish them to take. You can only ignore paths that are expressly forbidden by the constraints of the game system and your description of its rules. Every other path — every possible choice of action and every possible consequence of choice — must be considered. Only then will you know whether the game ‘works’ and whether the world that your players are likely to build bears any resemblance to the one that you have so meticulously designed and described.

We can use Lewis and Dave’s experiences as concrete examples, by considering what the play spaces implied by the behaviour and reactions of their players might look like. Few players confronted with a new game will immediately comprehend the entire game space, which means that the play space they immediately construct cannot perfectly match it. My three idealised diagrams above therefore do not, unsurprisingly, represent the common experience.

However, I think a generalized good experience of a game might be described as follows: players begin by comprehending only the most obvious sub-region of the game space but are then able to explore, over time and through continued play, and with a degree of ease afforded by the game itself, the remainder, reaching out from their origin until they have explored, if not the entirety, then at least a satisfying majority of the territory. I have visualized the game space representing such an experience in the left-hand panel of the diagram below.

Game spaces: Good experience. Weak experience. Bad experience.

But what of Lewis’ players, who seem to have missed the mark? Lewis speculates that they might yet discover the game as he imagined it through continued play, but their first play is likely to have created a weak experience — not broken, but thin and unsatisfying, and certainly not what the designer had intended or expected. As I say, this was after just one play, but a possible visualisation of that experience and of the play space constructed during it is shown in the centre. The players have explored some of the game space, but not enough to create an experience that truly reveals the game’s potential.

Unfortunately matters are possibly worse for Dave, since his players not only failed to explore the whole game space, but in doing so they missed the part that Dave considered the core of the game's experience: the trading. Fortunately for Dave the game survived, and the jurors still thought the game interesting, but if ‘the good bit’ gets missed by your players the risk is that the game may be prematurely written off as a bad experience and that players may simply never return, and therefore never have the opportunity to discover more of its true potential.

Although I think my physics analogy is useful, there is a very important sense in which it flatters to deceive. The laws of quantum physics are (more or less!) universal and immutable, even if they are not wholly understood, but the laws of a game, its rules, must be constructed and so represent yet another opportunity for the designer to stumble. And the fact that game rules may themselves not be wholly understood is just one more thing to worry about. The ways in which rules can fail are multifarious, but can be split into three broad categories: omission, inclusion and loophole, which I have visualised below.

Game spaces: The omission. The inclusion. The loophole.
  • Omission — Here the rules explictly forbid some action or consequence that should be allowed. This constricts the scope of the play space, preventing some part of the game space from being explored.
  • Inclusion — This is the precise opposite of the omission, in that the rules now explicitly allow something that should be forbidden. It is now possible for the play space to legitimately expand outside the scope of the game space.
  • Loophole — The effect of the loophole is similar to that of the inclusion, since the play space can now escape the bounds of the game space, but the logic is different. In this case the possibility of moving outside of the game space is neither explicitly forbidden nor allowed, rather the rules have created a ‘grey area’, a crack in the boundary drawn by the rules through which players can choose to play. Often players themselves will veto expanding the play space in this way by reasoning that to do so would break the ‘spirit of the game’, but there will always be others who seize the opportunity and point out, correctly, that no rule forbids it.

Often, if these flaws exist, then they exist only at the margins; indeed, when designers and players discuss ‘edge cases’ then you could say that the thin red line I have drawn is the very edge they are talking about! Large omissions, inclusions or loopholes are rare because they are easy to spot, but in more complex games smaller flaws become almost unavoidable. And in any case, players may deliberately choose to ignore or change, or may simply misunderstand, even the most well-formed rules and so contort their own play space in a way that is entirely out of the designer’s control.

As I said earlier, the notion that ideal rules and perfectly constructed play spaces can even exist is something of a pipe dream. Rules and play spaces will, except for all but the simplest of games, always be imperfect; it is only the degree of imperfection that we can seek to limit.

The good news is that I definitely believe that it is possible for game designers to protect their games from all these various and unfavourable variables. Indeed, I rather think that it is the designer’s duty to do so.

The bad news is that there are no quick fixes. The care, precision and clarity of thought necessary to create games, rules and experiences that are as free of these problems as possible is neither easy to attain nor simple to apply. But let me be clear: just because a particular game can deliver a good and rewarding experience does not mean that it will, and it is simply unacceptable to reward players who complain that your game delivered a bad experience with a shrug of your shoulders, and then offer the advice: “you’re not playing it right.”

If it’s possible to ‘play it wrong’ then the design, and the designer, have failed.

Anonymous said...

January 14, 2011 5:27 pm

Good stuff.

I particularly like how you've related the creation of rules of a game engine with the laws of physics. Although, it may result in designers developing god complexes.

David Brain said...

January 18, 2011 10:08 am

Lovely piece as ever, Brett, with almost too much to think about.
Some time ago, a friend of mine suggested that games comprise "laws" and "rules". Laws are the things that work logically and can often be deduced ahead of time (there is a discard pile, therefore you discard cards o it.) The Rules meanwhile are those things that, as you have noted, circumscribe the world in order that the game "works" and does not comtain loopholes. But they are the things that you could not possibly know ahead of time.

The trick is that although both the Laws and the Rules need to be spelled out, neuther of them sould necessarily directly instruct the player in "how to play".
Naturally, this leads to the initially described scenarios, and to those games that feel the need to include "tips for play" as a separate section in the rulebook because certain aspects of the game may not be immediately apparent at first glance.
And the designer's dilemma is, as Pulsipher observes, that the first game needs to be weak but not bad. Preferably it should be a good weak, but it must leave the players thinking that there is more to discover here and, crucially, eager to discover it.

(Of course, the real test is whether, having discovered more, the player feels that (a) it was worth it, and (b) still wants to play!)

Post a Comment

Older post / Newer post