The Problem with Adventure/Puzzle Games

Adventure puzzle games all suffer from an intrinsic problem: the puzzles there-in are contrived by the minds of man, and are not resultant of behaviour that emerges from a set of observable rules. As such they never really engage on the same low level of the mind as, say, a physics puzzle game.

This leaves designers with a fundamental puzzle of their own.




Truth In Game Design

At GDCE 2011 Jonathon Blow, creator of Braid and the upcoming The Witness, gave an excellent talk entitled Truth in Game Design. The talk was very insightful and took a long time to say something quite simple, which I shall paraphrase here:

Puzzles tug on the cognitive processes of affordance-identification best when they are borne of the observations of simple rules that create complexity. 

Rather than design puzzles, designers can create more intuitively engaging problems by creating the simple rules of a game-space, observing complex and unpredictable outcomes, and then reverse-engineering those phenomena to invite the player to create conditions which lead to their emergence.

Play and Consistent Observable Behaviour

A great book, although I don't agree
with all of their definitions

Humans are very good at experimenting with well-defined and constant rules in order to produce a favourable outcome. This process is commonly known as learning, and the iteration of inputs to these rules is called play. A simple example is a child playing with building bricks: the rules are those of physics, and the iteration is the act of stacking and re-stacking the bricks. The outcomes observed are whether the stack falls down, leans, or stands upright.

This process of learning is hardwired at the lowest levels into the human brain after millions of years of evolution. Even without stone implements people needed to understand how the most base of all tools, the body, reacted under different conditions. Becoming proficient in sport is a process of being able to more accurately predict the outcome of the combination of input and circumstance.

Opaque Behaviour

People are quite good at sussing each other out too. The taradiddles of many people's social lives does provide empirical evidence that as a species we're not terribly good at it. Society functions as result of a community acting with reasonable predictability, but there are often cases from the trivial (someone leave the pub in a tantrum) to the significant (sociopaths) that prove time and again that predicting and understanding the thought patterns of others is haphazard at best.

A good HCI introduction

It is also argued that a person cannot introspect their own cognitive processes. This is why asking people for insight into their actions in usability tests (or indeed after spur-of-the-moment events like reacting to violence) is fruitless. I shalln't recite the reasons why here - those interested can go and read Donald Norman's  Emotional Design, Design Of Everyday Things and read up on cognitive psychology - but if the field of psychology generally agrees that we can't reliably explain our own thought processes, then the chances of us being to explain those of others are similarly slim.

In human-computer interaction terms, the behaviour of others is less predictable because their state is less transparent. A stack of bricks immediately shows its state, meaning less interpretation is needed and a more accurate prediction of events can be made.

Contrived Puzzles

Puzzle games like the Monkey Island series, the Dizzy series and my own Clover: A Curious Tale rely on players being able to second-guess the designers' understanding of how objects (often identified by little more than a few words) interact. There is little or no 'organic' interaction between them like you might see in a game like Portal.

Sorry!

The lack of natural and observable interactions between objects is one of the reasons why 'combining item' puzzles are always infuriating. Why should I be able to combine these two items, but not any other two? Hopefully there'd be some hint in the description of the items - I'm guilty of this with an arrowless bow and an aerodynamic plunger. There's no observable reason why a rubber-chicken-with-a-pulley-in-the-middle should suggest use down a zipwire.

Most adventure/puzzle games do make limited use of more natural problems, normally involving the "I can see it, but how do I get there" mechanic. Clover: A Curious Tale had a small example of this with a drawbridge that could only be opened from one side.

If adventure/puzzle games can never be 'pure' then are they truly relics of a bygone era? An evolutionary offshoot that survived whilst competition was weak, but was driven to extinction when more successful beasts encroached on its turf?

Adventure/puzzle games clearly work, otherwise the genre as a whole wouldn't have sold millions of units. I suggest this is because they work at the reflective level and not the behavioural.

Reflective Play Spaces

Donald Norman's Model

Narratives are the prime example of reflective possibility spaces. An author introduces a set of characters and circumstances, and the imagination of the consumer is set ablaze churning over the possible outcomes of what might happen if Character X's secret gets found out by Character Y. Interestingly one can consider this entirely hypothetical play: evaluating inputs and circumstance to second-guess outcomes, when it is impractical to empirically test these ideas.

If we concede that adventure/puzzle games' real arena of play is the mind of the player and not the consistent rules of some virtual world, what does this mean for gameplay?

My argument is that adventure/puzzle games in their truest form should be stories first and foremost, with as little 'gameplay' as possible. The play occurs primarily in the players understanding of all of the constituent elements of the narrative, and not in the rules of the on-screen components.

Players of these games take primary joy from the play in their imaginations, and any hybridisation of game mechanics will alienate this audience to some extent (although hopefully attracting those of a different disposition). The game world and its layout should be considered an aid to the imagination of the player, and less a manifestation of the interacting elements within.