In this series I’m blogging about the development process for Delugional, the puzzle game I’ve been working on.
There a few broad choices you can make when you start to design a puzzle game. For one, you can broadly divide puzzle games into two categories: ‘authored’ and ‘generated.’ By ‘authored’ I mean a game where every puzzle has been designed by a human. Think World of Goo, or Cut the Rope, or Kami. By ‘generated’ I mean games where every puzzle you play is a fresh creation, generated in the algorithmic bowels of the game’s code. Think of Threes/2048/etc, or Minesweeper, or Dots.
From a player’s perspective, each mode offers a different appeal. In a game with hand-authored puzzles, there’s typically a finite number of puzzles, giving the player a clear goal to attain: solve all the puzzles (as well as sub-goals like finish one set of puzzles, get ‘perfect’ on all the puzzles, etc). When that final goal is attained, however, the game’s over.
Generated puzzles offer theoretically infinite replay value, since each new puzzle is unique. The trade-off is that you don’t have the larger-scale goal structure, so there’s no obvious point at which you’ve ‘beaten the game.’ (The same applies for freeform non-puzzle games, like SimCity or Crusader Kings.) Also, the infinitude is only theoretical; in reality, replay is limited by getting bored. People are born-and-bred pattern detectors, and we’re good at detecting the kinds of repeating features that are a signature of algorithmically-generated content. (Counterpoint: we love doing the same thing over and over, if we get some kind of virtual reward for it.)
From a designer’s perspective, too, each mode offers a different challenge. Generated games tend to require a strong, simple core mechanic. The puzzle-generating algorithm has to be able to reliably produce a satisfying game each time, which restricts the complexity of the puzzle’s rules. (Imagine a variant of chess where you placed all the pieces on the board in random locations. How many games thus generated would land in the ‘sweet spot’ of being neither trivially easy or impossibly hard for one of the players?)
Hand-authored puzzles allow more variety in game mechanisms, because you as the designer can control the interactions between different mechanics. The burden of game design is shifted from finding a single, simple mechanic which is strong enough and has enough ’emergent’ potential to sustain a whole game, to being able to take a mechanic and explore its potential in combination with complementary mechanics and variations on the theme.
Considering the trade-offs, I tentatively decided to aim for an ‘authored’ style. In the next post I’ll talk about prototyping ideas for the basic puzzle mechanic.
