Videogames are one of the most marvelous mediums when it comes to creating a fancy layer around the intellectual challenges posed by abstract puzzles. Game designers work very hard to create new, innovative representations and interfaces ("gameplay") around ageless logical recreations, puzzles. In the old gamethink tradition, let’s take a few steps back to analyse what different kind of puzzles we are talking about and hopefully draw some interesting conclusions for budding puzzle/game designers.
From arbitrary guessing to intuition
Guess or die
- Shadowgate, NES, 1987
- If you touch the candle, you will die. You had no way to know that you shouldn’t do that.
The lowest form of puzzle is one where the game asks the player a question with several choices, all of which but one being losing choices. The game doesn’t help the player by providing hints or anything that would back up their decision. Popular up to the 80s, this kind of puzzle that questions the ability to decide rather than that to judge, and that tests frustration rather than attention, is unsurprisingly one of the first victims of natural selection of the evolution of game design. Often rejected, as perceived as unfair, it still appears now and again under various formats, often for optional sub-games, gambles and so on.
Explore the combinations
Another extremely basic form of puzzle, this one requires the player to find the right combination of simple gameplay actions in order to progress. In this context, the number of actions the player can undertake is strictly limited. Sometimes, the player must find the right order in which to perform a certain number of tasks, which can or cannot be repeated. In other terms, the player is confronted to a number of combinations, one of which works, and has to test them in sequence until they find the one that works. This puzzle tests the player’s ability to list all possible actions and to be able to go through that list sequentially, without repeating a sequence which they know that it doesn’t work. This very basic construction is found very frequently and in a variety of shapes and sizes. It is often the core mechanism in text or point&click adventure games, when the hero has, by nature, limited choices and must perform them in the right order to progress in the game. This can be done in a more or less natural way, for instance using an item or performing an action may grant the hero access to a new area where new actions are possible. Or (and this is typical of the notorious Runaway for example) talking to a character or performing an unrelated action might enable the hero to do something which they couldn’t do before, although they wasn’t any physical impossibility. The guideline in this case should be, when the game prevents you from doing something or conversely when it allows you to do something that was previously impossible, there should be a logical reason for that, but at the very least it is the game’s responsibility to signal its change of policy.
- Captive, amiga, 1990
- In the classic RPG Captive, the player must open doors like this one by guessing the order in which to press the buttons on the corner. An example of combination puzzle.
This kind of puzzle can also be used in many other games. In the classic RPG, "Captive", some locked doors could be open by pressing 4 buttons in the right order (24 combinations) or by keying in the right combination of eight buttons (255 combinations).
Deduction
When a player has to explore combinations, that’s because there is no reason to try one first. When there are sufficient hints, the game would reward the player who perform tasks according to the set of rules they know. That is deduction: a finite situation to solve and a well-established set of rules. This is the most common form of "elaborate" puzzle in games today.
In Vagrant Story, in a recurring sub-game, the hero, Ashley, had to push crates in a correct position. But various crates followed various rules: some could be broken, not others, some could only be moved a few times, etc, etc. When they saw a room with crates, and with their knowledge of the game rules (which were explicitly detailed in the online help) the player could deduce the solution to the problem.
- A Link to the Past, GBA, 2003
- A wall that looks like the one on the left can always be broken. Blocks that look like the one towards the top of the screen can always be pushed forward. Knowing these rules, the player can solve most puzzles.
In Zelda, many of the tasks that the player has to perform also fall in that category. In the classic "A link to the past", for instance, the legendary hookshot could be used to cross gaps. All the hero had to do was to shoot it at a stake. Conveniently, all stakes looked strictly identical throughout the game. Therefore, whenever the player saw a stake, they could deduce that they needed to use the hookshot. (Same observation for cracked walls and bombs, unlit torches and the lamp, blue bumpers and the magic cape, etc.)
Analogy and intuition
In this far more interesting puzzle category, the player is invited more or less explicitly to imagine a solution based on a situation, the resources at hand and what they know. However, what they know and the written, explicit rules of the game are two different things and that difference is precisely what makes those puzzles interesting.
To keep on with the Legend of Zelda examples, most inventory items have, beyond their obvious usage, a certain number of extra functions and it is up to the player to imagine them. It wasn’t obvious for many of the "A link to the past" players, for instance, that the Bow - which everyone agreed was a great tool to kill enemies from afar or to activate the occasional remote switch - was also the item of choice when confronted to statues. Yes, statues. Why - Just shoot them in the eye! In "Ocarina of time", when the player is introduced to (yes!) flaming arrows, it’s pretty obvious that they damage enemies and that they can set fire to flammable items such as torches. Then, the player is confronted with ice blocks, ice enemies, etc. It’s not in the game manual, but as the player is supposed to know that "fire > ice", they’re supposed to find out that a flaming arrow will melt an ice block. They’re also supposed to figure out, later, that firing an ice arrow into water will create a little ice platform. In "The Twilight Princess", the ball and chain looks like an impressive weapon. The player is supposed to deduce immediately that swinging it towards any breakable item will break it. And then, later, that shooting the heavy ball towards a movable platform will put it in motion - just based on what they know on physics.
Often, these type of puzzles rely on observation, or the capacity of the player to see that a detail in the gameworld is different from "normal" - although normal is not explicitly defined. For instance, this wall feels different from the other walls - It could hide a secret door; those two statues look almost the same. If it were possible to make them exactly identical, something may happen...
One very common, almost canonical example of observation-based puzzles is boss battles. Most action-adventure games are divided in broad levels, each ending by a battle against a boss, which can only be defeated by finding its weak point. The player has to analyse its behaviour, find patterns, and form hypotheses as to how to beat the monster. When the player tries one specific strategy, there is no guarantee that it will be successful - the player is not deducing but inducing.
Management and configuration In the previous categories, the puzzles had one single solution: pass, you win, you continue; fail, you die, you lose, you’re stuck. But most often, the way the game will pick you brain is not so binary. It will constantly ask the player to take informed decisions, where most of the choices are valid, where some will have adverse consequences and a few could end the game outright. For instance, let’s imagine an action game such as an FPS or a survival-horror game. Suppose your character has plenty of ammo but little health. They spot a group of enemies in the distance. Won’t you try to attack them from afar? If the reverse was true - low ammo but high health - won’t you try to avoid them altogether and not waste ammo? In such games, you spend your time shooting zombies, or soldiers, or both, while trying not to get killed. But in fact, you are managing a situation according to a few variables, such as your status, that of the enemies, everyone’s relative position, cover, etc. Depending on what those variables are like, you are invited to take this or that decision which will affect your long-term success. So even in these seemingly action games, you have to devise a strategy. If you go a step beyond, in some games, more attention is spent in the preparation phases (ex. optimizing inventory or skills for heroes, picking the units of an army) than in the actual action phase. That is more and more the case with modern RPGs where character progression is more and more complex. In all these games, the real challenge lies therefore in these preparation, or configuration stages, rather than in the real-time part of the game.
Integration in the game
As we stated earlier, videogames are a fine abstraction layer for puzzles. That being said, all games are not equal in that domain. In some, it shows that puzzles have been added to a game structure that would otherwise work fine without them. In others, the puzzles are so naturally integrated that the player doesn’t feel that difference.
In the most extreme examples, the game interface completely changes to a screen dedicated to solving a specific puzzle. This is the trade mark of Resident Evil, and is frequently seen in games it inspired.
On the other hand, in Zelda, the items which are used as weapons or tools in the main game are the same which are used to solve puzzles. As mentioned above, the bow, which is a fine ranged weapon, is also used to trigger remote switches. In that case, there is no separation between "action mode" and "puzzle-solving mode". If this integration is well done, then it is possible to put much content in the game, including real intellectual challenges, while maintening the verisimilitude of the game, without showing that the puzzles have merely been added to a game that didn’t really need them.
This integration of the puzzle-solving mode in the main game is natural in an real-time game like the Zeldas, but it is also featured in RPGs of a more classical vein such as in the Lufia or Golden Sun series. In both games, the player can choose an action to assign to an action button, which they can use while exploring the world. This action (for instance shooting arrows in Lufia, or growing plants in Golden Sun) can then be used to solve puzzles.
- Half life 2, PC, 2004
- In this situation, by putting heavy objects on one end of the plank, you can balance it and climb on it for freedom!
More and more frequently, game developers are turning to a physics engine to allow players to solve situations in creative ways, without separating the main (action) mode and puzzle-solving mode. The best example of such a game is probably Half-Life II, where you can take, drop, or throw about every object and exploit the reactions of the environment to your advantage.