Puzzle Design & Decoder Reading

Each video game is a world all to itself. Each world contains its own rules and lexicon. Overcoming gameplay challenges requires player engagement and usually the player to look at the game screen (exceptions include Rhythm Heaven and memorizing levels). Through visuals players learn to recognize objects, symbols, and other important interactive visual elements to help them make informed player choices and strategize. Visual consistency and form fits function help create an interactive world where what you see is what you get. In this way players learn to read a game. 

Just reading my DS under a tree on a lovely day.

More so than other genres of video games, puzzle games have a much higher reading to action ratio. As a huge puzzle game fan, I've often stared at a puzzle for 20 minutes or longer coming up with a successful approach without ever making a move in the game. Many puzzle games feature a series of challenges created out of a few elements, rules, and dynamics. Each challenge is clearly presented and the player must figure out how to satisfy the winning condition(s). Sometimes players must clear/use all the pieces. Other times players must line up specific pieces in a limited number of moves. And other times the goal is to simply get to the end of the stage with your avatar character. 

Before I go on, I must briefly cover the different puzzle game modes. There's what's commonly referred to as a puzzle mode, a challenge mode, and an endless mode. With a little ingenuity any of these 3 modes can be designed for single or multiplayer. For the purposes of this article, we'll focus entirely on single player puzzle experiences. 

• Puzzle mode. The levels in this mode use the game's core puzzle/gameplay system and feature challenges with very limited solutions. The purpose is not to give the player the freedom to play around with countless, emergent, and divergent possibilities while exploring the results of their choices. Rather, the purpose of the challenge in a puzzle mode is to limit the player in such a way to test their knowledge of the puzzle engine and their logical strategies. Puzzle mode levels typically deemphasize real time, action based skills (timing, reflex) in order to completely test the player's knowledge, logic, and deduction skills. This means that many puzzle modes are turn/move based. Like a clever joke or plot twist, the challenge and engagement comes from learning the solution/punch line. Usually, once a puzzle mode level is completed, there's little challenge and therefore incentive for players to replay it. 
• Challenge mode. These levels take the puzzle engine and work in more real time elements into the challenge. The simplest examples involve completing levels from puzzle modes in a limited time. Other examples involve manipulating a fixed level that runs on a predictable cycle until complete.  
• Endless mode. This mode is like challenge mode except that the level arrangements/challenges are randomly generated. Every time you play, the level is different. Keep in mind endless mode doesn't mean you play until you lose. Rather, compared to the very limited replayability of puzzle mode, endless mode has endless replayability. 

LEARNING TO READ

Reading a puzzle mode puzzle (referred to from this point simply as a puzzle) involves simultaneously reverse engineering the challenge from the winning condition and deductively eliminating extraneous elements, moves, and strategies. Putting a deep, working knowledge of a puzzle system to good use involves visualizing the solution as the option that remains after eliminating erroneous options. There are usually far more ways to fail a puzzle than to solve it. After all, it wouldn't be much of a puzzle if it were the other way around. Learning to read involves converting many 'if's into one (or a few) 'it must be.'

If it can't be this, or this, or this, then it must be THIS!

The following story refers to the puzzle mode in a popular Nintendo game Tetris Attack/Planet Puzzle League. The puzzles system has been imitated in this flash game. Play a few levels to better relate.

When I was very young playing through the puzzle mode of Tetris Attack on the SNES, I only had one basic approach or strategy: to imagine the move before I made it, and try to imagine as many moves into the future as I could. Even with the ability to visualize several moves ahead, my approach was still rooted in trial and error. In other words, between my manual attempts and my mental projections I might consider every possible move for a puzzle before stumbling onto the solution. This is an ok strategy that works well enough for the easiest beginning stages (stages with 1 to 2 possible moves). But soon, the number of permutations or possibilities I had to go through became far too many to be successful with this method.

Soon after but still a kid, I developed a more advanced strategy. If a 3 step puzzle contained too many possibilities to work out, by eliminating a step from my strategy, I could reduce the puzzle to a 2 step puzzle and use my previous strategy. Still working with relatively simple puzzle levels, I learned to read the puzzle to find out the final move/state necessary to win. This strategy was effective, but I eventually hit a brick wall and gave up.

Years later, I revisited the same puzzles in a remake of Tetris Attack titled Planet Puzzle League on the Nintendo DS. Now much older and much more experienced in puzzle solving and complex logic systems, I was able to progress very quickly past my previous best. I realized that visualizing every possible move is like tackling a delicate problem with brute force. To solve more difficult puzzles with finesse, I needed to learn how to weed through all unsuccessful possibilities with theory so I wouldn't need to waste time visualizing them. I needed to learn how to read the dynamic challenge instead of the individual pieces. By developing and keeping a few simple rules in mind, I could solve far more difficult puzzles with greater ease than ever. And this is how I did it.

READING PLANET PUZZLE LEAGUE

The puzzle mode in Planet Puzzle League is limited by moves. In so many moves, you must clear the field. To clear any panel, you must match their colors/shapes in lines (vertical/horizontal) of 3 or more. After panels are cleared gravity pulls any floating panels down. From these simple rules I developed the following "rules of thumb" to read the puzzles.  

• Panels that are stacked on top of other panels are rich with potential energy. When they fall (and they must fall or be cleared) they can land on the right panels and be cleared in a chain. 
• Diagonally lined matching panels are already in a good position to be cleared. When the panels around them are cleared they'll fall into place.
• If possible, make every one move (panel flip) move two panels into two places you want them to be. Doing so can effectively double your move count. 
• If a puzzle contains 3-5 panels of a single color, they must all be cleared simultaneously. If there are 6 or more of the same color, they will most likely be matched in in two separate combos.  

Instead of thinking in thousands of possible move combinations, I can read a Planet Puzzle League puzzle with these four concepts in mind. When I look at the puzzle, instead of being overwhelmed, I'm calmed as my brain lets go of my inability to process thousands of move combinations. By thinking in a handful of concepts, I can read the game according to function, rules, and gameplay dynamics.  

THINKING ABOUT PUZZLE DESIGN 

The best puzzle systems can facilitate deep readings based on the coming together of mechanics, nuance, and gameplay dynamics. The best puzzle modes feature a series of challenges that cover a wide range of reading techniques showcasing the depth and flexibility of the puzzle system without bogging down the player with too many inefficient complexities or levels that are too similar. Boxlife and Zengage are good examples on the DSi. Minim, Use Boxmen, I wish I were the Moon, 3D Logic, and Exploit are all great examples that are also free flash games. Go here for links. 

Looking at how a puzzle can be read (depth) and how a game teaches players to read (development) are useful when critiquing a puzzle in any game. Video games of other genres have been incorporating puzzle modes and challenges since Super Mario Brothers (bonus levels). Remember that the difference between a puzzle challenge and a normal gameplay challenge is hazy and flexible at best. 

Many action and adventure games feature puzzle sections to break up the pacing. This can be highly effective because good puzzles are very logical and the process of reading them gives players the ability to engage with a single challenge at a (most likely) slower, self controlled pace. The best implementations of puzzle challenges are when a game teaches players how to read through a series of challenges that increase in difficulty like in The Legend of Zelda series. See the challenges of an entire dungeon of Twilight Princess detailed here. 

On a quick side note, I don't think there's such a thing as a well designed platforming puzzle (ie. a puzzle mode puzzle). At least, they're very rare. To be clear, there are plenty of excellent puzzles and puzzle games that are designed with platforming elements (see Braid one of my GOTY for 08). But creating a platforming puzzle inherently involves creating a challenge that must be overcome using platforming mechanics. Platforming mechanics are inherently action based making any platforming puzzle challenge fall on the side of challenge mode rather than puzzle mode. The key question is how can you read a platforming puzzle? Essentially, there's either mandatory "platforms" in plain sight, or hidden/obscured "platforms." If the "platforms" are obvious, then the "platforming puzzle challenge" is just a normal platforming challenge. Obscuring the mandatory path/platforms just make things more laborious. For 3D "platforming puzzles" the mandatory "platforms" are easily obscured behind level architecture. Moving around the environment to find the necessary "platform" is not reading. Doing so is more like... looking. I assume that none of us consider looking for a light switch in a dark unfamiliar room as solving a puzzle. Mirror's Edge is filled with frustrating and confusing challenges like this.

RIDDLES VS. PUZZLES

Puzzles that are designed or implemented poorly, usually offer shallow reading experiences. Sometimes, trying to read such a puzzle is functionally like decoding the message instead of reading it. The problem is, you don't have everything you need to properly decode it. Such puzzles are more accurately labeled as riddles (a type of puzzle). To understand what I mean by this, we'll need more definitions. 

Riddle

• 1. a question or statement so framed as to exercise one's ingenuity in answering it or discovering its meaning; conundrum.

Ingenuity

• 1. Inventive skill or imagination; cleverness.

Conundrum

• 1. a riddle, the answer to which involves a pun or play on words, as What is black and white and read all over? A newspaper.

Riddles are designed to challenge one to think outside of the box and/or to make long and uncommon connections between ideas. Solving a riddle often comes with a feeling of invention or a sense that one has a discovered the solution. Unfortunately, developers often design riddle puzzles that force players to "figure it out" without offering enough clues or teaching them to read the presented conditions. Essentially, if you don't know how to tackle a difficult problem and you don't have effective help (clues) all of your efforts will be trial and error work. If the best strategy you can hope to develop for a challenge is trial and error, then the challenge isn't deep or engaging past an elementary level. Some riddle puzzles do it better than others. 

Crimson Room? More like Crimson Riddle

The old school adventure games like The Hitchhikers Guide To The Galaxy- The Adventure Game and the series of flash games starting with Crimson Room are excellent examples of games filled with obtuse and difficult riddles. Go ahead. Try them out. There Is Only One Level is an example of a riddle puzzle game with very effective clues. It's not very deep reading wise, but it's very creative and unique. 

To complete the metaphor and to create an analogy, if solving a puzzle mode challenge is like reading a poem in a language created out of game rules, mechanics, and gameplay dynamics, then solving a riddle puzzle is like cracking a coded TV commercial slogan sometimes even without a decoder sheet. In other words, it's decoder reading. 

In the end, the concept of reading is about the same as developing a strategic eye for any game of strategy and logic. I've learned how to read extremely deep strategy games like Advance Wars and fighters like Smah Brothers. But reading is the cleanest when reading puzzle mode levels.