Advertisement

Puzzle design in Adventure games

Started by November 11, 2010 07:37 AM
5 comments, last by topright 14 years, 2 months ago
Creating interesting, original, logical, thoughtful puzzles is an art.

What techniques and methods do you use to create good puzzles?

Do you use any diagrams to document puzzles and their dependencies?

How do you balance difficulty of puzzles?

Do you write storyline before designing puzzles or combine these 2 processes?

---------------------------------------------------

Some general inspiring approaches to create puzzles:

1. One possible inspiring method is to use puzzle classifications.
Adventure UML (Visual representation)
Adventure UML (Source article)

Mark Newheiser's classification
"Making Better Puzzles" by Stephen Granade

All these classifications can be used to stimulate your creativity and brainstorming.

2. TRIZ - "a problem-solving, analysis and forecasting tool derived from the study of patterns of invention in the global patent literature". It was developed by Soviet engineer and researcher Genrich Altshuller and his colleagues, beginning in 1946. In English the name is typically rendered as "the theory of inventive problem solving", and occasionally goes by the English acronym "TIPS".

TRIZ is variously described as a methodology, tool set, knowledge base, and model-based technology for generating new ideas and solutions for problem solving. It is intended for application in problem formulation, system analysis, failure analysis, and patterns of system evolution. Splits have occurred within TRIZ advocacy, and interpretation of its findings and applications are disputed.
TRIZ (wiki)

3. Edward de Bono's methods of creative thinking.
Edward de Bono (wiki)

---------------------------------------------------

[Edited by - topright on November 11, 2010 1:09:55 PM]
Interesting topic.

Answers to your questions:

1) What techniques and methods do you use to create good puzzles?

Technique: Scrambling the Solution
1. Define a set of operations for the player.
2. Use reverse operations to mutate a state. Mutate more for harder puzzles.
3. Define the state before as the solution state, after as the problem state.
4. Present the problem state to the player.
5. Optional: Describe the solution state for the player to make it easier.
6. Optional: Describe any intermediate state to make it easier.
7. Optional: Describe any required operation to make it easier.

Characteristics of good puzzles:
o The effects of a basic operation are observable.
o Objects with different effects under the same stimulus are discernable.
o Operations (Action, Object, Effect triplets) are intuitive or memorable.

Characteristic of intriguing puzzles:
o Use a problem state similar to the solution state, but with no direct solution.

2) Do you use any diagrams to document puzzles and their dependencies?

Yes, especially when I don't know whether the puzzle can be understood,
or when I don't know the full set of operations for solving the puzzle.
This usually happens when I start with the problem state, and I don't
know what the solution is myself, but I know that the game ought to have
that problem. The purpose of the diagrams is to study the problems, to
discover the operations and the solutions.

3) How do you balance difficulty of puzzles?

You could always make a puzzle be optional. So I guess the question, in the
context of adventure games, is how do I know whether a puzzle is too hard
when it is a required puzzle. I suppose the right way of doing this is to
play test. If I am just doing it on my own I would tune it to something easier
than what I was willing to handle, because as the designer I have more
interested in solving the puzzles than the actual player. The hard puzzles
become optional puzzles.

4) Do you write storyline before designing puzzles or combine these 2 processes?

I have a design goal to make puzzle solving a narratable part of the story.
I mean that if the game is story and you replace a puzzle solving activity
by the sentence "I solved it." you would eliminate, in the story, not a
sentence, not a paragraph, but an entire chapter or even the entire story.
I guess you can say that I combine these 2 processes by creating a storyline
where the conflict of the story is a puzzle that the player solves in the game.

Think about detective stories. If the 'puzzle' isn't there, what is there in
the story?

I know there are other reasons and design goals for puzzles in adventure games,
I am just talking about what I currently, typically, try to do. Of course it doesn't
hurt if the game has a door that happens to have a puzzle as a lock.

----

I also want to talk about classification of puzzles. Like how you could classify puzzles by:

o Their roles in the game (i.e. optional vs required)
o The type of cognitive task required to solve the puzzle
o The type of solution (i.e. sequence of action vs vision of a state)
o The type of required operations
o How the player first realizes that there is a puzzle
o How the player knows what the goal is
o How the player discovers all parts of the puzzle
o ...

I think that the more way you know to classify a puzzle, the easier it is
for you to create a puzzle in an area that you haven't tried, this gives
variety to the game. Therefore when I see that you have references to
ways to classify puzzles, I know that this discussion is worthwhile.

I hope this is what you want to discuss.
Advertisement
Puzzle classifications are great. I'd use that as step number 2, where step number 1 is making a list of your favorite puzzles in other games. (I vaguely remember posting a puzzle classification I made in this forum a few years ago, wonder if I can google that up and if it will be interesting or useless when I find it...)

I use an iterative story development method, sort of like the snowflake method; first a vague concept, then some brainstorming, then beat it into a more solid and detailed shape, then more brainstorming, etc. Puzzles would go in the middle, after the general story concept was established but before the exact dialogue and other details were created.

I want to help design a "sandpark" MMO. Optional interactive story with quests and deeply characterized NPCs, plus sandbox elements like player-craftable housing and lots of other crafting. If you are starting a design of this type, please PM me. I also love pet-breeding games.

Hmm, found this:
Quote:
Original post by sunandshadow:
Good puzzle levels are the ones that DON'T require being super speedy, super dexterous, and aren't risky. Puzzle-solving is all about being able to relax and take your time enough to ponder a mysterious set-up and what you should do with it. What you need, to start with, is a list of puzzle-mechanisms in other games, then you could eliminate any you don't think you could implement, try to think of clever or humorous twists on them, and then finally assemble them into puzzles and sort the puzzles out by complexity and difficulty.

The most basic puzzle mechanism is a destructible barrier: I am here, goal is there, object is blacking the only path. This can be different kinds of objects, and the objects can be destroyed by different actions of the player character, depending on what tools and inherent abilities the character has.

Second most basic puzzle mechanism is hiding the goal, or a necessary subgoal like a key or a switch, inside/behind a destructible object. There could also be multiple keys or switches behind multiple destructible objects, or it might be necessary to collect all the items of a set, for example, oh, a statue of every sign of the zodiac.

Next step up, simple pattern recognition. Perhaps you need to light all the torches which make a square, but not the one which isn't part of the shape. Perhaps you need to press switches A B and C in the correct order.

Now add in another basic element, the ability to move objects around. Now you can slide blocks to make stairs, push a trampoline to a useful position, push balls that will roll down hill gaining speed until they smash a barrier, or a button that won't stay pushed unless you put a weight on it, or a button which you have to turn on and off at different points to make other puzzle elements work.

More complex, either multiple objects need to be slid in the right order or you need to use one object to get another to the correct position like ice blocks that slide as far as they can instead of a single square, or you need to do container math to get an equal amount of water in each of three containers, or all 3 objects to the opposite side of the river without any eating the other, or a towers of hanoi puzzle. Or you have 3 buttons and 5 panels and each button flips two panels and you need to get all 5 panels to match. Or a tangram or tri-ominos type tile rotation puzzle.

Then you can add complexity to get into systems of gears and rotating rooms, hydraulics and floodable rooms, , windmills and waterwheels, furnaces and boilers, ropes and pulleys, weights and balances, magnets or gravity control, modifiable tracks for rolling or sliding objects, prisms or mirrors that must be aligned properly to transmit a beam of light or defract it to a desired color.

Commonly, these are made less straightforward by having a broken or missing piece and a replacement must be found and/or a tool must be used. This expands easily into recipes: you have a cannon and a wall, but you need to find a cannonball and some gunpowder and a source of fire to use the cannon to knock down the barrier. Or an NPC won't help you unless you bring them bread, but the baker has left town so you have to find his recipe book, gather all the ingredients, bake the bread, then finally give it to the NPC.

Now if you give the playable character some kind of distance weapon (boomerang, slingshot, arrows, fire arrows, bomb arrows, ice spell, lightning spell, wind spell, etc.) you can add things that need to be done here while the player is there, or unorthodox ways to act where you can't go or transport yourself where you couldn't otherwise go (grappling hook, suction cups, sticky/magnetic boots, helpful animals, growable plants.)


also found this:
Quote:

4. Puzzle quest - A door is locked or an NPC will not cooperate until a puzzle is solved. For MMO use these should have a randomized start state or victory condition to thwart walkthrough-makers.
- A. Dialogue puzzle - make the correct choices in conversation with the NPC
- B. Press buttons in the right order or set switches to the right positions
- C. Slide and/or rotate blocks/tiles to the right positions
- D. Other physics/logic puzzle
- E. Math problem
- F. Trivia question
- G. Maze (personally I much prefer the kind where you slide a token through a maze you are viewing from the top to the kind where you are walking through it. And mazes which violate the laws of physics with strange warps are just annoying.)
- H. Disguise - you must wear all the pieces of a disguise to enter a restricted area or fool an NPC.


[Edited by - sunandshadow on November 11, 2010 5:58:09 PM]

I want to help design a "sandpark" MMO. Optional interactive story with quests and deeply characterized NPCs, plus sandbox elements like player-craftable housing and lots of other crafting. If you are starting a design of this type, please PM me. I also love pet-breeding games.

Archetypical vs Elemental classifications

It seems that the OP is gone. But here I get the patterns of Sunandshadow's short list. The purpose is to look at the features of those types to get the Elemental classifications from the Archetypical classification.

Archetypical Classification is a type of classification rule based on what you have seen. For example, if you see a bin of apples and oranges, you could decide that there are two kinds of fruits in the basket: The ones that look like apples, and the ones that look like oranges.

Elemental Classification is based on what the properties of the objects. To do this, you catelog both the properties that are in common and those that differ. For the apple and oranges example, a property might be color. You see that apples in the bin are all redish, and the oranges are all orange. This property may be so distinct that it is all you need to tell an apple from an orange if it is picked from the bin. But there are also other properties that are shared by both species, such as both have the seeds near the center of the fruit.

An interesting property about Elemental Classification, is that you could easily ask "what if" questions on each feature and imagine a new object. For color, you could think "What if it is green?" and you many imagine a pear or a lime. For the location of the seeds, you many imagine "What if the seeds are on the outside?" and you may imagine strawberries or pineapples.

The values that an object takes with respect to an Elemental Classification is like its DNA. Classifying objects in a mutable way makes it easier to get new objects. You could start with an Archetype or go totally random.



Quote:
- A. Dialogue puzzle - make the correct choices in conversation with the NPC
- B. Press buttons in the right order or set switches to the right positions
- C. Slide and/or rotate blocks/tiles to the right positions
- D. Other physics/logic puzzle
- E. Math problem
- F. Trivia question
- G. Maze (personally I much prefer the kind where you slide a token through a maze you are viewing from the top to the kind where you are walking through it. And mazes which violate the laws of physics with strange warps are just annoying.)
- H. Disguise - you must wear all the pieces of a disguise to enter a restricted area or fool an NPC.


A. Dialogue puzzle - This is a classification based on how the puzzle is conducted
B. Button in right order - This is a type of puzzle where the player must observe the effects of the buttons during the game to solve the puzzle.
C. Slide/Rotate blocks - This type of puzzle is characterized by letting the player imagine the completed pattern. The player moves a block with the intention of moving a block to a specific part of the pattern.
D. Physics/Logic puzzle - This is a classification based on the content of the puzzle
E. Math problem - A classification based on the content
F. Trivia question - Classification based on content: declarative knowledge
G. Maze - A puzzle where all parts may be visible and the player solves it by path finding.
H. Disguise - A puzzle about perspective

Elemental Classes:

These are just literally based on the above. The options are preceded by an 'o'. When you read the list you will find that there are more options for each class. You may also see more classes.

Medium - How the information about the puzzle is communicated to the player
o Dialog
o Text
o Map
o Patterns on objects
o Equation

Learning - Whether the player is expected to do something to learn more about the puzzle to solve it.
o Required
o Optional
o None

Learned Knowledge - The type of knowledge the player must learn to solve the puzzle
o Agent Intention
o Effect of an operation
o Boundaries
o Connectivities
o Perspective

Goal Identification - How the player knows the goal
o Given
o Judgement
o Guess
o Prompt
o Scrambled Pattern
o Incomplete Pattern
o Need
o Assumption
o Intention defined by the player

Interactible Object - The objects that accepts user input
o Checkbox
o Boxed Text
o Buttons
o Switches
o Slidable Block
o Rotatable Block
o Token
o Avatar
o Clothing
o NPC

Interaction Layer - The layer where the interactible objects reside
o GUI
o Scene

Solution Execution - What te player must do to execute the solution, given that the player already knows the solution.
o Choice
o Sequence of action

Solution Type - The difference between the solution state and the state when the problem is presented to the player
o Reply
o Sequence
o Image

Cognitive Task - What the player needs to do in their mind to solve the puzzle
o Judgement
o Path finding
o Simulating effects
o Visualization
o Perspective thinking

Knowledge - Tye type of knowledge the player must have, that is not given by the puzzle, to solve the puzzle
o Facts / Declarative knowledge
o Behaviors
o Desires
o Logic
o Math
o Physics



If you start with an Archetype of a puzzle and you know its Elemental values, it is easy to see how changing a value could give you a puzzle idea. But what would happen if you go random?

A random composition:

Medium = Map
Learning = Optional
Learned Knowledge = Agent Intention
Goal Identification = Need
Interactible Object = Boxed Text
Interaction Layer = GUI
Solution Execution = Sequence of action
Solution Type = Sequence
Cognitive Task = Perspective thinking
Knowledge = Behaviors

Interpretation:

In this puzzle, you see a floor map of your apartment where the locations of your roommates A,B,C are marked. Your goal is to get the piece of chocolate that your roommate A had forgotten. You know that roommate B had heard your conversation with A, and you suspect that B might take the chocolate for herself if she finds it before you and A find it. To find the chocolate, you click through boxed text describing A's intentions and what she might have done after bringing the chocolate home.

[Edited by - Wai on November 16, 2010 2:33:36 PM]
That's a nice way to list out the different dimensions of the cognitive space of puzzles. [smile] When I made the short list I was specifically thinking of the "interactable object" dimension, but the others are important too. "Learning" and "cognitive task" are particularly relevant to audience - different game audiences enjoy or dislike different mental activities, and you should create puzzles which will be enjoyed by your target audience. "Interactable object" and "medium" may have technical limitations due to the game engine used, and being aware that activities which look different on the surface use the same underlying mechanics is helpful to planning your design in such a way that code can be efficiently reused in multiple puzzles.

I want to help design a "sandpark" MMO. Optional interactive story with quests and deeply characterized NPCs, plus sandbox elements like player-craftable housing and lots of other crafting. If you are starting a design of this type, please PM me. I also love pet-breeding games.

Advertisement
Wai and sunandshadow, thank you for brilliant answers! Very interesting and thought-provoking.

I asked the same questions on gamedev.stackexchange and recieved many useful answers. May be you'll find those answers interesting, too.

This topic is closed to new replies.

Advertisement