Induction to Inductive Games
This Induction to Inductive Games is an attempt to gather all games falling in the inductive category, sometimes also named Instructionless, Rule-discovery or Black box games.
Having grown from research notes for the unreleased game "Puzzle Type" (2021), this living document may now be useful to players looking for inductive game gems, perhaps curious about the origins of the genre, as well as to inductive game designers, for whom the final Design Considerations was written.
Hallmarks of inductive games include:
- one or more underlying rules to be discovered;
- progress equates with a deeper understanding of the rules;
- no instructions given, intentionally,
Those can be condensed in the following definition:
Inductive games are primarily about discovering a set of underlying rules, using inductive reasoning: prior examples or past interactions with the game world are used to construct hypotheses (possibly subconsciously), to be confronted with new examples and interactions, until they can no longer be disproven.
by Pedro
This definition excludes:
- Guessing or Concealed Information games, where a limited number of questions can be asked before trying to guess not an underlying rule, but a piece of information such as a code (Mastermind), pattern (Breaklock), word (Hangman, Wordle), concept (20 questions), movement sequence (the Inductive Game of Rubik's Cube), personality (Akinator), etc…
- Puzzlehunts, chiefly because they are an heterogeneous category, as individual puzzles may possess (or not) an inductive nature.
- games whose primary theme is not inductive even when occasional inductive elements are present (e.g Tunic, Fez)
With this in mind, a tentative classification of Inductive Games could start by dividing them in three major groups: Atomic, Closed and Open.
Atomic
Atomic Inductive Games comprise a single unexplained situation, usually presented as a single level-. Occasionally this situation may be spread across multiple levels (perhaps to clarify the idea or present more challenges) yet without adding new rules.
| Title | Author | Year | Interaction | Question |
|---|---|---|---|---|
| Angry man's contest | 2006 | Observation of picture clues | What is the man angry about? | |
| Tempres (with a demake) | 2017 | Click/tap (anywhere) | How to light up the next bar? | |
| Color Theory | 2020 | Placing lines and dots | How many dots should be placed, and how connected? | |
| Proof of Equivalence | 2022 | Place and displace vertices | How to move in 3D? | |
| Secure | 2022 | Pushing blocks to "secure" positions | What kind of positions are secure? |
Closed
Closed inductive games tend to be self-contained, with the number of rules limited at the outset. Often, each of these rules is explored more deeply than those in Open Inductive Games, with particular emphasis on edge-cases and on their mutual interactions.
A common eureka moment occurs when rules previously tested in limited scenarios suddenly generalise to encompass a new situation.
As soon as all rules in a Closed game are fully understood (and subsequent game goals attained) the game is complete. As the rules never change, replaying the game, or playing it after watching someone else offers diminishing value.
At least two subtypes of Closed inductive games may be distinguished: Symbolic and Environmental.
Symbolic
In symbolic inductive games, particular rules are tied to particular symbols (or combinations thereof).
| Title | Author | Year | Interaction | Question | Commentary |
|---|---|---|---|---|---|
| Signal | 2013 | Moving and pushing blocks with different labels | How are movement patterns of different labels linked? | Intentional use of visual noise to disguise mechanisms "hidden in plain sight" | |
| The Witless | 2017 | Drawing lines on the board | What does each symbol mean, in relation to the lines and the board? | The progression though panel sets, which have consistent rules, kindly provides a way to assess one's partial understanding towards the whole solution (rare in paper-based games) | |
| Illiteracy | 2018 | Clicking symbols to swap other symbols | What is the effect of each symbol? | - | |
| MISLED, Puzzled | 2019 | Placing symbols on a grid | - | - | |
| Gestalt OS | 2019 | Placing symbols and watching their effects | - | - | |
| Cypher | 2019 | Activating tiles | What do the annotations imply? | A manual rule check lets players ask whether the current solution is correct, avoiding "chance wins" | |
| Char Min Glee | 2020 | Clicking Buttons | - | Impressive codegolfing | |
| Understand | 2020 | Line drawing through a grid with symbols | - | - | |
| Artefact | 2020 | Inputing symbols into text fields | - | - | |
| Cypherset 1 | 2020 | Moving around a grid while activating tiles | - | - | |
| From Muddled to Clean (FMTC) | 2020 | Drawing loops | What rules govern each loop, so the puzzles are uniquely solvable? | Pen and paper based | |
| Fremda | 2021 | Drag symbols to certain positions and a chain of tones | Which symbol configurations produce the required chain? | The musical and visual feedback indicate partial solutions, thereby guiding the player. | |
| Insight | 2021 | Region division | Which constraints on the possible divisions are imposed by the symbols? | - | |
| Autoscript | 2021 | Placing symbols | - | - | |
| Stoff OS | 2021 | Placing symbols and watching their effects | - | - | |
| Null Hypothesis | 2021 | Combining gems | Which combination of gems will produce an effect? | Non grid-based | |
| Funny Looksy | 2021 | Drawing lines to divide symbol tiles | How does each symbol behave in relation to the line and other symbols? | Witness like (core mechanic) | |
| Symmetry Hall Problem | 2022 | Activating tiles | What do the numbers and underlays imply? | Cumbersome input, also annoying that the default "empty" colour is not recognised as a legitimate colour, requiring more clicks. | |
| Discovery Jigsaw | 2022 | Placing polyominos on a board | Which placements are valid? What are the rules for each symbol? | Levels can be played in any order, adding a new layer to discovery by letting players take full control of their inductive process. | |
| Catalyst | 2022 | Combining gems | See null hypothesis | Inspired by Null hypothesis, but using a search algorithm to comb through possibly interesting rule sets | |
| LOK Puzzlebook | 2022 | Crossing letter sequences on a grid | What rules govern each word, so the puzzles are uniquely solvable? | Pen and paper based | |
| Gridspech | 2023 | Flipping tiles on grid | - | ||
| The Runic Conjecture | 2023 | Converting letters to a path | What effect does each letter have on the path, and on other letters? | - | |
| ABDEC | 2023 | Placing "letters" on a board | What shape does each "letter" represent and what are its rules? | Pen and paper based | |
| FMTC "Ripoff" | 2023 | Drawing lines | - | - | |
| Agriculture | 2023 | Placing and removing symbols (animals) | - | - | |
| Bilattice 1 | 2024 | Shading tiles on a grid / dragging clues | Which rules are imposed by each symbol? | Pen and paper-inspired, with feedback e.g. in the form of exclamation marks/checks on number clues, etc… | |
| Bobogram | 2024 | Shading tiles on a grid | - | Pen and paper based | |
| Alphabet soup for picky eaters | 2024 | Typing letters, to wake up colourful characters | Which letter combination does each character require? | Typing-based |
Environmental
Environmental inductive games do not restrict rules to particular symbols only, but tie them to physical objects in a simulated environment, akin to traditional escape room games (too many to list here).
These games add a focus on lateral thinking and attentive observation, because it may not be obvious what is a "symbol", where to look, and what can be interacted with — though some objects may be actual drawings of symbols, e.g. in a piece of paper.
| Title | Author | Year | Interaction |
|---|---|---|---|
| Starseed Pilgrim | 2013 | Placing seeds | |
| Quern - undying thoughts | 2015 | Activating mechanisms | |
| Rose | 2016 | - | |
| The Witness | 2016 | Line drawing through a grid with symbols; activating mechanisms; observing the environment | |
| Sensorium | 2016 | - | |
| Islands | 2018 | Activating hidden mechanisms in a block-pushing world | |
| Compass Rose | 2019 | Answering riddles and interacting with the world | |
| Un-lock | 2020 | Opening "locks" | |
| Lingo | 2021 | - | |
| Taiji | 2022 | Activating tiles; activating mechanisms; observing the environment | |
| Logicality (witness clonish) | 2022 | Pressing buttons in order | |
| Syllogism | 2024 | Turning arrows in a 2D map (interactign with the environment) | |
| Awakening | TBA | Extracting clues from videos |
Open
Open inductive games have no beginning nor end, because rules are derived on the fly for each new game. When relying on symbols or pieces, those need never be tied to particular rules, but may change meaning with every replay.
Nevertheless, some well-known open games may develop a "canon" of known good rules, usually aimed at beginners, plus, sometimes, a "uncanon" of unacceptable rules, infamous for frustrating players. As long as new interesting rules keep being devised by their players-creators, Open inductive games are intrinsically replayable.
| Title | Author | Year | Interaction |
|---|---|---|---|
| Eleusis | 1956 | Determining why a card may or not be played | |
| Mao (also Mau) | ??? | (see Eleusis) | |
| Bongard Problems | 1967 | Finding the distinctive feature in two sets of 6 images | |
| Queries 'n Theories | 1970 | Reconstructing a hidden language with a formal structure (basic sentences and replacement rules), by formulating questions (queries) | |
| Petals around the rose | 1970s | Observing dice rolls and results, then trying to deduce a formula | |
| Penultima Chess | 1994 | Attempting to move a chess piece | |
| Zendo | 1999 | Asking the master whether a piece configuration follows the hidden rule | |
| Eleusis Express | 2006 | (see Eleusis) | |
| Luminare | 2015 | Classifying new examples as black or white | |
| Riddle Runes, formerly Sigils | 2016 | ??? | |
| White Pillars | 2018 | Pushing blocks | |
| Devil's Calculator | 2019 | Typing (numbers) | |
| Caterpillar logic | 2020 | Classifying new examples as valid or invalid | |
| Send A Damned Message | 2021 | Typing | |
| Exactly 7 | 2021 | Increment or decrement a score counter by activating tiles, picking letters, etc… | |
| Instructionless Grid | 2021 | Paper puzzle drawing | |
| The Name of the game | 2021 | Variable rules, multiple question checks | |
| h | 2022 | ??? | |
| -vxv- | 2022 | Drawing in a mysterious interface | |
| House | 2022 | Visual clues | |
| VIX 1 puzzles: I,II,III , | 2022 | Placing Roman Numerals | |
| Puzzle Type (unreleased) | 2021 | Typing | |
| Forgetting | 2024 | Shading (tiles on a grid with symbols) | |
| INTAO (direct link) | 2024 | Guessing a word based on visual clues (colour blocks) |
Social games where players keep providing words or objects as examples until they find a hidden rule, such as Umbrella, Green Glass Door, Black Magic, Music Box, Scissors and Triangle Mind could be seen together as components of a larger open game. However, as the supply of such hidden rules seems relatively limited, it seems more appropriate to call them a collection of games.
Design considerations
Designing inductive games requires additional considerations from those that apply to standard puzzle games.
Here are a few that
Blind vs Kind
In a blind game, you are playing blindly, attempting things without any feedback, except a final "yes or no" — only after understanding absolutely everything can you win the current situation.
In contrast, a kind game, you receive some feedback after every action, which reduces frustration because you can gauge your own understanding.
When making Inductive games, it is of utmost importance to make them kind.
Auto vs Controlled
As players interact with an inductive game with partial understanding, every so often they produce an action with unexpected results. They may even wish to explore randomly to see what happens.
Thus there is an unavoidable chance that players "win" without fully understanding the system. Some games enforce a manual "win check" to prevent this, but this actively slows down players, creating friction. Much better to provide them them more control, by letting them go back and forth in their actions, "unwin" or pause anytime.
Complicated vs Complex
The complexity of an underlying rule relates to how many test cases that need to be checked by players so they are convinced they understand the rule, as well as to the cognitive effort of each individual check.
Rules in inductive games should be complex enough to feel interesting but not overly complicated, such that they feel longwinded (too many unrelated checks) or burdensome (each check is hard or slow to reason about).
Forgetful vs Recorded
As the number of past interactions grow, a point will be reached where players will find it hard to keep all past information in their mind. Whereas in paper-based inductives players can be expected to take notes, players may expect digital inductive games to take notes for them.
Digital inductives should at the very least allow players to replay any particular situation, thus discouraging forgetfulness. At their very best, they'll provide a handy, concise record of past interactions that serve as handy test cases (without revealing the underlying rule) for players evaluating new hypotheses.
Fictional vs Real
Since a deeprer understanding is the result of finishing an inductive game, it is natural to ask: what was understood? By playing a fictional game, you gain a deeper understanding of the game's fantasy world; but in an real game, you are rewarded with a new understanding of (some aspect of) the real world.
References
The following game lists partially overlap with those in this article:
Acknowledgments and Changes
2021-04-14 A rough first draft incorporating feedback from the Thinky Community: thanks
2022-12-05 Thanks to
2023-02-05 Thanks to
2024-02-21 Thanks to
2024-12-02 Added a recent entry to the table
2025-07-17 Added INTAO by