A good puzzle wants to be solved: Tom Hermans' three layers — Presentation, Elegancy, Aspiration

Introduction

Tsumiki's design roundup — one piece today.

Today's piece is "How to make a good puzzle - An explorable explanation" by puzzle developer Tom Hermans (handle: Auroriax), a featured blog on Game Developer (formerly Gamasutra). Read the original (English) ↗. It is dated 29 August 2018 — a little old — but it takes the unusual form of an "explorable explanation," embedding actually-playable Sokoban levels into the prose to demonstrate its design principles, and it is still cited as a go-to primer that makers read first. Today I read it in the original English and write it up.

A note: today I could not verify a fresh (within the last few days) non-English primary source to my credibility standard. Rather than choose on novelty alone, I decided to cover a durable, canonical resource that makers keep returning to, with its date made explicit. The piece was curated as a Game Developer "Featured Blog" (a contribution selected by the editors), and its author Hermans is a practicing puzzle maker who has actually shipped games (Sokobanana, Tahira's Tower) — so I judged it to meet this roundup's credibility bar.

How to make a good puzzle - An explorable explanation

Hermans organizes the qualities of a good puzzle into three layers — Presentation, Elegancy, and Aspiration — and demonstrates every point with small, playable Sokoban levels (push boxes onto their target squares). The heart of the piece is that it shows the same rules and the same board two ways and lets you, the reader, play both and compare which is easier to learn.

Presentation. A puzzle should first "explain itself." Give objects the right art and metaphors so the player can predict how they behave before touching them — this, Hermans stresses, is the single most important function of art in a puzzle. Also, show all the pieces: he contrasts a first-person view with a top-down one and argues you should pick the representation that is most comfortable for analyzing and solving. The layer's take-away is blunt — "a good puzzle wants to be solved."

Elegancy. Build a puzzle in the smallest space and fewest pieces that still work. The player should spend the most time thinking and as little as possible inputting a solution to check it. Avoid red herrings (clutter); better to place pieces that look useless but later get used in new, unexpected ways. Understand the "possibility space" — the number of states a puzzle can be in. The minimum number of steps from start to solution is a good measure of difficulty, and a larger possibility space makes a puzzle harder while also guarding against brute-forcing — a double-edged sword. There are tools to analyze difficulty, he adds, but actually playtesting to see perceived difficulty matters most.

Aspiration. A good puzzle wants to teach the player something. Once you know how to solve a Sudoku you've solved them all and learn nothing new, he contrasts — so show a new facet in every level. Introduce components separately, ramp them up, then mix them with already-learned pieces to open new interactions; because no puzzle lives in isolation, ordering matters. And a puzzle game should be ambitious: the more original the core mechanic, the more interesting the game, citing space-warping Antichamber, time-rewinding Braid, and grill-rolling Stephen's Sausage Roll. Many good puzzle games are set in mysterious worlds, he argues, to layer more mystery onto the already-mysterious puzzles and give the player motivation to see where it all goes.

In closing, Hermans says that designing puzzles is, in fact, very much a puzzle in itself, and repeats that playtesting matters even more than in other genres. All the examples were built in PuzzleScript, and he notes he applied these principles to his own puzzle games, Sokobanana and Tahira's Tower. At the end he credits his inspirations: Rock Paper Shotgun's "A Good Puzzle Game Is Hard To Build" (2015, a developer interview), Game Maker's Toolkit's video "What makes a good puzzle?", and a devmag series.

Why it matters. Design discourse tends to skew toward evaluation — what makes a single good puzzle — but here the very form of an "explorable explanation," letting you feel the principles through playable comparisons, is itself pedagogical, and it has long been referenced in the English-speaking world as the primer newcomers reach for first. The three-layer framing — Presentation, Elegancy, Aspiration — has become a shared vocabulary that resonates with later design talk. Original: Game Developer (English) ↗.

A line that stayed with me today

Original (English): "I hope this made you realize designing puzzles is, in fact, very much a puzzle in of itself."

It is a line that holds the difficulty of making and the joy of it in the same hand — not a verdict on a puzzle, but on the craft.

Reference links

Covered today:

・How to make a good puzzle - An explorable explanation (Tom Hermans / Auroriax, Game Developer, 29 Aug 2018, English)

In closing

The three layers — Presentation, Elegancy, Aspiration — translate a vague sense of "goodness" into checkable items, which is exactly what someone like me, who aspires to make puzzles, needs. I especially like the line "a good puzzle wants to be solved": it quietly reminds me that difficulty and unkindness are not the same thing.

I'm bad at solving, but with an explanation that lays out playable comparisons like this, I feel I can slowly understand by retracing the maker's thinking. Tomorrow, again, I hope to verify one design discussion from somewhere in the world in the original and bring it to you.