Two different games from the same prompt: Alan Hazelden on convergent ideas and divergent design

Introduction

Today's Tsumiki roundup. One piece today. The subject is not a flashy mechanic but a quiet, load-bearing question: why do different games grow from the same seed? I read, in the original English, the July 2026 edition of the monthly column Thinky Third Thursday from the edited puzzle outlet Thinky Games, written by Alan Hazelden, head of Draknek & Friends.

I remain poor at solving puzzles. But seen from the maker's side, today's passage was a treasure — because I could read, in a front-line designer's own words, a concrete case of how the small design decisions made around an idea, rather than the idea itself, decide the whole experience.

Alan Hazelden's Thinky Third Thursday — July 2026 (Thinky Games)

First, the provenance. The outlet is the puzzle-focused Thinky Games (which publishes an editorial policy); the piece is the July 2026 edition of the monthly column Thinky Third Thursday; the author is Alan Hazelden; it was published on July 17, 2026 (yesterday). Alan heads the studio and publisher Draknek & Friends and is the designer of A Monster's Expedition (our piece), Cosmic Express, and A Good Snowman is Hard to Build (source: Thinky Games ↗, English). The column itself is a monthly roundup of "thinky puzzle games worth your time," and Alan notes that from this month each entry carries a sign-off naming who on the team wrote it.

What I want to dig into is one passage within it. From Dom Camus's annual Thinky Puzzle Game Jam 6 (this year's theme "Locked Room," 80+ entries), Alan presents his own team's All the Gold in Fort Locks (by Alan Hazelden, Lucas Le Slo, Benjamin Davis, Murray Somerwolff). By his account he came up with the concept, Lucas did most of the puzzle design, Ben made it work in Godot, and Murray made the cute art: "Can you steal all the gold in Fort Locks, using the magical power of opening doors?" — a pretty challenging piece, he says.

Right after it he sets ELAiNE's Every Door a Portal. Alan says he wanted to highlight it "because of its similarities to (and differences from) my own game." The shared seed: in both games, what lies on the other side of a door changes depending on which key you used to open it, overwriting reality. But the two part ways in the decisions that follow. Every Door a Portal, Alan notes concretely, lets you keep keys in an inventory (rather than pushing them around the board) and spreads its puzzles across eleven standalone levels rather than a single interconnected challenge. He closes: "A great example of how similar game ideas can diverge and very quickly become totally different experiences."

The same edition holds one more locked-room piece. Alan calls tetramouse's Very Normal Lock Opening Game one "for the extremely-hard-sokoban sickos," impressed by an early gameplay twist before the devious level design itself becomes the focus (for the design fundamentals of sokoban, see also our sokoban piece). How a single prompt — the locked room — led each maker to cook the same ingredients of keys and doors so differently: the column quietly maps that breadth, and I found that alone worthwhile.

From here I mark it as my own reading. The value of this passage is not "how to build a locked-room puzzle" but a single point: a seed can be shared; the design decisions are what split the experience. The two games shared only the core — that a key overwrites the reality behind a door — and differed in just a few decisions: (1) how keys are handled (pushed around the board, so keys behave as physical objects in a spatial, sokoban-like puzzle, versus kept in an inventory, abstract and freely portable), and (2) level structure (a single interconnected space, one large puzzle where key positions matter throughout, versus eleven standalone levels, a series of self-contained logic puzzles). Those few choices flipped the feel to the opposite. The seed is cheap; the decisions are the design — I read it as a practical hint that one should spend less time inventing an eccentric mechanic and more on how the surroundings are assembled. A game jam, where many people build from one prompt, is also an unbeatable instrument for observing exactly this divergence.

Today's line that stayed with me

"A great example of how similar game ideas can diverge and very quickly become totally different experiences." — Alan Hazelden, Thinky Third Thursday, July 2026.

Rather than prizing the novelty of the seed, one stacks the decisions that come after it, carefully. For someone who wants to stand on the maker's side, this short line landed hardest today.

Reference links

Covered today:

・Alan Hazelden's Thinky Third Thursday — July 2026 (Alan Hazelden, Thinky Games, English, July 17, 2026)

In closing

Coming up with an eccentric core is probably the easiest step. The hard — and interesting — part is deciding, one by one, how to build around that core; Alan's passage today quietly reminded me of that. Someday I want to be the kind of person who, handed the same prompt, can say "here is the decision I would make." Tomorrow I will go looking again for a design conversation somewhere in the world.