DESIGNER-STUDY · 2026-06-17

Inside Alan Hazelden’s Philosophy — Make many first, then name the 'thinky'

Reading Draknek, maker of A Monster’s Expedition and Sokobond, through his own words

Reviewed by Kizuki · #designer-study #alan-hazelden #draknek #thinky #puzzle-design #philosophy

Introduction — The man who named the 'thinky' puzzle

Alan Hazelden is a London-based solo puzzle designer and the founder of the studio and publisher Draknek & Friends. Sokobond, A Good Snowman Is Hard To Build, Cosmic Express, A Monster’s Expedition — all of them look quiet on the surface, yet each aims at that moment when something rearranges itself inside your head the instant you solve it.

I write about him now because he is tightly tied to designers this site has already covered — Arvi Teikari (Hempuli) and Jonathan Blow among them. To describe his games as puzzle games distinct from things like Tetris, he reached for a single word: “thinky.” This essay is not about his games but about the person, read only through what he himself has said in interviews and talks.

Background — From student hobbyist to puzzle community's standard-bearer

On his own site Hazelden notes that he has been making games since 2006, starting as a student hobbyist and going full-time with the July 2013 release of Sokobond (his site ↗). Draknek & Friends, founded the same year, has since grown from a personal banner into a label that publishes other people’s puzzle games too (Draknek & Friends ↗).

His title is fluid. In the TouchArcade interview he calls himself “Head Draknek,” and says that on internally-developed games he is creative director and main puzzle designer, while also contributing to production, testing, bizdev and UX design (TouchArcade, 2024 ↗). He is not merely someone who makes games; he tends the genre itself.

He launched a Discord for thinky puzzle games in July 2018 and runs the monthly Thinky Third Thursday newsletter and the yearly Cerebral Puzzle Showcase (same source). Less famous in Japan than his games are, in the English-speaking puzzle scene he is known as the community’s standard-bearer.

Philosophy — Make many first, and don't fear the bad ones

The most consistent thread across Hazelden’s statements is the belief that quantity brings quality. In his older talk “Learning through failure” he states flatly: “I believe that the most effective way to create a great game is to first create a large number of games which are not great. Many of these games will not be worth polishing.” (Draknek “Learning through failure”, 2010–2011 ↗). Bad games are not a source of shame but the necessary sediment on the way to a good one.

That “make it first” philosophy survives, reshaped, a decade later. Asked what skill grew most in his puzzle design, he points to “improving my ability to have a mental model of player understanding — internalising what I do/don’t expect players to pay attention to or what’s going to be necessary to teach a certain gameplay concept” (TouchArcade, 2024 ↗). The more games he makes, the faster that internalising goes — the philosophy of volume feeding the craftsman’s eye.

When he praises Hempuli’s five-block-pushers-a-month output in his own newsletter and says he is “delighted every time” (Thinky Third Thursday, 2024-03 ↗), it reads as the same value in mirror image. Being able to genuinely celebrate another’s prolificacy comes easily to someone whose own creed is “volume first.”

Obsessions — Narrowing language to 'thinky', and designing the teaching

Hazelden’s obsessions show up first at the level of language. He says he “may be responsible for coining the word ‘thinky’ as a genre” because he was “trying to describe my games as puzzle games distinct from things like Tetris” (TouchArcade, 2024 ↗). He wanted a clear line between reflex-driven Tetris and the stop-and-think games he makes. The very act of naming becomes a statement of design intent.

His other obsession is teaching. His games use almost no tutorial text, letting level layout alone convey the mechanics. The “mental model of player understanding” remark reads as the craft behind that wordless instruction: what to show, what to withhold, in what order to let the realisation land. What he has honed again and again is less the solution itself than the design of the learning curve.

Failures and how he overcame them — Shipping, then wishing it had been there from the start

Hazelden talks frankly about his own misjudgements. In A Monster’s Expedition he originally meant to place the discovery of raft travel late in the game — in his words, “the epitome of the ‘woah, I could have been doing that the whole time’ moment we were interested in exploring” (TouchArcade, 2024 ↗). In practice, moving the mechanic earlier worked better as a marker for the end of the tutorial and gave the world more flexibility. “It’s now hard to imagine it any other way,” he reflects — a case of letting go of an early conviction when playtesting said otherwise.

A more painful admission concerns hints. The hint system in that game was added after release, and he says “the process of adding hints to A Monster’s Expedition (after release) made me wish they’d been there from the start” (same source). He adds that his future games “are all likely to have some form of hint system at launch.” Rather than letting a failure end as one-time regret, he institutionalises it into a default for everything that follows — that, I’d say, is the shape of how he overcomes things.

Dilemma — Between difficulty and approachability

What Hazelden names explicitly as a struggle is the tug-of-war between difficulty and approachability. Listing what he wants to carry forward from A Monster’s Expedition, he cites “the value of approachability,” admitting that while his earlier games “weren’t completely unapproachable, they were a bit easier to bounce off” (TouchArcade, 2024 ↗). The urge to build genuinely chewy puzzles and the care not to lose players midway do not reconcile easily.

His change of heart on hints sits on the same line. In a genre that sells itself on difficulty, adding hints after the fact surely raised the worry of “making it too easy.” His answer, as I read it, was not to lower the difficulty itself but to build an escape route for the stuck player in from the start — protecting the purity of the challenge while widening the entrance. How he splits that difference shows where he stands today.

Influences — SpaceChem, English Country Tune, and his peers

The influences he names are concrete. Recalling when he first became conscious of the genre, he lists from the early 2010s “Manufactoria, SpaceChem, and English Country Tune” (TouchArcade, 2024 ↗). English Country Tune in particular is by increpare (Stephen Lavelle), and Lavelle’s tool PuzzleScript is the very thing Hazelden used as the level editor for A Good Snowman (Draknek Talks, GDC 2016 ↗). The influence runs through tools, not only ideas.

His other “influence” is the contemporary circle he actively created. He celebrates Hempuli’s output in his newsletter (Thinky Third Thursday, 2024-03 ↗) and boosts younger developers through his Discord and showcases. Designing not just to receive influence but to build the very space in which influence circulates makes him an unusual kind of author.

Kizuki's Reading

From here this is my reading, Kizuki’s. The Hazelden of 2010 preached “make a large number of games that are not great” — a young volume-ist unafraid of failure. The Hazelden of 2024 speaks instead of “approachability,” and of the careful regret, after shipping, of wishing hints had been there from the start — a man determined not to drop players. I read this not as contradiction but as a shift of weight. Someone who once bet on “how many games can I make” now bets on “how many games can a player finish.” The philosophy of volume was externalised intact — into supporting others as a publisher, and into community as a machine for producing volume together. I read Hazelden less as a lone author than as a gardener trying to design the metabolism of the puzzle genre itself. His best move is probably not any single game, but the word “thinky” and the scene that word grew.

Closing — Where to start

If you want to understand Hazelden, start with A Monster’s Expedition: it is where his current answer on hints and approachability takes its most straightforward form. For the artistry of how mechanics are taught, try Sokobond; for difficulty and kindness living side by side, A Good Snowman Is Hard To Build.

A few threads to related designers, too. Stephen Lavelle (increpare), whom he names as an influence and whose tool he borrowed, and Arvi Teikari (Hempuli), whom he champions most eagerly — this site already has a study of Teikari. Draw the lines of “volume” and “peers” and the three of them start to look like one constellation.