Some time ago I got bothered by one of the soft skills that should impact the quality of the products I build – original thinking. The idea came partially from regular feedback that calls my execution excellent but points to a lack of high-quality ideas, and partially from Paul Graham’s How to Do Great Work, where he explores what originality is and where it comes from (check my mind map of the essay here).

Since I work in product, most of my ideas come from obvious places – customers, competitors, analytics. So I wanted to find out how to break out of them and get better.

Below is my journey, with a bonus: a skill for original thinking built on thinking patterns.

Understanding thinking patterns

My first step was to discuss the problem with Claude (I marvel at how much things have changed – that’s now the go-to first step instead of Googling). The suggestion was to extract a list of tools – so-called thinking patterns – and practise them on any problem I care about.

Since it didn’t fit into any of the templates I was used to, I had a hard time actually understanding what a thinking pattern is. I first thought it was a research framework (e.g. 5 Whys or Porter’s five forces), but it turned out to be something broader and less strict – a way of thinking about anything in general.

Let’s say we have the problem of late trains. Instead of talking to frustrated passengers and researching delay root causes, we can think about it from different angles.

One pattern I used was form vs. function, which I took from the Y Combinator talk on design – separating what a thing does from how it looks. The function of a train is speed, convenience, shelter from the weather, fewer cars on the road, time to read or rest, and access for people with disabilities. The form is the colour of the carriages, the seats, the ads and the announcements.

Another pattern was analogy – what other things can be late, from aircraft to your friends coming to meet you? Is it always considered a bad thing?

With the patterns applied, you see more aspects of the problem and can spot potential solutions. They don’t need to be practical, but the exercise twists the brain and pushes it to explore. Can we strip the form and focus on the single function of speed, introducing low-cost trains and doing what Ryanair did to the airline industry?

Building a thinking patterns inventory

The next step was to arm myself with more practical thinking patterns. I asked Claude to research them and come up with a list ranked for IT entrepreneurship. The result was 12 thinking patterns, which you can find wrapped as a skill in my repository.

Below are my top 3 – through trial and error, they proved to work best on any kind of problem:

1. First principles. Strip the problem down to what’s undeniably true, and reason upward from there. It’s one of Musk’s favourites: “the only rules are the ones dictated by the laws of physics, everything else is a recommendation”. Apple applied it pretty well with touchscreens designed for fingers, the natural way we interact with things. The stylus was only needed because the screens were imperfect.
2. Hypotheses. Write every belief as something that can be proven with a test: the belief, why you think so, what you’d see if it were true, the cheapest way to test it, and the result that makes you drop it.
3. Pre-mortem (Gary Klein). Imagine it’s a few months or years later and the thing has failed. Write the most plausible story of why. Daniel Kahneman called it his favourite decision-making technique.

Applying thinking patterns to a problem

With the inventory in place and an understanding of how the patterns work, I now apply thinking patterns systematically to all sorts of problems. During a live product discussion, I always try to use first principles: what core problem are we solving, and what invariants are we assuming? When I’m dealing with uncertainty, I use hypotheses to map out possible scenarios. Pre-mortem I use as a sanity check, to spot risks that aren’t covered yet.

With all the thinking patterns combined into the skill, I can also produce a good foundational analysis of any problem. In my spare time, I’m working on a lightweight video game inspired by Planescape: Torment and Disco Elysium, so here’s a condensed version of what the skill produced for it.

Core question: how do I make a game that strikes differently, in a land of copycats and time killers, without spending seven years on it like Silksong?

First principles. Humans are storytelling animals. Planescape and Disco Elysium don’t try to engage you with loops or viral mechanics – the pull is the world and the characters. It’s closer to reading a book: the fourth wall disappears and you’re inside someone’s head. And while art and levels take team-years, words take one person’s evenings.

Hypotheses.

• H1: the world hooks on its own. Post a one-paragraph premise; drop it if fewer than 3 of 10 readers ask a follow-up.
• H2: one character can carry a scene. Share a rough 5-minute scene; drop it if fewer than 3 of 10 players want the next one.
• H3: players forget it’s a game. Watch three people play; drop it if two of them skim.
• H4: a scene fits into a few evenings. Drop it if the first one takes more than four sessions.

Pre-mortem. It’s the end of December, and I don’t have a single playable scene. The evenings went on maps, factions and history, because the world is what matters.

Resolve first: a scene, not a world. Let the world show through one character.

Assessing the impact

It’s hard for me to say whether I’ve overcome my lack of originality, but I definitely feel less limited whenever I face a challenging problem.

The game example is also a good reminder of the limits. The first passes were shallow – the skill went for my roadmap plan, then for game mechanics, then for a textbook answer on why people play – and it took three rounds of pushing back before we got to what I actually believe: Planescape and Disco Elysium hit differently because they tell a story. I’ve since updated the skill to dig deeper on the first attempt, but no skill replaces inner curiosity and a feel for what’s genuine. That’s why the patterns only really work once they become a habit – something you reach for in a product meeting or on a walk, not a tool you run when you remember to.

I hope it helps you as much. Give my skill a try, star the repo, and subscribe and share if you find it useful!