The House Rules Framework

House Rules, for short.

I'm a systems thinker, and over my time in games dev, I've noticed that good companies, full of good people, have pipelines and processes that can still go sideways. House Rules is a framework for diagnosing the 'why' behind it all.

You might be metabolising people's wellbeing in order to keep the dashboards green, or be permanently stuck in firefighting mode, or be putting too much faith in process. Or, you mightn't even notice that sensible-sounding policies and practices are derailing some of your best people.

Every piece of work that every organisation delivers is down to a system, and every single one of them was designed. Even the ones that feel like they just… happened.

I'm talking about the machinery underneath the gym perks and the duvet days – the actual nuts and bolts of how a business goes about their business. How decisions actually get made, what happens to bad news, and whether one specific person getting hit by a bus will kibosh your output for a year.

Nobody knowingly sets out to build a bad system. They all made sense to someone, on the day they designed it. But they stack, and interact, and times and expectations and experience levels all change. The rational decision that was right three years ago might just become the thing leading to people quitting right now.

I've spent twenty-five years inside these systems – employed and on contract – and all could have benefitted from House Rules.

Including the ones surest they'd cracked it. The 'Good Places' weren't wrong, exactly. The results were there, the people were happy, and they were more thoughtful about their staff than most. But they'd attached themselves to a localised maximum, blinkered to the possibility that there's usually a better way.

And on the flipside, some of the most crushing places I've worked have got some stuff very very right. The inspiration for House Rules' 'Deliberate' Capability came from somewhere that would likely have crashed and burned without it.

House Rules is a framework that allows us to read the systems issues inside a company. Picture a swan: gliding serenely up top, paddling like hell underneath. House Rules is the metaphorical swim coach.

It works in three layers, and they connect simply:

Traits: what an organisation is like. Its raw character, informed by systems design, culture, and trust: whether ownership is clear, whether people feel safe saying negative things, whether the company has a choice of responses or just wields a proverbial hammer. There are fifteen fundamental Traits at the moment.

Capabilities: how it does what it does. Every organisation executes across four Capabilities in order to do its actual work: Discuss, Deliberate, Decide, and Do. These are broadly separate from Traits, but interact with them to give the final layer.

Conditions: what goes wrong as a result. Put the Traits and Capabilities together, and certain troubles become predictable: Hero Dependency, Hidden Burnout, Trust Debt. Some might be obvious, some might be hiding. All are fixable.

That's the whole engine: who you are, plus how you do what you do, produces the trouble you get.

So, in practice: you come to me with one or more symptoms, or a nagging feeling that you could be doing better. I spend time in your company – pun intended – and analyse your Traits and read your Capabilities until we can see what's happening. Not only will we know what Conditions you've got, or have lurking in the background; we'll know where they're coming from and why they're happening.

You get it as a Stat Block – a high-level RPG-inspired summary of your outfit's Traits, Capabilities, and Conditions.

You bring the symptom. House Rules finds the causes. Then I help you fix them.

Here's an example of a company you miiight just have heard of. InGen's Stat Block, the day before everything went pear-shaped.

InGen

Stat Block

InGen – International Genetic Technologies – is John Hammond's bioengineering company, the one that cracked cloning and built a what-could-possibly-go-wrong dino theme park.

House Rules finds an organisation that excels at doing, but that's dangerously poor at questioning. It executes both flawlessly and tirelessly. But, behind the scenes, it rests on a handful of irreplaceable people, and has almost no slack for the day things go sideways. The strongest predicted conditions are Hidden Burnout, Hero Dependency and Resilience Deficit – a place that looks to be performing beautifully while quietly running its reserves down. Unusually for a company this size, there aren't any political machinations at play; rather, its failures are baked into assumptions that no one thinks to challenge.

Traits

Ownership Clarity

Accountability sits – and sits obviously – with the right people. Hammond owns the vision, and everyone below him knows exactly where their responsibilities start and stop: Wu has the genetics, Arnold the systems, Muldoon the animals, and Nedry the code. No ambiguity about where the buck stops, and no committees, no fuzzy edges, no 'let's circle back on this.'

Psychological Safety

People say the unsayable – often straight to Hammond's face. Muldoon openly comments – to the boss – that the raptors should be destroyed, and Wu defends the science to a bunch of openly sceptical experts. Nobody softens their doubts to spare Hammond's feelings, and nobody gets punished for voicing them. The one wrinkle is Nedry – the single person whose real gripe never quite makes it out into the open. And that's probably not going to turn out well.

Information Flow

Knowledge moves well across InGen. The control room sees the helicopter view of the whole park, specialists brief each other openly and honestly, and expertise crosses disciplines freely. When something is known, it generally travels. No-one's sitting on what they know in order to look clever.

Follow-Through

Commitments consistently become completed work. You don't get actual, real dinosaurs without planet-sized Follow-Through. Whatever else is wrong, InGen finishes what it starts – which is precisely what makes its blind spots so dangerous. A place that can't execute never gets the chance to build something this fragile.

Trust Default

Hammond assumes people are trying to do good work, and builds accordingly – oversight is support rather than surveillance, and nobody's counting hours. The exception is Nedry, who wasn't distrusted so much as underpaid. Different failure entirely.

Feedback Fidelity

Bad news does reach leadership, but it arrives softened. The warnings are all raised; they just get sanded and smoothed before they land. 'We spared no expense' is a catch-all response that ignores the truth underlying what's had money thrown at it.

Resource Visibility

Operational fragility is... worrying. So much focus is on the output, and the expenses not spared, that other types of resource – and the fact that those things are running on fumes – aren't even considered. Every person that hands in their notice will come as a surprise. Hammond can't conceive that people could want anything else.

Incentive Alignment

Opening the park is the be-all and end-all. Nothing else gets a look in. Every decision bends toward getting those famous gates open on schedule. This isn't an organisation that actually wants to run a safe park; it's an organisation that needs to get one open, right on schedule.

Learning Rate

InGen is brilliant at 'how' and utterly uninterested in 'whether'. Hand it a problem and it'll engineer a gorgeous answer – to something it should probably never have said yes to in the first place. It solves like a champ, but never stops to question the question.

Adaptive Capacity

Reality moves faster than the plan does. Hammond's picture of his park updates grudgingly, and late, and by the time he's in a place to do something about it, things have... 'moved on.'

Competence Distribution

The whole park rests on just four irreplaceable people: Wu, Arnold, Muldoon and Nedry. Four brilliant specialists, four single points of failure. Lose any one of them, and an entire capability goes with them. My sense was that at least one of them had the look of a resignation waiting to happen.

Principle Coherence

'We spared no expense' sits awkwardly beside an underpaid, resentful engineer and a skeleton weekend crew. The values are trumpeted; that the actual underlying truth is saying something different is never considered.

Slack

There's no give. Everything's tuned to run exactly as planned, with nothing kept aside for a rainy day – literal or metaphorical. It'll all work beautifully, right up until the first thing that doesn't go to plan.

Capabilities

Discuss

There's no shortage of talk, and it can happen anywhere: the dinner table, the control room, the tour. Things get mentioned, discussed, and then... it all stays there. There are few mechanisms in place to capture and action meaningful feedback, because doing so might waylay getting the gates open on schedule.

Deliberate

InGen can only be in the position it's in with rock-solid deliberate capabilities. Genomic mapping and reconstruction – not to mention actual dinosaurs – doesn't happen by accident, and it's obvious that the park's scientists have benefitted from expense not having been spared. But – perhaps due to flaws in Discuss – not everything that needs deliberation has had it.

Decide

Decisions get made, and made cleanly. Hammond's vision is clear, and he'll decide in that direction every time. And therein lies the problem – once his mind is set, it's set like concrete, and no amount of new evidence gets the issue back on the table.

Do

This is what InGen is frankly frightening at. Whatever it commits to, it delivers – at a scale absolutely nobody else has managed. Which is the whole problem in one word: execution this good papers over every crack underneath it, right up to the day the cracks win.

Conditions

high confidence

Hero Dependency: active

Everything critical lives inside a handful of heads. Wu on genetics, Arnold on systems, Muldoon on the animals, Nedry on the code. Lose one and you don't slow a capability down, you delete it entirely – there's no subs bench, no backup, no 'second person who also knows.' You don't need to dig for this one; it's visible from the car park. The place runs beautifully because these few are exceptional, which is precisely the trap.

Politics: not indicated

Almost none, and that's genuinely to InGen's credit. Clear ownership, good information flow, proper technical authority – there's barely any room for empire-building or backroom games. When people fall out, it's about the work, not the pecking order. This isn't a low score, it's a Condition that isn't showing – because InGen's strengths are actively crowding it out. Which is the unsettling bit: there's no villain here to root out. Just a very capable company strolling confidently off a cliff.

medium confidence

Hidden Burnout: active

A skeleton crew running a park that ought to need an army, and doing it brilliantly – so brilliantly that nobody's clocking the cost. All that Do and Follow-Through is being paid for out of Slack, Resource Visibility and a tiny pool of overworked specialists. I can see the crew, and I can see the pace. What I can't see – because nobody here is measuring it – is how close to empty the tank already is. That gap is the whole problem. It's assumed to be full right up to the moment it's bone dry.

Institutional Rigidity: active

Brilliant at solving problems inside its worldview, hopeless at questioning the worldview itself. InGen will happily engineer a world-class answer to 'how do we contain actual dinosaurs?' and never once entertain 'should any of this exist?' You can watch it in how every problem gets framed – always an engineering puzzle, never a question about the premise. It bends its version of reality to its will, never the other way round. This is the one that quietly underwrites most of the others.

Commitment Inflation: emerging

Once Hammond's committed to something, that iron Follow-Through teams up with a weak Learning Rate to make it almost impossible to walk back – even when the evidence starts screaming. You can see it every time someone pushes: the answer's already decided, and the pushback just bounces off. The very thing that gets InGen over the line becomes the liability – a decision, once made, gathers weight, and reopening it starts to feel like betrayal rather than sense.

Resilience Deficit: predicted

Tuned to perfection for the plan, the plan, and only the plan. No slack, no adaptability, one kind of answer to everything. I didn't watch this one bite – on the day I visited, everything was holding perfectly. But the profile is unambiguous: the moment reality wanders off-script, things won't wobble, they'll cascade. This is the prediction, not the post-mortem. It just happens to be a very confident prediction.

Trust Debt: active, but only just

There's a little rot in the floorboards – 'we spared no expense' doesn't quite match what's on the ground – but at least we still have floorboards. People still argue in the open, still collaborate, and there's none of the cynicism or covering-your-backside you'd expect if trust had properly gone. Plenty of frustration; nowhere near exhausted trust. A rating of two, not five, precisely because it's real but nowhere near its ceiling yet – the debt's mounting, it just hasn't been called in.

Learned Helplessness: seeded but suppressed

The ingredients are all here: feedback gets softened, leadership reinterprets concerns rather than absorbing them. What's stopping it setting in is one specific strength – people still feel safe speaking up, and Muldoon, Arnold and Wu are all still shouting, repeatedly. That's the only thing holding the shrug at bay. What I saw was the moment just before it, not the moment itself. But it's a balance, not a resolution: the day Psychological Safety slips and people decide there's no point saying anything, this stops being a risk and becomes the thing quietly running the place. No star, because right now it's contained – but if that suppressor goes, it's a bad one.

low confidence

Delivery Theatre: active

The gap between the show and the substance. InGen is superb at the presentation layer – the tour, the merch, the oft-repeated 'we spared no expense.' The demos I saw felt built to reassure rather than reveal. But that's the honest limit of a flying visit: I watched the show, I couldn't get behind it. My read is that what's holding it all up is thinner and more rickety than the polish suggests – and a closer look is exactly what would settle it.

Summary

None of InGen's problems need a genius to fix – they need someone to actually look. And the good news is there are fewer of them than the Stat Block makes it feel: those three headline Conditions all trace back to the same short list of causes. Too much resting on too few people, and the whole place running flat-out with nothing in reserve.

So the structural fixes are dull and completely doable. One: break the hero dependency – get what's in Wu's and Arnold's and Nedry's heads written down, cross-trained and backed up, so no single resignation walks off with a whole capability. Two: build some slack – stop running every system at full tilt, and keep something back for the day the plan doesn't survive contact with reality (or a rampaging Tyrannosaur). Do those two, and all three headline Conditions lose their teeth.

But that only catches the fall. It doesn't explain why InGen built something this precarious in the first place – and that's the deepest fix, sitting under all the others: the Learning Rate. Give the place a real way to interrogate its own assumptions, because the companies best at answering 'can we?' are exactly the ones that lost sight of asking 'should we?'

Principles in Practice

I've done a lot of different jobs, at a lot of different scales, in a lot of different places.

But looking back, no matter where I've been placed nor what I've been doing, I make things work for the people on the other end. Human-centred, always. Because you get that right, and suddenly you're not fire-fighting, you're watering.

Nobody Is Average: Project CARS 2

Every team is different. Good systems don't expect people to conform to an idealised process; they adapt to the people doing the work.

pCARS2 had a large – and fully remote – multi-discipline team. And it had significant expectations. The creative direction was strong – but before production scaled, the underlying structure needed to match the reality of the people running it, not the idealised version of them.

Different disciplines needed different things. Some leads worked best with high-level direction and room to breathe. Others needed granularity – task-level clarity, tight feedback loops, someone absorbing the noise above them. I built production systems around how people actually worked, not how a generic model said they should.

The result wasn't just a smoother production. It was a team that could sustain pressure without fracturing. One clear signal: after the turbulence of the first title, Project CARS 2 passed certification first time. Not through heroics at the end, but because the underlying structure held.

And all with a production team one-tenth the size you'd normally see for a project like this.

House Rules lens: Response Variety. A system that only copes with one kind of person only copes with one kind of day. I built production around the actual people, so the team could bend instead of snap.

The Plan Isn't Sacred: Fast & Furious Crossroads

A system's job isn't to reassure leadership. It's to tell the truth, especially when the truth is uncomfortable.

Deep into Crossroads dev, and I could see it coming. The planning approach being used looked clean on paper – capacity balanced, tasks assigned, boxes ticked. What it didn't account for was dependencies. Or critical paths. Or the gap between what the plan said and what was actually happening.

We in production couldn't do anything about it at the time. So instead, we built the right system in parallel – quietly, without interrupting anyone's workflow – and waited.

When the moment came, the switch took no time at all. No interruption, and infinitely better information. Which was, at that point, largely: yes, we're in trouble. But knowing you're in trouble, clearly and early, is the system working. Pretending you're not is the system failing the people inside it.

The game was an absolute disaster, but that's more down to the various egos involved, and a tale best told over a pint or two. But I made a promise that we'd deliver, and so I made damn sure we did, and 150+ people kept their jobs.

I shipped this and Project CARS 3, in the same calendar month, in the middle of a pandemic. Handed my notice in the very next day.

House Rules lens: Feedback Fidelity. A system that can't tell leadership the truth isn't a system, it's a comfort blanket. We built one that could – quietly, in parallel – and had it ready for the moment the truth actually mattered.

Convenience Isn't Good Enough: Timesheets

If two hundred people are each spending five unnecessary minutes a day on administration, the system has been designed for the convenience of the wrong person.

A client needed documented evidence of hours worked across projects. Not too much to ask, really. The standard solution would have been: ask everyone to fill in a timesheet.

Instead I built a system in an afternoon that defaulted each person to their correct project, logged a full standard day in a single click, and only asked for more input when something had actually changed. The heavy lifting – setup, processing, maintenance – sat entirely on my end.

The people exposed to it needed a second or two. Most days, less.

It's a small example. But it's the same logic at every scale: complexity belongs with the person whose job it is to absorb it. Not distributed across an entire team as a daily tax on their time and attention.

House Rules lens: Slack. Five wasted minutes a day times two hundred people is capacity the system nicked, because it couldn't be bothered to absorb the complexity itself. So I made it absorb it.

Humans Aren't Resources: Threkka

People don't produce their best work by ignoring how they function. They produce it when the work is designed around sustainable human rhythms. Not everyone's in the middle of the bell curve.

It started with six words scrawled on a sheet of paper: "Joe Wicks makes Clash of Clans." Five years later: a fitness app without precedent, live on the App Store, designed from first principles, and built entirely solo.

No team means no buffer. Every decision – game design, production cadence, technical architecture, behavioural science, App Store delivery – landed on the same desk. The challenge was to juggle all this, without getting lost, without losing momentum.

I designed my own days. What worked for me was – 0700 to 1030, hard at work. Then exercise, lunch, and a snooze; then an afternoon shift from 1400 until the work was done. Very much marathon; very much not sprint. Output-based, not tick-box task-based. No holidays, but – and this is super-important – no burnout either. The system was designed around the human running it, not the other way round.

What that bought me was years of sustained, high-level output. And at times I was outpacing entire teams I'd worked with in the past. Not because I'm exceptional, but because I wasn't fighting the structure I was operating inside.

The core lesson I can take from this? Treat people like grown-ups. Let them work closer to their ideal. Design the system around them. Get it right, and the output takes care of itself.

House Rules lens: Hidden Burnout, dodged. The Condition sitting at the top of a lot of companies I've seen, run as an experiment of one. Design the days around the human and you get years of high output with no wreckage at the end. Turns out the framework works pointed inwards, too.

Three Ways In

I don't really do off-the-shelf. Every company's house rules are different, so every job is too. But people tend to start in one of three places.

Fresh Eyes

- the read -

A defined chunk of time and an honest outside opinion.

Bring me a specific problem, an awkward question, or just a nagging sense that something's off.

You'll get a House Rules read on it: clear, specific, and exactly where I'd start pulling. The Stat Block, pointed at you.

Get Close To The Work

- the diagnosis -

I come inside and follow the work as it actually happens – not as the process diagram reckons it does.

Together we work out which house rules are helping, which are quietly costing you, and which are one bad week away from taking the roof off.

Stick Around

- the redesign -

Finding the problem is usually the easy bit.

I stick around to help redesign the systems, support the change, and make sure the good ideas survive contact with reality.

The aim is to leave you needing me less than when I arrived.


If in doubt, we start small.

© 2026 Chock Hoss