Alternative title:

10 years in tech

Weather Factory

  • Cultist Simulator





  • award(s)
  • mentoring

Failbetter

  • Fallen London
  • Sunless Sea
  • client projects
  • awards

Solo

  • Paradox
  • Bioware
  • Telltale
  • Burden of Command

FALLEN LONDON:

  • content-heavy free-to-play browser game
  • designed for web not mobile
  • no genre
  • grindy, divisive
  • kinda shonky-looking, 0 USD budget
  • cult hit, making c. 60K USD a month seven years after launch

SUNLESS SEA:

  • the world’s slowest RPG
  • RPG with permadeath
  • no genre
  • divisive
  • 4.5m USD year 1 against a budget of 350K USD, very positive reviews

CULTIST SIMULATOR:

  • deliberately obscure occult solitaire
  • with no tutorial
  • or genre
  • divisive
  • looks like >2m USD revenue year 1 against a budget of 190K USD
  • 2 BAFTA nominations

"Experimental"?

It is an advantage to do things no-one else is doing

Often no-one is doing those things because they look stupid

Often, they are stupid

It's hard to tell until you do them

"Experimental"?

...but I signed off on every one and bear responsibility. All were failures.

  • Black Crown Project (dumb commercial model)
  • Cabinet Noir (have you even heard of it?)
  • Machine Cares! (incoherent creative vision)
  • BELOW (not the Capy one) (not distinctive enough)
  • Silver Tree (badly implemented)
  • Fallen London mobile (swollen budget, botched launch)
  • Storynexus (actually just a dumb idea in lots of ways)
  • Distinctive and divisive
  • Reasonably good
  • Small budgets and constrained timelines
  • Apophenian design

How do we get to not-stupid?

TO SUM UP

The M577's chassis is made of bonded titanium and incorporates a 5 cm foam-packed floor cavity to protect against forged-fragmentation mines. Ground clearance is normally only 22 cm, but the vehicle employs a hydro-pneumatic, fully active suspension to allow a clean ride over rough terrain. The suspension is capable of boosting clearance by a full 30 cm and allows the M577 to comfortably tackle vertical obstacles up to 0.5 m. The hull is made from welded light alloys and is latched and bonded (rather than welded) to the chassis in order to prevent fatigue and failure from piezoelectric effects associated with an alloy-titanium interface.

The inside of the hull is lined with boron carbide ceramic tiles, each of which has been coated with a polymer resin to prevent cracks or shattering during normal travel; this resin is 2 mm thick on the outward-facing surface of the tile and is said to provide limited ablative protection against pulsed lasers. The tiles are backed with a thick layer of woven fire resistant polymer armor to limit spalling in the event of a

Newt is ten years old and has a second grade citizenship award.

She can fit through ventilation ducts.

Lean into your strengths.

  • Distinctive and divisive
  • Reasonably good
  • Small budgets and constrained timelines
  • Apophenian design

...

IT'S REALLY HARD TO BE SPECIFIC HERE

BUT OK AN EXAMPLE:

(to my annoyance)

Distinctive doesn't always mean expensive

Do you need a genre?

Do you want a genre?

...and when I say ‘divisive’...

...and when I say ‘divisive’...

Communities are essential for indies

Communities are defined by their boundaries

...you must only use this power for good

  • Distinctive and divisive
  • Reasonably good
  • Small budgets and constrained timelines
  • Apophenian design

MAKE YOUR GAME REASONABLY GOOD!

Thank you for coming to my talk

any questions?

MAKE YOUR GAME REASONABLY GOOD!

"It's art, it doesn't need to be good"

has limited effectiveness (not no effectiveness)

MAKE YOUR GAME REASONABLY GOOD!

Design your game in a way that lets you be a feedback sponge

  • simple core loop playable from day 1 - super important for iteration
  • constant community QA and iteration
  • OPEN PRODUCTION
  • educated community gives better feedback
  • deliberately roguelike for feedback

  • simple core loop playable from day 1 - super important for iteration
  • constant community QA and iteration
  • OPEN PRODUCTION
  • educated community gives better feedback
  • deliberately roguelike for feedback

YOU CAN DESIGN YOUR GAME TO MAKE YOUR GAME EASIER TO DESIGN

  • Distinctive and divisive
  • Reasonably good
  • Small budgets and constrained timelines
  • Apophenian design

"Don't spend too much on experiments"

Let me hammer that point home...

Jake Birkett does a talk on the risks of spending too long making small games

Cellar Door gave a (really good) GDC talk after Rogue Legacy on their philosophy of making games as cheaply as possible... then their most ambitious game to date, Full Metal Furies, was a 'pretty massive failure' and they think the extra years development hurt rather than helped.

Derek Yu: 'finishing games is a different skill than making them.. I tend to sprawl, like many of us.

Personally: we all need validation. Not many people will just have sold a studio, but nearly all of us will agonise about whether people will like our work

Jake Birkett does a talk on the risks of spending too long making small games

Cellar Door gave a (really good) GDC talk after Rogue Legacy on their philosophy of making games as cheaply as possible... then their most ambitious game to date, Full Metal Furies, was a 'pretty massive failure' and they think the extra years development hurt rather than helped.

Derek Yu: 'finishing games is a different skill than making them.. I tend to sprawl, like many of us.

Personally: we all need validation. Not many people will just have sold a studio, but nearly all of us will agonise about whether people will like our work

  • Distinctive and divisive
  • Reasonably good
  • Small budgets and constrained timelines
  • Apophenian design

The M577's chassis is made of bonded titanium and incorporates a 5 cm foam-packed floor cavity to protect against forged-fragmentation mines. Ground clearance is normally only 22 cm, but the vehicle employs a hydro-pneumatic, fully active suspension to allow a clean ride over rough terrain. The suspension is capable of boosting clearance by a full 30 cm and allows the M577 to comfortably tackle vertical obstacles up to 0.5 m. The hull is made from welded light alloys and is latched and bonded (rather than welded) to the chassis in order to prevent fatigue and failure from piezoelectric effects associated with an alloy-titanium interface.

The inside of the hull is lined with boron carbide ceramic tiles, each of which has been coated with a polymer resin to prevent cracks or shattering during normal travel; this resin is 2 mm thick on the outward-facing surface of the tile and is said to provide limited ablative protection against pulsed lasers. The tiles are backed with a thick layer of woven fire resistant polymer armor to limit spalling in the event of a

Newt is ten years old and has a second grade citizenship award.

She can fit through ventilation ducts.

Lean into your strengths.

  • Don't simulate where you can imply
  • Make mechanics visible to players
  • Embrace the limitations of production. You can't do everything
  • Don't simulate where you can imply - draw lessons from other disciplines besides game design

  • Don't simulate where you can imply - e.g. visually

Kuleshov Effect

  • Don't simulate where you can imply - e.g. textually

  • Don't simulate where you can imply - e.g. mechanically

CORRUPTION

TRANQUILITY

Heretic Operative, C Prompt Games

'Curious'

'Open-minded'

  • Don't simulate where you can imply

STEER THE PLAYER

THEME FIRST: YEARNING, APOCALYPSE, EXPERIMENTATION

  • lore serves theme
  • mechanics serves theme
  • narrative serves theme
  • art serves theme
  • music serves theme
  • ...NB if a player doesn't like the theme, you're f*cked, but that's same as it ever was

2. Make design visible to players

Make design visible to players!

2. Make design visible to players

3. Embrace the limitations of production. You can't do everything

This is really important for a constrained, experimental game. You *have* to leave things out, so it's easier to think earlier about what to take out

The St Exupery quote everyone knows

Garfield / Rosewater on complexity budgets

  • cutting helps focus on what is necessary
  • super useful for an experimental game
  • the designer 'grieving period' for losing a feature or content - be conscious of it, negotiate it
  • Distinctive and divisive
  • Reasonably good
  • Small budgets and constrained timelines
  • Apophenian design

How do we get to not-stupid?

“Ever tried; ever failed. Try again. Fail again. Fail better.”

  • Beckett

“Any fool can learn from their mistakes. The trick is to learn from someone else’s.”

  • Bismarck (approximately)

THE FAULT LINE

Nothing is 100% original. 'No genre' inevitably means borrowing from multiple genres. Each of those comes with assumptions, expectations, solutions. Some are more compatible than others

  • Sunless Sea: RPG vs roguelike
  • Cultist Simulator fault line: game of exploration vs RNG clicker progress?

The GDC board actually asked:

"It seems a bunch of the criticism of the game lay in the UX of the game. Is this something you felt was as designed, or was this a fall out from the process and decisions you made?"

intentional

obscurity

mechanical limitations

lack of

polish

In experimental games, it's arguably even more important to get the UI & UX right

  • give players hand-rails to hang on to while they're dealing with your experiment
  • prevent your intentions getting lost in the fog
  • helps keep the designer honest

Paul K-T quote: I had said that I didn't think about the UI because the underlying mechanics were the important bit, and the UI came later. I was overstating for hyperbole, but still, he helped me realise that this was sillier than I'd intended. The medium is the message.

Lean into your strengths.

slide design: @tronbevan

art: @ungapants


UI: @mnerurkar

v3.0 GDC 2019 - Cultist Simulator: Designing an Experimental Game for Commercial Success - Google Slides