PRACTICE 2016 conference notes
Laine Nooney
Laine Nooney is a professional historian of computers and video games, currently writing a social and business history of Sierra On-Line. As an Assistant Professor of Digital Media at Georgia Tech, Nooney travels internationally to speak on the history of personal computing and computer gaming. Her work has been covered by NPR, The Atlantic, and Killscreen, and she is a founding co-editor of ROMchip: A Journal of Game Histories, the first peer reviewed journal on the subject, set to launch Spring 2018. @Sierra_Offline
Beneath the Flesh of Walls: On the Making of Mystery House
Roberta and Ken Williams: built Mystery House & Sierra Online. How did they do it?
Mystery House: A game that launched a thousand checks for $24.95
First game Roberta designs. Game that launched Sierra. First adventure game with graphics.
Fall of 1979: Jimmy Carter’s 3rd year. Walkman released. ESPN. Arrival of personal computer. Voyager
Ken used a teletype. Linked it up to a mainframe. Poked around the files to look for games. Found Adventure
He didn’t like it; called Roberta in, who took over the machine
She wasn’t coming to it as a technical or game novice. Did data entry in COBOL. Ken tried to interest her in games before.
What’s important are: adventure was that it spread. and it spread because it was designed for children. No abbreviation, syntax, etc
No interface interference. Adventure had the right affordances to appeal to someone who didn’t like computers.
Roberta tore through it. She played other adventure games. Around this time they got a personal computer. “She loved them all. and then there were none left.”
Inspirations: Clue, Agatha Christie’s And Then There Were None
The innovation in Mystery House was graphics. which was Roberta’s idea.
Not a lot is known about the process by which she produced the design. No documentation or notes of this game.
She kept designs hidden from Ken for weeks. Roberta blindsided him at dinner one night.
her ideas for the game was expressed as sketches. to show relationship with spaces showed network graph of rooms
her design practice required separation of roles between designer and programmers. but two not completely separate. Ken had to tune Roberta into logic of programming
3 columns: verbs, nouns, what happens. This helped shape her ideas into something that could be expressed in code.
Mystery House graphics were coded as vector graphics
The core of this design process was dialectic.
Mystery House proved itself as 1) something you COULD do with a computer. and 2) something you could do SPECIFICALLY only on an Appple 2
1980s review: grpahic was ‘very nice’ ‘showed room and objects in detail’
Came in ziplocked bag with a single piece of paper documentation
ad copy: “Tired of buying games that become boring after a few hours of play? Online systems is dedicated ot deliverying seriosu software for hte discriminating gamesman… six months in development”
Mystery House: made $11,000 in May, $20,000 in June
Mystery House was first game ever set in domestic space. She designed in the home setting. Came from a different culture than was prevalent among game designers at the time.
Fundamental arc of King’s Quest game: family flung apart and needing to get back together
Each of her Sierra games were about getting over a particular technical issue. The flagship games were always a debacle. It was the 2nd game that they figured out what they were doing.
Game engine developed in mid 1980s allowed them to separate the design from the programming. It grew out of the way she had to work, but also allowed them to do cheap labour. Had machine that could standardise process.
They typesetted everything themselves
On being a historian and nostagia: Wary of nostalgia. It’s a frame that tends to make games history about itself where she’s interested in game history that leads people outside of games.
Robin Hunicke
Robin Hunicke is the Co-Founder of the independent San Francisco game studio Funomena, building PC, VR, AR and console game titles: Luna, Woorld and Wattam. A game designer and producer by training, she has a background in computer science, fine art and applied game studies. She has been designing, making and teaching about games for over 12 years (Journey, Boom Blox, MySims, TheSims2). An Associate Professor in the Arts Division at UC Santa Cruz, she is Director of the Art, Games & Playable Media B.A., founder of the Ludo Lab Center for Games, Play & Social Practice, and principal faculty of the Digital Arts and New Media MFA. She co-organizes the annual Experimental Gameplay Workshop, sits on the PlaySFMOMA, FDG & VRDC advisory boards, and volunteers in support of a variety of causes – including the original Indie Game Jam, Global Game Jam, GDC Game Design Workshop, IGDA Education SIG, Lexus Design Award, Girls Make Games, Anita Borg Foundation, Google Made with Code, HEVGA and the CS For All initiative
The Hustle
There’s a lot of success bias in our industry. Making games is difficult; making experimental games is really hard
- Spent five years just hanging out
- First job working at big company on establisehd franchise even though heart in indie games
- People she did random projects or initiatives with, helped her find jobs for rest of your life
- Didn’t have a reputation among gamers until very recently, but had reputation in industry because of the initiatives and other things she had done
- Did not have a female mentor until she hired an executive coach
Knew what her dream job was, but started by finding a low-level job where she could be of service
Creative entrepreneurship: What do you need in your toolbox?
When you start, you have…NOTHING But over time you do things. Your experiences crenulate into failures. But failures crenulate into stories. Stories crenulate into intuitions, and THAT’S what you get in your toolbox
When you go do something, ask for funding, etc, the only thing you bring into it are your intuitions. (and same for the person on the other side)
On Failures
Failure #1 - Too much shit to do
Your title is not reality
Intuition: You get what you want. She really wanted to be a people person. But was suffering in her role as executive producer (while she had a lot of fun working as a junior object designer)
Failure #2 - Blaming other people
Took a long time to realise she wasn’t really working for others, was working for herself.
But also: Your memories are not reality. The things you remember are the little things you take out and repeat.
Failure #3 - Blaming yourself
Intuition: You can only do what you do.
Failure #4 - Not asking for help
Intuition: Good people understand
Good people know you’re doing your best, and bad people can shut the fuck up.
Practice:
- Know your value and your values It’s better to give up if you can’t do it right than to continue doing a shitty job at it
- Decide what you will learn
- Pack your toolbox. review, and share your failure.
- Play the long game
Life is a lot like Carcassone. Don’t spent your life bellyaching about a meeple you could’ve popped off the board if only you drew the right piece. that’s absurd.
Worse thing is to try to convince someone on the other side of a burning bridge they have to pour water on it now.
Just try to be a non-force. If you can get through life
Sustainable funding for experimental video games: have a fund that’s dedicated to supporting under-represented minorities making experimental games. A lot more like microlending.
On firing people: make it really clear to people at when things are going south that they are not meeting your expectations. thing you want to avoid is either side having regret or resentment, which comes out of mismatching of expectations
Dain Saint
Dain Saint is the co-founder of Cipher Prime, an award-winning Philadelphia game studio known for uniquely captivating experiences. As a programmer, musician, writer, and activist, he loves figuring things out and giving other people the tools to do the same. He says “Philly” a lot, and wonders if you’ve got a problem with that.
Genesis of this talk was his 2015 GDC talk that explored the possibility that we all live in a simulation Q from audience at the time: if it is simulation and you cna peek behind curtain, would you want to? He answered “absolutely”, but auddience follow-up: “Even if it would break the illusion?”
Made him realise that “We are, all of us, magicians. We live in the 2nd age of magic.”
First age of magic: When we first controlled fire we attributed magical significance for it. Followed by earth, water, wind. So elements became how we understand the world. Magic = the power to influence events by mysterious forces.
Second age of magic: In the 20th century we are again creating worlds and influencing events in them with mysterious forces
Experience vs Games. Experience is an event that leaves an impact
What’s a good experience? intentional, holistic and alive
Your goal when creating an experience: create emotional responses to leave an impression.
First principle of XD: An experience is intentional
so, what’s the purpose of your game?
- Know what you are crafting.
- Why you’re crafting it?
- Make sure every piece serves the goal
- Cut away every piece that doesn’t serve the goal
Case study: Auditorium;
Goal of the game: connect player to a childlike senes of play. Implication: no lives, no score
Second principle: An experience is holistic
Holism: systems should be viewed as wholes and can’t be fully understand solely via its parts.
This means your game is also your menus, your pause screens etc. Everyone hates making menus. but you can’t skimp on them.
All of the uncontrolled variables are part of your experience and you have to take them into account.
Case study: Splice;
Goal of the game: make you feel like you’re a scientist in a lab uncovering the secrets of life Implication: menu colours designed to bathe the surroundings. Colours in game matched by colour of light given off by the controllers. Music is soothing to counteract sense of frustration. Menu as barcodes.
Third principle: An experience is alive
Juice: adding satisfying meanignful animation to every interaction. but Juice can go too far
He prefers the term ‘alive’. Juice vs alive. Juice is just a layer of polish on the surface. Alive builds out from core
Example: you have two blocks. One wood, one metal. how to convey that? First instinct is skeumorphism (textures). But that’s not only way. and what if you’re going for a flat design style? can use motion animation, particle effects, sound, etc to convey the difference
Case study: Tailwind;
Goal of the game: “before you die, live” Implication: cinematic approach. ship has 20 animation scripts
There’s a zeroth principle
An experience is experienced: if I can watch a video of a game and get your game then that’s no good.
- you watch a movie
- you read a book
- you play a game
- you experience an experience
You know it’s special when people say: You just have to experience it yourself
The lessons that we learn (through games) provide ways for us to look at our lives in earnest and make changes that have real impact
Implications for team structures if you want to create experiences: You need to pay attention to your words. Have a nomencalture conversation where you define terms. You need to build a common language. i.e. ‘objects’ is vague and confusinng
Also: we have one person that’s The Keeper of the Intention. at the end of the day that person gets to decide if it fits the intention or not
People like to talk about harmony in the workplace but it’s bullshit. if people care deeply about things they are going to disagree about things.
Ken Perlin
Ken Perlin, a professor in the Department of Computer Science at New York University, directs the Media Research Lab, and is a participating faculty member at NYU MAGNET. His research interests include future reality, graphics and animation, user interfaces and education. He received an Academy Award for Technical Achievement from the Academy of Motion Picture Arts and Sciences for his noise and turbulence procedural texturing techniques, which are widely used in feature films and television, as well as the 2008 ACM/SIGGRAPH Computer Graphics Achievement Award, the TrapCode award for achievement in computer graphics research, the NYC Mayor’s award for excellence in Science and Technology and the Sokol award for outstanding Science faculty at NYU, and a Presidential Young Investigator Award from the National Science Foundation. He serves on the Advisory Board for the Centre for Digital Media at GNWC, and external examiner for the Interactive Digital Media program at Trinity College. Previously he served on the program committee of the AAAS, was general chair of the UIST2010 conference, directed the NYU Center for Advanced Technology and Games for Learning Institute, and has been a featured artist at the Whitney Museum of American Art. He received his Ph.D. in Computer Science from NYU, and a B.A. in theoretical mathematics from Harvard. Before working at NYU he was Head of Software Development at R/GREENBERG Associates in New York, NY. Prior to that he was the System Architect for computer generated animation at MAGI
Grew up in the Bronx. played with plastic dinosaurs. As a little kid, you have this superpower where you could take two plastic dinsoauars and play with them, and come up with stories and worlds. But at 6 or 7, your adult briain kicks in and you lose that magic, and you spend the rest of your life doing game design and coming up with these prosthetics to try to recapture that past
References Rainbows End by Vernor Vinge
VR kits are no different from Sir David Brewster’s stereoscope from 1851. Somemone said “hey, we could put an iPhone on that”
We’re in an interesting transition time. Right now when you think of VR it’s this box you stick on your face. Immediately it’s isolating. But everyone now also working on spectacles - goal is to make them social, and socially invisible. It’s like eyeglasses: if you don’t think about it you don’t remember who’s wearing glasses.
“The holodeck is other people”
The tragedy of the future is that it lasts for about 5 mins. It’s amazing, but it stops being amazing.
2010: Microsoft Kinect was when he realised: the physical modality is going to come. So, how do we design a language for it? What’s the grammer of drawing in the air?
The concept isn’t new: the idea of a visual programming language has been around for over half a century. But I want this to be real time, low cognitive load.
So, what happens when you have AR? You can start thinking about things in the world as being phsyical and also as an information object.
One of the non-obvious things: If you have glasses that let you see things that aren’t there, they also let you NOT see things that are there (lets you hide the plumbing) So you can use gesture control of robots that are invisible to you in AR.
The use of performative visual languages:
When you draw on a blackboard you have this storytelling aspect in the sequence of how you draw it, which is lost when you are just showing a program. It enhances teaching when you can draw and show virtual concepts, and allow people to intearct with and play with it. The watershed moment will happen when this can be done casually and conversationally between people.
Ken: I’m really interested in bringing characters to life. But we’re not fundamentally interested in interactive characters; the most important thing about interactive characters is that they are interested in us.
The power to develop natural language is an instinct, not a technology.
Real breakthrough is when little kids who grew up with that start evolving the language.
“Technology is neither good, nor bad, nor value-neutral”
Nikita Mikros & Joshua DeBonis
Nikita Mikros is the CEO of BumbleBear and a veteran game designer. He is a three time IGF finalist, Indiecade and Come Out and Play winner. In addition he has worked on many games for clients including Nickelodeon, Cartoon Network, Adult Swim, LEGO, and others. He has taught game design and programming at the School of Visual Arts since 1994. Nik is currently very focused on creating new types of competitive games for arcades. @nmikros
Joshua DeBonis is an award-winning game designer and President of BumbleBear Games. Joshua is a co-founder and organizer of NYC-Playtest. He has taught game design and development at Parsons the new School for Design and the NYU Game Center. Josh is particularly interested in creating games that provide a deep and meaningful experience in a short play time. @joshdebonis
Killer Queen sucks because it can’t be played by monkeys
Arcade industry prevailing attitude is that game is good if it could literally be played by monkeys. (KQ is not about that)
Originally envisioned that the game would be a festival game where people use NES controllers and all look at a projection screen But was destined to have it in an arcade. But, they had no idea how to make that transition. No mentors.
Modern arcade games: oversized. gaudy. Specialised in one-time thrill rides. Make the best use of affordances one can imagine. (if making motorcycle game, what better affordance than to actually ride a motorcycle) and makes the experience difficult to reproduce at home.
But when they started, mental model was as arcade player in 70s and 80s.
In order to convert Killer Queen to arcade, needed to make a series of adjustments, all constrained by the realities of manufacturing and making them easy to repair:
- attract mode
- tutorial
- business model
Attract mode: What’s shown when no one is playing. intended to cactch your eye. lots of contrast. not just monitor but also the display lighting. Problem: When no one at the game, really hard to get a game going. Solution: Show film footage of a bunch of people playing and having fun. Shows 20-somethigns drinking beer (because everyone wants to be 25)
Tutorial: show gameplay in attract loop. Also, film footage of people playing shows orange and blue, and the two sides of the cabinet. Show that it’s a team game and you are playing against the other cabinet
(But one problem. too short: need to give enough time not just for undrestanding but also for players to explain how it works to other players)
List of specific problems the tutorial had to solve (things it needed to teach):
- players didn’t know it was a team game.
- players didn’t know you could fly.
- players didn’t know how to capture.
- players didn’t know you could kill from above and from behind.
- players didn’t know queen has special move, which is the dive.
thought also went into design of the controls. No superfluous controls. Needed to fit five players into a very compact space, so controls are staggered with so one set deeper and then the adjacent set shallower to create more room.
Business model: one fee for 10 people. in barcades rather than family arcades. community aspect (more on that below)
Things that are different about KQ compared to other arcade games/makers
They treat this as a living game. Always adding content and iterating on the game. Also they help create context for the game. Deliberately made no home version of the game so you have to organise over Facebook and go out to play it with friends (or to make new ones)
Received a lot of pressure to put in AI so you could play with fewer than 10. But, liked social/virality aspect of people having to recruit. So, compromise was to build a simple bot that would be v efficient but v robotic.
Established arcade industry see arcade owners as their customers and no regard for the player.
What does it mean to regard the player?
Treat them as smart, have game literacy, and empower them.
Empowerment examples: fan art, fan merchandise. Only rule it that it has to be transformative. delibertately not fill in backstory and give names. Had national championship and asked each city to make a video to introduce themselves
Players turned it into esport, which they didn’t anticipate. Try to empower that: created Killer Queen League. Both are part of the community as players and not just as developers.
“What we’re doing is really to build a new, more engaged community around arcades. We’re seeing revitalisation of arcades and it’s exciting to be at the forefront.”
How to avoid exlusion when building a community? Hard to answer. Several possible answers: Show at big festivals so it’s a destination (but then can’t play regularly). Extend the game: make home version that’s set in same world but it’s a different game so it doesn’t compete with arcade version. Wary of direct port: Nidhogg got criticised after releasing home version because it didn’t live up to expectations.
How to maintain a living/evolving game when there’s lots of stakeholders? Updates - send downloadable file to the operators. download onto usb stick, put into machine, and reboot. Local mods: nearly all machines are same. In settings, can put in URL for local community to meet up and when local league night is. Map changes weekly
Sam Barlow
Sam Barlow is a BAFTA Award winning game director and writer who has been making games since his cult 1999’s title, Aisle. Barlow has an extensive history of making games that create deep personal connections with their players. With Silent Hill: Shattered Memories he created a classic that psychologically profiled Wii gamers. In 2015 he released Her Story reinventing the FMV genre for the YouTube generation. Barlow is now working for NYC based Interlude using the new medium of interactive streaming video to create the next generation of stories
Game designing like a writer: How Her Story was constructed entirely (on paper) (in excel) in Google Sheets
Why he made Her Story: Had always been fascinated by screenwriting: How this document on paper would be a blueprint that describes visual storytelling, and could (more or less) survive the production process. Also, the creators of the screenplay could sit in this very protected, safe space to create things on paper
The enemy of creation, of writing, is distraction. In bigger videogames, narrative creation was a very messy process.
Wanted to make something that could be rapidly prototyped and could be done without being distracted by having to put things into a game engine.
Also, fascinated by how you can look at the 200 pages of a script and see if the movie is going to work. Half of movie-making is figuring out the 40 plot beats on paper, and there’s no point in progressing until you get it right. You can’t say that my story is 99% there, and the visuals and special effects etc will fix the last 1%
Creating Her Story
When first started, was just doing a lot of research. Also downloaded a lot of flowchart software and drew out a lot of stuff. Each time it didn’t work, would blame the software.
Was essentially searching for experience of figuring out a story that’s very fluid. A promise that was held out (but undelivered) in games like Portal or Deadline from the 80s, because in those games you would start to get a sense of it, but then, would run up against scripted narrative.
Breakthrough came from finding transcript of interviews with James Porco, who attempted to murder his parents. (One died and the other survived). Had all the interviews with him in transcripts online in a txt file.
Crystallisation of the idea came when going for a walk one day. “Subconscious taps me on the shoulder and says, here’s the game.”
Porco transcript essentially provided a prototype. Dumped all of Porco’s lines into google sheets, and used text filter to replicate the idea of searching through it, and just started playing.
“Through playing this crappy prototype, the sensation I was feeling was slightly magical - the sense of exploring with some level of blinkering. If I read the whole transcript linearly, it would’ve been very dull. The experience of having it chopped up for me, and have me work at exploring the story and figuring out the conflict made it fun.” Noticed that typing words and not getting any results back didn’t feel crappy.
“I got very interested in how the words you searched for curated and created this set of search results”. Here was a narrative videogame where the magic was happening in the arrangement of the clips.
Her Story is essentially a game about what it is like to be a writer.
So, went back and uninstalled all the flowchart software.
Since, with Porco transcripts, there was no game designer behind it. Just a situation with these characters. He decided to just to write the scene and not worry about the game design part at all.
Writing the scenes
7 interviews, wrote it out chronologically. In each interview, knows: agenda of the woman. agenda of the cop. what happened between this intv and the previous one. this was important because it allowed him to write from the character outwards.
Felt that: as much as the Porco transcript worked, he should be able to make it a lot better as a game designer.
Then came the editing process, which he described as a balance/Chiascuro process: alternating between zoomed in and overview mode
how it worked: took script, ripped out questions. put into google sheets. applied formula that gave score of how ‘connected’ it was.
He focused on those that have 5,4,3,6,7 results. These are exciting because constraint in Her Story is 5, so it’s exciting to get a 6 or 7 result, because it feels like something is just out of reach. So, he looked for evocative words like ‘wedding’ or ‘husband’ and if it’s in 6 or 7 results, tried to see if he can use a different word to reduce it to 5. It’s chiascuro-like because alternating between curating just this little piece of the painting in the moment, and finding the spots to touch up by looking at story as a whole.
Another way to identify the points that need touching up were to look at lowest-scoring clips, because those are the problem clips. eg. One that says ‘cute, what ages are they’ (in a convo where the detective handed his wallet to her). This clip is exciting becasue there’s action in the scene (amid hours of footage of just talking) but the problem is: No one is ever going to find this clip becuase it was so disconnected. So, he looked at word list, and saw what additional words can be used. Solution: clip now says “Cute what ages are they? [looks at photo] you must really love them” (so it surfaces when searching for “love”)
A marker of good screenwriting is you don’t have redundant systems of imagery. and this process of weeding out low-scoring clips forces him to do that.
Aside: had file named ‘Latest Word Analysis’ “This wwas very early in the process. V good indie game developer source control” [jokingly]
There are limits to how much the excel analysis can surface. For example, clip of “im ok im ok fuck” has a low score, but people are gonna find it because they’re gonna search for swear words.
The key to it all
Find a way to keep the story alive in the writing. Writing a story is the dullest thing you can do when there’s all this other interactivity going on. But what you don’t want to do is have your story feel mechanical, so you have to write it in the moment and forget about the bigger game structure.
Hardest part of the craft of telling interactive stories: have this thing that is by nature more mechanical and designed than other stories, and yet have the audience not see that mechanism.
So, with the spreadsheet analysis, “the numbers were just telling me there’s a problem here, but I’d fix it as a writer.” Using numbers, and number crunching, to figure out a structure then allows us as writers to write from the moment, outwards.
one of most interesting things in this medium: author/designer on same footing as player. “Some of that magic felt to me like the same magic of writing”.
Asides: Project was orginally called The Box when it first started. Sharon Stone’s casting tapes for Basic Instinct was one of the aethetic inspirations
Question: How to keep the end from happening too soon? If you turn on a movie halway through on TV. within 2-3 mins you’ll have picked it up. All performances encapsulate wider context That means both that it’s quite hard to break a story. But also, if you have a favourite movie you’ll watch it more than once. You’ll have free dramatic irony on repeated viewing
Longer answer: See the history of detective stories: whodunnit (agatha christie) -> howdunnit (columbo) -> whydunnit (scandinavian series: popularity of serial killer series came out of this)
Therefore, even if you knew who did it. Even if you knew how it was done, you still have the meat of figuring out why things happened.
Max Zolotuhkin
Max grew up in Gainesville FL, where he set a route the first day he ever climbed in the summer of 2000. After graduating from the University of Florida in 2007 with a double major in political science and psychology, Max has devoted his life to crafting thought provoking climbing movement in gyms and at competitions across the country. Max considers his greatest indoor climbing successes to be making the US Bouldering Team in 2009, setting for multiple Adult Bouldering National Championships, and most recently the IFSC Bouldering World Cup in Vail Colorado in 2015. On rock, Max has bouldered V14, clipped up 5.14, and is well known for once foolishly trying to boulder a sport route. Aside from climbing with his partner Clara, he also enjoys watching and playing basketball, drinking craft beer, street fasion, and occasionally writing
Mastering movement
Climbing is a naturally social sport. It’s categorically different from, for example, two surfers compete for the same wave or two basketball players go for the same ball. When one climber succeeds it informs and inspires the other.
Route-setters: puts holds into walls and designs routes and boulder-problems
A bit of definition:
- Route: something you tie into a rope to climb
- Boulder-problem: something you do without a rope
On a daily basis, route-setters clears all holds from gym, cleans holds, and sets new route
The fact that he was urged to look at route devleopment critically from day one had a big influence on him.
“I’ve never thought about what I do as game design before.” But, points out that there are actually a lot of similarities. Many of the smae soft skills to excel as a developer also works for route-setter:
- creativity
- ability to take feedback
- conceptulaise 3d space
- wide-ranging knowledge of what’s out there (other games, or other routes)
Both professions also require hard skills: Game dev: programming languanges Route-setter: handyman skills
Commercial setting for climbing gyms is comparable to creating traditional AAA single-player game - a designed environment for a single person.
Setting for competitions more like Call of Duty or Counterstrike - multi-player tournament. spectator sport.
“Climbing is basically one long learning process. You can win a climbing competition but you can’t win at climbing”
Commercial setting pays the bills. We get to set maybe only 2-3 competitions a year, but those are most rewarding. Takes 4 days with crew of 4-5 setters and a chief to do a tournament.
After each round is set, the setters climb on the problems to get a sense of it. take photos, then take it off. Major element of tournament is climbers don’t know in advance what they’ll come up against. Setters also watch climbers do earlier rounds and make on-the-fly adjustments to the routes for the next round 3 rounds per competition. adjust based on field of athletes.
The goals: to achieve a clean break-down of results, with a clear winner and no ties. To test them fairly. And put on a good show for the crowd.
RIC scale for judging difficulty: risk/intensity/complexity
Not that any one problem will score high on all three. but collectively, the problems on a run should test all 3 aspects
** The one thing we’re not worried about, is what the athletes think of the problems. That’s fundamentally differnet from what we do in the climbing gyms - there we absolutely care what the people using it think. **
Mechanics of designing a problem
Does it have one solution or multiple solution? With multiple solutions, climbers with different strengths will be separated by their decision making. Problems like these are a dice roll. What if everyone chooses most difficult path and fail b/c easier path is not obvious enough?
There is no other sport where the outcome of a tournament is so beholden to the organisers.
How to deal with difference in height and arm span? (really hard to set for)
Invovles both art and science. Can break down difficulty of route into several factors, but, setters also have their own style. His style: put eye-popping shapes onto routes. Says there’s usually a correlation between how good a route looks and how well it climbs
One of the challenges is being in touch with your own strengths and weaknesses as a climber. An example of this: Low-risk, low-complexity movements that were just intense, top athletes were considerably better than Max. So in a tournament he set a few of these problems expcting them to be too difficult and it turned out just right. Really eye-opening for him because usually his understanding of difficulty is really close
Current maximum is a four-move sequence
One of the most interesting aspects of setting is the need to visualise the sequence without doing them. This is hardest to teach to new setters. Trial and error is only the way to create those schemas in your head.
“It’s what I imagine designing VR must be like. you know your body and how it might interact with the environment” but you can’t be in the environment all the time when you are designing
Specific tricks of the trade
- Can sand holds down: to force maneuvers you designed for (remove other options)
- Fill hold with putty.
- Special holds for competition. looks like a hold they’re familiar with but is actually substantially worse. can also do opposite
- esoteric finish holds
More general advice
- “Always make things as aesthetically pleasing as possible, but not at the expense of functionality”
- Keep the big picture in mind
- Always have a team mentality. You can design a route but you lose control of it once you put on the wall (and other setters and climbers ‘play’ it). Need to be open to take in feedback
- Be willing to take big swings and miss
Steve Pardo
Steve Pardo is a professional game composer and audio designer currently residing in Boston, MA with his wife, Amy, and two kids. He works at the video game studio Harmonix Music Systems as a lead composer/sound designer. He also composes music for games independently and through the audio production house SkewSound. Steve serves as audio lead at Harmonix on titles including the upcoming Rock Band VR for Oculus Rift, and Beat Sports for Apple TV, both titles where his audio prototypes heavily influenced the game’s fundamental game mechanics. His early audio design work is showcased in both the Rock Band and Dance Central franchises. He has composed original music for games including titles such as Grim Dawn (Crate Entertainment), Fantasia: Music Evolved (Harmonix/Disney), Chariot (Frima), The Magic Circle (Question), and Fated (Frima). Steve studied music at the University of Miami, FL, where he received his Bachelor’s in Studio Music and Jazz (2006), and his Master’s in Studio Jazz Writing (2008)
The modern game soundtrack is a film score
At Harmonix, games are built out from the music
Referenced A city sleeps
Game music systems are inherently more exposed in music games
sound engine tools - fmod and wwise
Showed a Venn diagram: linear music (static music. loops and stingers) / Interactive music (dynamic to player’s choices and game state) / Generative music (soundtrack created on the fly)
Games scored and lessons learnded
Part 1: Linear music
Beat Sports. Why linear? because players need to focus on interacting with the music. If backing music changes, then it’ll be very distracting. This also allows the designer to have full control over the music: they chose songs that were rythmic and melodic. To make player’s action feel more impactful, they add a little bit of generative music on the swing sound effect
Grim Dawn: There was no 3rd party audio engine or audio code support. No budget for live music. Post mortem thoughts: people seem to appreciate the moodiness and the composition. Due to random playlist, some pieces didn’t fit well. If there were intensity parameters built into the game, they could’ve made adjustments. Limited soundtrack on a 60+ hour game could get repetitive.
Part 2: Interactive music
Fantasia: gave players 3 mixes to choose from with 3 unque stems. Entire gameplay was built around manipulating the music
Gigantic: MOBA with large mythical creatures. Music post mortem thoughts: took a modern approach where the music fits with current gameplay mode and intensity.
The Magic Circle: (music making was part of the game)
Part 3: Generative music
At the moment, stage of development is basicallY: How much can we get out of stem swapping?
Rockband VR: music performance crafted for VR. Optimised for being on stage and being present in the space. Binaural. Spatialised stems. Reverbs. But… traditional Rockband gameplay not working well in VR. So, crafted brand new mechanic where you are generating the guitar part. This generated music is the game. The music you created is then given a score by the game.
A short primer on how to make a generative virtual electric guitar performance simulator:
Sampler: a >3gb multi-sampler containing >4300 audio files 1 multisampler with 11 channels, for single notes 1 multisampler with 16 channels, for chords Sound effects amp simulators (authored parts of the song)
How does it sync up? every song is a MIDI file marked up with chord tempo and rhythm info. With this info, can dynamically generate solo licks, sustain/tremolo, chords, and single notes
But to make it all sound ‘good’, the core part of the song is hand-authored.
In summary
- new tools are making a lot of new stuff possible
- work with what works for your game. Don’t feel pressured to have to use generated or interactive music just for the sake of it
- music system can be a two-way street where the music-making impacts gameplay
Celia Hodent
Celia Hodent is director of User Experience at Epic Games and holds a Ph.D. in cognitive psychology. Celia joined Ubisoft Paris in 2008 to help the editorial team apply neuroscience knowledge to game design. She later worked for Ubisoft Montreal and LucasArts. She joined Epic Games in July 2013 to help guide the studio toward improved user experience practices and develop the studio’s UX strategy and process. Celia is the founder and curator of the Game UX Summit, which launched in Durham NC in May of 2016, hosted by Epic Games
Applying UX and Cognitive Psychology to Game Design
When we get ‘trick’ questions wrong, it’s not because we are not logical, but it’s just easier for your brain to use heuristics.
UX definition: how the target user perceives and interacts with the game, including how engageign the experience is, relative to the design intentions
A lot of designers say ‘I don’t need a psychologist i’ve been doign this for years’ - feel like an alien somtimes but ‘don’t worry - i come in peace’
How the brain learns: Perception – Info processiong –> Memory
Attention and emotion/motivation influence info processing
Perception: 3 levels of processing info
- Sensation - physics orientation, sptital, frequencye, brightness
- Perception - Organisation of the visual field
- Cognition - Knowledge: access to semantics
Memory, how it works:
- Sensory memory - part of perception, importance of attention | encode info | v
- Working memory - short-term memory. requires heavy attentional resources. easily disturbed
- Long-term memory - potentially unlimited. largely unknown
The brain doesn’t work like a computer even though a lot of people use that analogy.
The forgetting curve (worse-case scenario): 60% lost within 19 mins. after a day you lose 70%. but long-term 20% retained.
“Also if you’re colour blind you don’t give a shit about this slide”
Memory lapses:
- Encoding deficit (lack of attention). solution is to draw attention
- Storage deficit (paid attention but weakens in time). Solution is repetition
- Recall deficit (available in memory but temporarily inaccessible). Solution is cues.
Recall deficit example:
In theory this is best way to teach people: give them a problem. let them figure it out, give reinforcing encouragement when they succeed\
(but 10 mins later when confronted with same situation player didn’t craft anything even when prompted by giving ingredients) Feedback from player is that: I couldn’t remember how to get to the craft screen
Onboarding is not just the first hour into the game. you have to remember to repeat things
We all suck at multitasking.
Limitation: cognitive load theory
A lot of designers have this feeling that people are going to figure it out. It’s not true. You have to be very motivated to figure it out. most people are not motivated to figure it out. They’re motivated to have fun.
Among most impactful factors of churn were understanding the basics of the game.
Analytics is very good at telling you what’s going on, not why it’s happening.
Brain limitations is at the core of UX heuristics
Good UX = Usability + (Game) Flow
4 kinds of affordance in UX: Physical (button) Cognitive (button label) Sensory (label font size large enough) Functional (the button triggers, say, a sort functionality)
False affordance example: axe as melee weapon but not designed as harvesting tool. Solution: take away the axe because you’re never going to teach player that something they’ve known as true all along (axes chop trees) is false
False affordance example: decoration assets that look like gameplay assets
How to teach: put the player in a position where they are motivated to learn the mechanic you want them to learn.
Prioritise the learning you want to happen
Cognitive bias codex one notable one: “We are logical beings” - It’s a lie.
If you craft your tutorial in a way that’s embedded in the experience, then people are actually ok with that.
We’re not designing a tool, we’re designing games. If a game is super useable but not fun, then it’s not a good game.
Tarn Adams
Tarn Adams is the co-founder of Bay 12 Games. His main project is Dwarf Fortress, a fantasy simulation game which he works on with his brother Zach. The game is distributed free of charge and has developed a strong niche following, but it is also known for its difficult interface and text-based graphics. The brothers work on game design together, while Tarn does most of the programming
Long term simulation in game design
On the development on Dwarf Fortress: See this as a case study of Aleissia’s conceptual-level systems talk.
DF started off with very restricted world design.
We had 2d model of the world - created problems v quickly
One thing to think about: local exploration vs dimensions. The more dimensions you have, the more objects have to be bigger to be discovered. In 1d you’re guaranteed to run across the river. In 3d if river is just a line, you’ll never come across it - needs instead to be lakes/caverns/etc
How to have so much complexiity without eating up all computing resource: different levels of realisation based on different levels of zoom.
On the level of detailed information DF captures and simulates: At some point it became a sort of ‘why-not’ thing, which is not a healthy impulse…”
Had a lot of trial and error - bureaucracy system abandoned. Individual coins system abandoned.
Trade/economy system basically abandoned, but by that point difficult to remove: tendrils run deep into the system.
Obviously a lot of these numbers are not exposed to the player, but he argues that’s not necessarily a bad thing. It adds to the sense that the world is self-consistent and rich. Also, if it behaves as players expect them to, it’s not necessarily a problem that they can’t directly influence the system.
DF started as being very game-y When first started. Became less so with move to 3d and over time, but starting off with a very game-y game is not a bad thing.
The point of DF is that the game is an emergent narrative generator
example: Personality system in DF grew a lot more complicated over time. Why have all these detailed personality? One interesting thing is how it ended up being used in RPG mode - it created gameplay incentives that rewards roleplaying and acting according to your character’s personality.
Anecdote:
The cats were dying because they were walking into bars. track alcohol onto their feet and were cleaning themselves - but the bug was that it was as if they drank a whole glass. so cats were vomiting and dying of alcohol poisoning in bars.
Totally didn’t plan that, did not remember cleaning mechanism. Fortunately it’s funny, most of the time.
We’ve filled, probably quite a while ago, the bandwidth of how much reading the player is willing to do (in DF).
At the same time, not surfacing all the information doesn’t feel crucial, as long as people are willing to roll with the punches a little.
You can’t avoid putting your own crap in the game. There are so many implicit assumptions already in DF.
Defining gender roles in the game, for example, was avoided deliberately
“There’s no race stuff in the game, which is not surprising in fantasy world. We’re considering how to deal with the problem.”
Vi Hart
Vi Hart is a mathemusician and philosopher known primarily for work in mathematical understanding, musical structure, and social justice. Hart has publications in the fields of computational geometry, mathematics and music, mathematical art, and math education, and speaks internationally on a variety of topics. Hart is best known for the film series “Doodling in Math Class”, the stand-alone philosophical work “Twelve Tones,” and as co-creator of “Parable of the Polygons.” Current research focuses on how virtual reality technology can impact human understanding and the human experience
Noticing by Design - how games change how we see the world
Not the talk I want to give. Do that one in your head in 5 secs
Noticing-design - a practical guide to manipulating player attention
Not the talk I want to give either. Do that in your head in 10 secs
Designing noticing - Games of paying attention and noticing thing
“First Person Noticers” term from Kate Compton (not the talk I want to give either)
“Noticing is a game mechanic, not an incidental part of human experience”
Designing, Noticing
The actual talk: how designing games affects us
You can’t design a VR game right now because the technology is so visible and new. In the first 5 secs, all people see is the headset and the technology.
Imagine the game world bubble here and real world bubble there. There’s a weak arrow of noticing from gameworld -> realworld Some games emphasise that. i.e. The Witness.
But this needs to be designed/reinforced. You can present the gameworld and convey understanding of a theoretical model, but if you don’t reinforce the noticing in the real world, then it doesn’t happen.
“How do you see something that’s not even in the model of your world?”
Completing the circle
To complete the feedback loop, how to bring real-world noticing back into the theory/model/game?
The key thing here is that you can’t just take everything in the real world back into your model. You need to filter what you bring back into the model/game/theory
To have those filters you need to look beyond your model for wider context. Otherwise, “this is how conspiracies can get made”: If you are great at self-consistant systems…
Aside: when she makes a doodling game on youtube and people are at home drawing along with her on their own notebooks, it’s “real-time client-side rendering”
A lot of the problem with education is that people are just taught the rules of the system without asking critical questions about context. So when Vi makes a doodling game, she names the rules but she also tries to explain why those are the rules.
But also: No amount of perfect filtering will get you a model of the real world
How do you see something that’s not even in the model of your world? (Her answer: listen to and believe in personal stories, and take them into account when you construct your models of the world) In political discourse, it’s all hypothetical and very little to do with personal stories
It doesn’t matter how logical you are or how good you are at judgement if you don’t even notice in the first place or can look without judgment in the first place.
Question: How do you have a game that makes you notice things in the real world and also have your real-world experiences be thigns you can take back into the game and make it better? This is what we get as game designers/artists, But the player doesn’t get that experience
She runs through some of her VR mapping work: “More fun than a hypercube of monkeys”
Aleissia Laidacker
Aleissia is a game programmer-designer-producer who has been working as a developer for 16 years and developing games at Ubisoft for 10 of them. She was Lead AI on multiple Assassin’s Creed games. Her passion for design has pushed her teams to think about why we are developing our games rather than just how we develop them. She empowers her teams to think outside of the box through game jams and experimentation. As part of the Diversity Committee at Ubisoft, she mentors kids and gives talks at schools to inspire young girls to pursue careers in tech and games
Systems are Everywhere… Right Elon Musk?
- Often the obivious and simple things end up being very complicated, especially in games
- She went through a lot of failures to make games that are truly systemic becasue she made them too complicated. Along the way, learnt when to stop, ask questions and then to iterate. There is a method and a process you can use to create systemic games. this talk explains that
-
Game systems: a set of elements with own inputs and outputs, and elements that should influence each other. For a game to be ‘systemic’, it means that there’s a link between all your systems, and they affect each other
- Why should we care about creating systemic games?
- Many oppportunities: the more the player acts upon the systems, the more the game adapts
- Emergence gameplay: allows for unpredictability. As a developer, she loves when players discover ways of interacting with the system
- Quick iterations: forces use of modular framework in developing the game, and therefore allows flexibility
- Replayability: Different story/experience for the player every time. It doesn’t take a monster team to develop systems, and usually invovles creating a lot less content (for same amount of replayability)
- An example magic of systemic games: You are a kitten in a catnip forest. She started playing this and left the browser tab open, and left for a week. “When I came back, unicorns were created in my absence and the cats were sacrificing the unicorns for their tears to satisfy their god in the game”
The Framework
- Systems: made up of game objects and their connections. Goal is to create as many connections as possible between the objects
-
Objects: has visual, sound, physics and gameplay components. Has inputs and outputs. But (in her approach) all of this is controlled through the system’s main logic, optional logic, and feedback
- Think of main logic as the objects in your system. Make the main logic as ‘light’ as possible
-
Think of optional logic as classes, with attributes. It’s where you put the bulk of your code and logic. So for example:
-
nourishment class has attribute is-a-resource-type. You might have all sorts of main logic items (cake, steak, salad) that belong to this class.
- Make sure your classes are as generic as possible but have as many attributes as possible. Once you set up your main and optional logic, the inputs and outputs will flow from that.
-
Very important to build in feedback from the start.
- Once you have your basic main and optional logic set up, you need a way to have an overview of your system
Visualising your system
- She uses visjs.org for visualising node networking
- The goal of the graph is to find islands that aren’t connected to anything. Also, to enable a conversation: Is this game object useful?
- So for example, leaf nodes: are they weak in gameplay terms by definition? Not necessarily, b/c could be story-meaningful, and be an input into an object that affects lots of others
- Graph is also useful when adding new objects: figuring out what effects it would have
Example: adding a new plant object
- think first what inputs and outputs plants should have
- create new main logic object plant
- find what existing optional logic classes already exist
- find what existing feedback you already have
In going through this process, find that she doesn’t have anything in her world that has water as output. So, does she create that part of the simulation?
Think through: water - what input and output should it have? Sample water outputs: impacts organic life, fire, temperature, etc See if those things are already in the world, and whether adding water would create interesting gameplay. Also, does the plant object need water? But the thing is: the more you add, the more you are going to have to think about feedback. The player always has to understand what’s happening in the game
The point is that this gives you a framework for decision making.
You’ve built a sandbox. Now what?
So now that you have a network of game objects, defined main and optional logics and have feedback. But… what you have is a sandbox. A play space where things fit together. but it’s not a game yet
In this sandbox, your player can feel smart (by figuring out the systems), but it doesn’t necessarily mean your game is fun
How to make it fun? Ask, how does the player play? What tools have I given the player?
game mechanics: what tools? what interactions (verbs)?
useful tool for figuring out what you have and have not given players: build cross-matricies that map out what the intended interaction would be between:
- system vs system
- system vs ingredient
- npc vs npc
- (i.e. do all of them)
The goal of doing this is to see what verbs exist. Are any too specific? too common? too overpowered/underpower? for example, so many actions are coded under ‘kill’ this hides interesting gameplay because it reduces everything to that one interaction. What to do: create more tools to create more/different verbs
Matrices help you identify blank holes. For example, game has no tools to manipulate atmosphere. This raises question of whether for objects with a lot of blanks, should they be cut?
How to answer that question? Develop player stories
Example: Make a tools/game objects cross matrix. but instead of filling in matrix with verb, write player stories:
Rope/bucket + fire. Player story: I can grab water with bucket and rope to put out fire but maybe also burn rope to spread fire?
Goal of doing this:
- tools that seemed less useful might actually have use when mixed with other actions
- player stories might identify other tools/systems/ingredients
Tips for finding the fun:
- iteration is key
- the more player stories you develop, the more you refine your systems/tools/verbs
You have a game?
At this point, player feel smart, are having fun, but still something missing: World Impact.
You haven’t yet considered:
- Can a system overtake the world?
- Can a system become extinct?
- Does this impact narrative?
- What’s the ultimate goal/possible ending?
(But, if you want to put a bunch of constraints on your system, why make it systemic in the first place?)
How to think through these problems: again, build cross-matricies. This time, of game objects/impact (lack of, too many, player fantasy)
Goal of this exercise: identify break points, questions to ask, how to refine the system Overall goal: By catering to player fantasies you give them a sense of agency.
Back to the first question
Why make a systemic game in the first place?
Quote from Westworld: “It’s not about giving the gamer what YOU think they want.”
“They are not looking for a story that tells them who they are, they already know who they are.they are here because they want a glimpse of why they could become”
- players want to feel they’ve experienced something that othrers may not have found
- opportunities for creating are for a diverse audience
- stop devleoping games for one type of player/person
One day AI will make games that adapt to any player and the way they play it, but until then we can use smart systemic design to give space for exploration and to achieve those goals.
Jennifer Shahade
Jennifer Shahade is an author, chess champion, commentator and poker player. Her books, Play Like a Girl and Chess Bitch, aim to attract more women to chess. She is the Mind Sports Ambassador for PokerStars and a board member at the World Chess Hall of Fame. She has explored the artistic aspects in chess and poker in projects such as hula chess, Roulette Chess, and in co-writing Marcel Duchamp: The Art of Chess. She graduated with a degree in comparative literature from NYU in 2003
- A lot of people think talent is something you have or not (but that’s not true)
- Originally, the Queen in chess was not ‘the queen’ but vizier and a weak piece until about 1500
- Came from a family of strong chess players. When growing up, was teased: What, you’re a Shahade and you only have 1200/1400 [ELO]? Which, she notes, was not bad for a 10 yr old!
- Chess & Art: Yoko Ono designed a chess set called ‘Play it with trust’. It had all white pieces. The idea was that eventually, ppl would forget whose pieces is whose, and will call a truce. But Jennifer played with that set with grandmasters and no one forgot and there was no truce. “When you get a bunch of gamers together, they will want to find a winner”
- People are often v impressed by chess simulatenous b/c they think the player remembers the board or is simultaneously playing lots of games in their heads. But that’s not the case. 90% of the time, you’re just moving from A to B and making the first move you see (but when you’ve studied chess, your first move is often enough to beat people who haven’t studied it)
- However, what is hard is blindfolded chess. (shows Magnus Carlson doing a blindfold exhibition). To do that you do have to memorise the board and play it entirely in your mind. It requires memory tricks. For example, one of best blindfold players said: had phonographic memory rather than photographic memory so used opponent’s voice to remember the moves.
- Anatolya Karpov: Said he doesn’t like blindfold chess cos it hurt his head
- Jennifer: Blindfold chess feels difficult just for difficulty’s sake. This applies to a lot of memory games
- Chess: meta game now has too many draws. It’s anticlimatic for newcomers
- Recommended house rules for poker: more forced bets; everyone should look at cards immediately upon being dealt.
On the parallels between chess and poker
- very few female professionals, especially in high buy-in games. “Maybe women are just better at managing their money!”
- Women who tend to be more successful in poker and chess tend to play with a more aggressive style. Jennifer’s theory is that it’s mostly incidental rather than due to some causation. Women tend to get a lot more attention in both those games, and aggressive styles are more celebrated
- On playing style in math-driven games: Poker almost all math-based. But sometimes there are two moves with same ‘equity’, and that’s where style comes in. Players who are too mathematical tend to be only ok. and only being ok in poker is not enough to make a living because there’s a rake.
- “You’ll be amazed at what poker players think is a reasonable contract”
- Good poker play is a lot like finding arbitrage opportunities - there’s what the mathematic models say is the correct move. i.e. if you raise, then your opponent should defend his hand 50% of the time. But if you observe that this population of players (in this tournament, for example) defend only 25% of the time, then you make an adjustment and form a heuristic to take advantage: raise a lot more often.
- But this is different in Chess. In chess generally best to just play the optimal move. Because if your opponent is going to be fooled by a suboptimal move, then s/he can be beaten by you playing optimally.
- Sad part about chess: much of the magic of chess is lost to players who are not experts. for example, magical to achieve flow state in a long classical chess game: v visual experience. magic happens when you see an amazing move in your head. But most of the time that move doesn’t show up on the board)
- That’s why commentary is so important for chess.
- Poker is more attrative to viewers than chess because of the money, and also becuase with ‘hole cam’ viewers can play along
- Money stakes inherent to poker but more as a way of keeping score for competitive/expert players. those who actually worry about the money don’t play well
On learning
- It is more important to do lots of chess problems rather than one really hard one. Because you need to do the easy stuff over and over again in practice, so you can see the solution even when you are in throes of competition and stressed out
- AI in chess: bad is that ppl can’t play online for money because of cheaters. Good thing is that everyone can play against AI and study 24/7, so there are now really great players coming out of disadvantaged areas or places that aren’t traditional chess hotspots.
- Studying the right things is important. Some things are so fascinating because they’re crazy situations but they are so rare it’s not worth studying them. much better to study the situations that come up all the time
- Sometimes you get lost in problems that excite you personally. That’s where discipline in learning comes in