Design Attention Not Immersion (Richard Lemarchand)

A4en5on,
Not Immersion 7th March 2012
Richard Lemarchand 1
A4en5on, Not Immersion 7th March 2012
In the minutes leading up to the beginning of my presenta5on, I showed a short film
called Immersion by the photographer and video ar5st, Robbie Cooper.
You can find it on the internet at:
www.youtube.com/watch?v=HfOUhwhdUV0
Good aJernoon, everyone, and welcome to my talk. Please silence your phones, and
fill out the Electronic Evalua5on forms when you receive them. Thanks in advance to
the awesome Conference Associates who are helping us out here today, and all
throughout the conference, and thanks to Robbie Cooper for the use of his short film
Immersion, which I just showed!
My talk today is the first arty, wannabe-‐intellectual game design talk that I’ve given at
GDC, so I’m pre4y nervous in case you think it’s a load of old bollocks, but please bear
with me. It’s also rather long, and I might go over the hour, but I’ll try and go fast.
Let’s do it!
My name is Richard Lemarchand, and I am lucky enough to be a Lead Game Designer
at a fantas5c studio called Naughty Dog. I’ve been a professional game designer for a
li4le over twenty years now, working mainly in character-‐ac5on games, and I’ve
worked at Naughty Dog for nearly eight years.
I was either the lead or the co-‐lead game designer on all three games in the
Uncharted series…
…culmina5ng in Uncharted 3: Drake’s Decep6on, which we released last November.
We’ve had terrific success with Uncharted 3, and even though I’m not going to put
too much of a focus on the Uncharted games today, later on I’ll tell you about some of
the playtes5ng techniques that we used to improve and polish our latest game.
But as well as leaving you with just a li4le bit of prac5cal takeaway today, I have two
other goals for my talk, both of them to do with game design philosophy, and both of
which I hope will be useful to you in your game design prac5ce.
One of them relates to my thoughts about the value and nature of what gets called
‘experien5al’ gameplay.
We’ve used this kind of gameplay to powerful effect in the Uncharted series – for
example, in the peaceful village sequence in Uncharted 2 – but it’s also being used in
many different ways by many different types of games, from the world of triple-‐A, to
indie and art games.
We’ll get back to this towards the end of my talk.
But my first and biggest goal is given away by the 5tle of my talk.
In the grand tradi5on of Jesse Schell’s Game Design Lenses, where Jesse gave us a set
of language for talking clearly about many different aspects of the game experience, I
want to give us just one new lens: the concept of ‘player a4en5on’.
A4en5on, in the way that psychologists talk about a4en5on, is very important when
we’re discussing the core of a good videogame experience, but it rarely gets discussed
by designers here at GDC. I’d like to see if there’s something we can do about that.
Part of the reason I’m interested in this is become I’m hungry for knowledge about
the human mind. And not just the kind of knowledge that science gives us, but the
insight into the lived human experience that we can get from art. And what be4er art
form to explore these ques5ons with, than that of videogames?
The other reason that I want us to start talking about a4en5on is my dislike, some
might say my unreasonable dislike, of a couple of innocent words.
Here they are: poor old ‘immersion’ and ‘immersive’.
Their rela5ves ‘engaging’ and ‘engagement’ are a bit be4er, but not much.
We use these words all the 5me when we’re talking about what makes games great,
but do we really understand what they mean?
Their literal sense seems confusing: when we’re immersed in a game, what are we
under the surface of?
Are we inside the gameplay, or the graphics, like James Woods making out with his
creepy TV in David Cronenberg’s film, Videodrome?
In their landmark 2003 book, Rules of Play, Ka5e Salen and Eric Zimmerman talk
about something that they call, ‘ The Immersive Fallacy’.
They say that there’s an idea, prevalent among game designers and media theorists,
that “the pleasure of a media experience lies in its ability to sensually transport the
par5cipant into an illusory, simulated reality …(one) so complete that ideally the
frame falls away so that the player truly believes that he or she is part of an imaginary
world”.
They go on to claim that this concep5on is a mistake.
The use of the word immersion in this sense, which refers to our going beyond the
boundary between our day-‐to-‐day world of IRL and into some machine-‐mediated
consensual hallucina5on – to borrow some poetry from William Gibson – comes from
the world of virtual reality.
And it’s an alluring idea – certainly the level of iden5fica5on that we have with the
protagonists in a movie or the hero of a game would suggest that we get caught up in
fantasies quite thoroughly, and the amount of 5me we spend ‘grounding’ the reality
of the world of Uncharted in lots of different ways is testament to the importance of
helping an audience suspend their disbelief.
But the claim that you can make someone forget who they are, and start to believe
that they’re someone else, just by enveloping them in pictures and sounds of another
place and 5me, doesn’t seem quite right to me. It doesn’t do jus5ce to the game
ar5sans who craJ experiences that get and hold our a4en5on in many different,
ingenious ways.
And even if we could pull off this trick, of making you believe that you are someone
else – would we really want to?
If I could reach into your mind, make you forgot who you were while leaving your
skills and emo5ons intact, and have you literally believe that you were Nathan Drake,
hanging out the back of a cargo plane with the desert floor a quarter-‐mile below you,
and gun-‐wielding enemies above you, you probably wouldn’t be excited and
entertained in the way that everyone at Naughty Dog hopes for, for players of our
games…
…you’d almost certainly be scared witless! No disrespect, you understand.
I do think that games, interac5vity and what you might call ‘overwhelming sensory
inunda5on’ can change our iden5ty in interes5ng ways that are worthy of a lot more
explora5on. The interac5ve theatrical produc5on Sleep No More and Gaspar Noe’s
film Enter the Void are two great, en5rely different demonstra5ons of that fact.
And here are some recent games that have played with the strange rela5onship
between a game’s protagonist and its player: Jeroen Stout’s Dinner Date, and Swery’s
Deadly Premoni6on.
Both of these games show us that things aren’t quite as simple as our having a one-‐
to-‐one iden5fica5on with a game’s hero.
…as do just about all of the games by Belgium-‐based artgame developer, Tales of
Tales.
Now I don’t want to be that seman5cs-‐quibbling guy. I actually think that the word
immersion is fine when we use it casually, and when we’re sure what we’re talking
about.
Virtual reality is very interes5ng and important, and high-‐fidelity wraparound
simula5ons are immersive in the true sense of the word. But in the context of TV-‐
based videogames the word immersion is a bit of a buzzword, and can point us in the
wrong direc5on when we’re trying to analyze the games we love, especially if we start
talking about being immersed in gameplay.
So how do I want us to talk instead?
Well, it’s really pre4y simple.
Videogames entrance us by geqng our a4en5on, and then they give us what we’d
call a compelling experience, by holding our a4en5on.
Loosely defined as the cogni5ve process of paying a4en5on to one aspect of the
environment while ignoring others, a4en5on is one of the most widely studied and
discussed subjects in the whole of psychology. So let’s take a closer look at it.
For many years, scien5sts believed that the human mind could only pay a4en5on to
one thing at a 5me. Psychologists call this our ‘a4en5onal bo4leneck’.
The game designer Colin Northway has talked interes5ngly about this on his blog in
terms of ‘a4en5on spliqng’.
For example, StarCraJ players call a4en5on “the third resource”, and expert players
a4empt to steal a4en5on from their opponents with raids and harassment.
A4en5onal Bo4leneck is important for game designers to think about. In many games
we have to keep track of lots of different pieces of informa5on: how much health we
currently have, how much ammo we have leJ, where the enemies and other threats
in the level are, and so on.
If the player only has so much a4en5on to go around, we as designers have to stay
aware of this so that we don’t overwhelm them.
We now know that that’s not quite true, that we can only a4end to one thing at once:
our unconscious minds are always monitoring the world for the arrival of important
new informa5on.
We call this ‘vigilance’, and it has been studied a lot since the 1940s, because it’s
central to some very important jobs…
…like air traffic control, military surveillance and even lifeguarding.
Totally relevant to game designers, of course: we’re beqng on the fact that our
players are going to be able to be vigilant, for hours on end, whether you’re keeping
an eye out for aircraJ in Flight Control…
…or staying eyeball-‐dryingly vigilant for incoming enemies in Ac5on Bu4on’s
awesome ZiGGURAT.
But it’s tough to be vigilant, though: we all suffer from something called ‘vigilance
fa5gue’, which means that, aJer the first fiJeen minutes of paying close a4en5on to
something, we become much more likely to miss new relevant informa5on if the
sensory footprint of the new informa5on is small or weak.
Scien5sts used to think that the more boring the task, the quicker vigilance fa5gue
kicks in, but recent research shows that it’s just as much of an issue for interes5ng
vigilance-‐based tasks, like the ones in games.
People focusing on the same task for a long 5me eventually become distracted,
irritable and impa5ent – think of your mental state at the end of a six-‐hour online FPS
jag, or a protracted raid in an MMO. If you’re anything like me, it’s probably
somewhat agitated.
But switching to another ac5vity – especially one that is low stress, or, interes5ngly,
contains pastoral scenes of nature – can restore our ability to be vigilant.
en.wikipedia.org/wiki/Attention_restoration_theory
That probably explains the pa4ern of self-‐guided player ac5vity in a game like…
…Skyrim, as we cycle between explora5on, combat and lengthy periods of messing
with our inventory.
This need that we have, to rotate between different tones and intensi5es of visual
ac5vity, explains why movie-‐makers talk about ‘pacing’.
Here’s an ‘intensity graph’ that writer-‐producer Warren Skaaren made for the movie
Top Gun, as well as Freytag’s famous triangular graph of drama5c structure, and Kurt
Vonnegut on The Shapes of Stories (which you can check out on YouTube):
www.youtube.com/watch?v=oP3c1h8v2ZQ
And we certainly think about pacing a lot when we’re planning the Uncharted games,
and we space out intense periods of combat with less intensive explora5on and
problem solving, and with experien5al sequences like Uncharted 3’s “Lost in the
Desert” montage.
To wrap this sec5on up, I want to quickly tell you about the two different kinds of
a4en5on we have.
The first kind is something I’ve been exploi5ng every 5me I bring up a new slide, using
what’s called your ‘orien5ng reflex’.
This big change in your visual field has almost certainly succeeded in a4rac5ng your
a4en5on to the screen, and it does so in a way that you don’t have any real control
over.
Sudden loud sounds and mo5on have the same affect, as do the men5on of your
name, anything that threatens your survival, and just plain novelty -‐ anything new
and different.
This is called ‘reflexive a4en5on’ and, broadly speaking, it happens at the back and at
the sides of the brain.
The other kind of a4en5on that we have is ‘execu5ve’ or ‘voluntary a4en5on’.
This is the a4en5on that we’re in charge of, which we choose to direct -‐ which we
have what’s known as ‘execu5ve control’ over.
Execu5ve a4en5on is one of a group of execu5ve func5ons, which also include
problem solving and the way we monitor our own ac5ons, and it takes place primarily
in the front of the brain.
But our execu5ve a4en5on can be tricked, as it struggles to organize the world, and it
has a tendency to assume that the world is con5nuous and predictable.
You might be familiar with the brilliant “Person Swap” stunts pulled by the Bri5sh
illusionist Derren Brown, where a vic5m ends up giving direc5ons to two different
people, without even realizing that the person they’re talking to has changed.
You can see video of Derren Brown’s “Person Swap” stunts, here on YouTube:
www.youtube.com/watch?v=UYeJ1BHHDIg
The whole spectrum of execu5ve func5oning, where we make choices, and take
ac5ons, is right at the heart of our craJ as videogame developers.
This is what we’re talking about whenever we talk about ‘agency’, but it seems that
we rarely talk about what the player is choosing to pay a4en5on to.
I wonder if that’s because our execu5ve a4en5on is “out of sight, out of mind”?
In any case, I think we shouldn’t forgot that the players of the games we design have
a freedom and a capacity to express themselves, simply by choosing what to pay
a4en5on to next, before they even act. I don’t know that the mainstream of
videogames oJen take advantage of this in a thoughtul way, and I’ll admit to oJen
being guilty, as a character-‐ac5on game designer, of thinking of the player as pre4y
reflexive, running from one shiny thing that we’ve placed in their path to the next.
Games like the contempla5ve Dear Esther and the atmospheric Proteus take a quite
different approach, of course.
Proteus, developed by Ed Key with a reac5ve ambient soundtrack by David Kanaga, is
one of my favorite games of 2011, and if you enjoy medita5ve interac5ve experiences
and haven’t yet played Proteus, you should seek this game out as fast as you can.
Anyway, that’s a ten-‐cent tour of what cogni5ve scien5sts know about a4en5on. If
you’re interested, there’s lots more good stuff to learn about on Wikipedia.
So a4en5on is probably important for game designers, right? Like other kinds of
entertainment designers, we have to guide our audience’s a4en5on to things that
enrich their experience, and away from things that don’t.
For example, game designers specializing in level layout use the ‘weenie’ technique
borrowed from theme park design, to orient the audiences’ a4en5on to landmarks in
the environment, and then provide sequences of cues and revela5ons that guide the
players to where they choose to go.
If we’re going to talk about the ways that videogames get and hold our a4en5on,
instead of talking about immersion, we’d be4er take a quick look some of the ways
they do it.
When I first sat down to think about this, I decided that most a4en5on ‘grabbers and
keepers’ in videogames fall into one of these three categories:
…beauty, story and gameplay.
This list seemed like a pre4y good star5ng point, but it was going to change, the more
I thought about it.
Let’s begin with beauty. The simplest thing I could think of that gets my a4en5on in
videogames is great art.
You can talk about beauty in terms of symmetry, coherence, harmony and ideal
forms.
You can also express it with logic and number: we know about complementarity of
color, the Golden Ra5o and the Fibonacci sequence.
But when I thought about it some more, beauty didn’t seem to be quite the right
word. Not everything that gets our a4en5on like this is tradi5onally beau5ful: think of
the striking work of H. R. Giger, or how much of the art we love is cute or zany.
So I thought that maybe a be4er word than beauty was aesthe5cs. That way we can
include a wider range of tastes and philosophical ideas.
But then I started looking for a way to describe this in even more general terms, by
talking about something that draws a4en5on to itself by virtue of its difference from
those things around it. So I decided to think like a graphic or interface designer and
call this ‘contrast’.
That way, this category can include anything in our sensory field that leaps out at us,
that grabs our reflexive a4en5on in a way that is useful to our execu5ve a4en5on.
Things in this first category, ‘beauty – aesthe5cs – contrast’, easily get our a4en5on.
They grab our a4en5on quite powerfully, in ways that we don’t have much resistance
to.
However, these kinds of things don’t hold our a4en5on for that long; think about the
way that people only spend an average of 30 seconds looking at even the most
famous pain5ngs in the world. We easily 5re of the contrast in a scene, and our
a4en5on needs to move on to something else.
My second category starts out as ‘story’.
Of course, story is a hot-‐bu4on topic for game developers. My hope is that if we talk
about it in terms of a4en5on, we can make it less loaded.
Story and games are almost certainly oldest human cultural tradi5ons, and both
demand and occupy a4en5on.
Think about an early human shaman holding his village entranced around a fire with
his tales of the gods and the birth of the universe.
However, unlike the linear story that’s woven into the playful space of Uncharted, not
many videogames have a story in a tradi5onal sense, so I pre4y quickly changed this
sec5on to ‘narra5ve’, to encompass many different kinds of games, from Rock Star’s
brilliant systemic narra5ves to Daniel Benmergui’s new game, Storyteller.
Even the fact that a rook is shaped like a castle, and a knight is shaped like a horse, is
narra5ve.
But to generalize even more broadly, this is all to do with social stuff. A 2007 study at
Florida State University showed that we exhibit ‘a4en5onal adhesion’ not only
towards people that we’re a4racted to, but also towards people that we perceive as
rivals for their affec5ons.
Suddenly, it seems scien5fically comprehensible, the interest that we all, men and
women, gay and straight alike, have in the a4rac5ve, noble, craJy cast of…
…Downton Abbey.
Opening the door to the social in this way will also let us talk about how social games
of all kinds hold our a4en5on…
…whether it’s a split-‐screen co-‐op game of Uncharted 3, or the social aspects of
Facebook games.
The developmental psychologist Michael Tomasello has wri4en about something

called ‘the joint a4en5onal scene’, and describes how a shared experience of a third
thing plays a fundamental role in our intellectual and social development, and might
even be what gives us our essen5al humanity.
Professor Janet Murray, author of Hamlet on the Holodeck, has run with the idea of
‘joint a4en5on’ in the direc5on of games – you should check out her work.
One paper in which Professor Murray of the Georgia Ins5tute of Technology discusses
joint a4en5on is:
Toward a Cultural Theory of Gaming: Digital Games and the Co-‐Evolu5on of Media,
Mind, and Culture
www.lcc.gatech.edu/~murray/PC0403_Murray.pdf
Thinking about the social also made me think that this is why we’ve had so much
success with the character-‐driven stories in the Uncharted series.
By puqng a focus on the people in the game, their fears and hopes, and the moment-‐
to-‐moment nuances of the rela5onships between them, we’ve been able to get and
hold people’s a4en5on much more easily than by trying to use more abstract, plot-‐
driven stories that focus on the power structures or the history of a fic5onal world.
Now, I think it’s averagely tough to get someone’s a4en5on with this ‘story –
narra5ve – social’ stuff. Screenwri5ng books like Blake Snyder’s Save the Cat are full of
advice about the challenging task of quickly establishing characters that are likable or
empathic, so that we’ll want to follow them along their path.
I think that narra5ve also has an averagely strong hold on people’s a4en5on, and in a
way that is curious because it doesn’t decay very quickly over 5me.
Episodic fic5on, where we might have to wait days or even weeks between chunks of
story, is a testament to that.
There’s an anecdote about people in the 1800s lining the docks in New York City,
wai5ng for the latest episode of Charles Dickens’ serialized story, The Old Curiosity
Shop, to come in on the boat aJer a par5cularly spectacular clivanger.
My last category is gameplay.
Games are made of several parts. They have rules – their mechanics – that interact in
a systemic way with players and resources to produce dynamics, which in turn yield
emo5onal responses – aesthe5cs – for the players.
You’ll recognize this as the ‘MDA’ model proposed by Robin Hunicke, Marc LeBlanc
and Robert Zubek – if you haven’t read their paper about MDA, you should definitely
look it up on the web. It acts as a great, concise explana5on of how gameplay works
to hold our a4en5on and impact our emo5ons.
www.cs.northwestern.edu/~hunicke/pubs/MDA.pdf
The philosopher Bernard Suits defines a game as the voluntary a4empt to overcome
unnecessary obstacles in pursuit of specific goals.
Goals are what make a game different from a toy or a free-‐play session. We work
towards long-‐term goals by means of a4ainable short-‐term goals that we put
together strategically to get there, like “roll the dice” or “shoot the enemy”, or even
as simple as “take cover”.
These build into compulsion loops by way of leqng us reach sa5sfying points of
resolu5on – the end of my turn, the dropping of the enemy…
…or the collapsing of three clumps of grass into a bush in Spry Fox’s brilliant Triple
Town. We string these loops and goals together into a gameplay experience of many
hours.
This is also related to Jaime Griesemer’s famous concep5on of the ‘thirty seconds of
fun’ in Halo – where an underlying pa4ern of player ac5vity is repeated, with
con5nual varia5on, to create the long-‐term enjoyment of the game.
But there’s a problem with calling this category ‘gameplay’: not all the videogames we
love even have goals, and so by our best defini5ons they’re not very gamelike, strictly
speaking.
MinecraO, before the single player game came along at least, and SimCity are two
videogames that are very open-‐ended, but which I’ve played compulsively for dozens,
maybe hundreds of hours; they are clearly doing something right to hold my
a4en5on, but quite how they’re doing it has always been something of a mystery to
me.
I’ll come back to this later.
So I almost changed this category to ‘interac5vity’, but let’s call it ‘ludism’ instead, like
‘ludic’, from the La5n ‘ludus’, which describes play of all kinds, not just games, and
which acknowledges the playful aqtude at work here.
This scheme needed a bit of preten5ous La5n, anyway.
I think that when the human mind finds a system, it loves to study it, if the moving
parts of the system are easily visible to us. Chris Burden’s art installa5on Metropolis II,
which I recently saw at the LA County Museum of Art, is a system whose dynamics
are plain to see, and it’s fascina5ng. Those are all Hot Wheels cars.
You can see a short film about Metropolis II on YouTube, here:
www.youtube.com/watch?v=llacDdn5yIE
We come for the regular structure and repe55on in the system, which gives our
minds something to grasp quickly -‐ like the way that waves lap into the shore…
…or the self-‐similarity of this Romanesco broccoli -‐ but then we stay for the variability
in the system -‐ the ever-‐changing pa4erns in the flickering flames of a fire, or the
variable schedule of reinforcement in the payout of a slot machine.
So to make things even more inclusive, I’ll also call this category ‘systems’.
When a system of rules, resources, interac5ons and goals are well designed,
gameplay is like mental catnip to us. I think that this gameplay – ludism – systems
category is the hardest to get a player’s a4en5on with; for example it can be tough to
teach a board game to someone who hasn’t played it before and isn’t that interested,
and we have to work very hard to design videogame opening sequences, like the
beginning of Uncharted 2, which grab new players’ a4en5on quickly. Simple
mechanics probably get your a4en5on more quickly, since they’re easier to
understand.
But as Jason Rohrer pointed out in his keynote at the Montreal Interna5onal Game
Summit last year: games are clearly the form that is best at keeping people’s
a4en5on. While a film struggles to keep you for longer than three hours, tops, a great
videogame can hold your a4en5on for dozens, hundreds, even thousands of hours.
This is also where the depth of a game becomes important. Game depth is a li4le
hard to define, but let’s say a deep game is one that produces a large number of
interes5ng choices in a comprehensible way from a compact, elegant set of rules. The
deeper the game, the greater the opportunity it has to hold your a4en5on for a long
5me with its gameplay alone.
It’s probably true to say that a game holds your a4en5on best when you feel that you
haven’t completely understood its complexity. Thanks to my colleague Kaitlyn Burnell
for that insight – check out her talk on Friday.
I hope you’re somewhat agreeing with me by now, that we can cut through the
confusion about what ‘immersion’ and ‘engagement’ mean simply by saying: good
videogames get and hold our a4en5on, using a mix of the elements in these three
categories.
Also, you can start to see how we could freely mix and match stuff from these
categories to make different styles and genres of game, drawing on them to serve
each of our unique, individual crea5ve goals.
We can also keep them in mind to use them when we need them. Just when we’re
losing the player’s a4en5on with one category, we could pull them back in by
deploying another. A very simple example of this would be the way that we vary the
graphics and add new mechanics into the mix as a game progresses.
We oJen get into difficult arguments, when we try to compare videogames to other
forms like the movies.
By using this concept of a4en5on to look at other forms, I hope we can get a clearer
idea of what videogames are and aren’t like. This framework cuts us loose from
subjec5ve, ideological evalua5ons of the value of different forms and leaves us free to
evaluate what does and doesn't work: func5onally and ar5s5cally
Whether we’re discussing games, film, sports, or even just some drunkard falling over
at a party, the kinds of statements we can now make are simple and to the point, and
we can more easily draw out the ways in which each of these things are similar to,
and different from one another.
As we learned from Colin Northway and the StarCraO players: human a4en5on is a
limited resource – there’s only so much of it, and there’s a cap on the rate at which it
can be parceled out. It diminishes and recharges over 5me.
But you can improve your a4en5ve capacity. In fact, the games we make are helping
people to do just that. Ongoing research from Daphne Bavelier’s group at the
University of Rochester shows that people who play ac5on videogames have be4er,
faster visual a4en5on skills than people who don’t.
psych.wisc.edu/CSGreen/csg_nature_03.pdf
www.bcs.rochester.edu/people/daphne/VisionPDF/hubertwallander.pdf
Since a4en5on is a resource, you’d think that it would be great subject ma4er for
videogames.
There haven’t been that many games that are explicitly about a4en5on – you can find
some very simple a4en5on-‐training games on the web – but there are a couple of
interes5ng ones that I’m aware of: Eric Zimmerman and the team at Gamelab made a
game a few years ago called Arcadia, where you had to play mul5ple ac5on games at
once…
…and Chris Hecker’s game Spy Party is all about using the player’s a4en5on as a
resource that they have to manage.
In fact, I’d like to claim that a4en5on is the basic currency in which videogames trade,
and in which nearly every other cultural form trades, too.
I really wanted to talk about game addic5on today, which is a difficult, polarizing but
interes5ng subject, and I can’t help but feel that we should talk about it more as a
community of game designers. I kept coming back to it when I was wri5ng this talk,
because of the rela5onship between game addic5on and a4en5on.
Addic5ve games pull your a4en5on back to them over and over again, and then hold
it for a long 5me. Games that cause people addic5on-‐related problems in their lives
are perhaps too good at geqng and holding that par5cular person’s a4en5on.
Unfortunately, it would take more than an hour to do this subject jus5ce, but I hope
that someone at another GDC can come back to this subject and run with it.
So the player has all this a4en5on, which they’re bringing to bear on our games. How
is that prac5cally relevant for game designers? I don’t have 5me to speak to every
kind of game, so I’ll tell you about our experiences from the development of
Uncharted.
OJen in our games, we want to direct the player’s visual a4en5on to an aspect of the
environment – a ladder, a door, an enemy, or even a simple ledge.
There are lots of things that impact the visibility of things in an environment: the
silhoue4e of an object, the background that it’s against, its hue, satura5on and value,
all of which are affected by the ligh5ng of the scene.
Then there’s the composi5on of the view of the scene and the lines and shapes
within it, the density of detail and any anima5on that it contains. If even one of these
things is out of whack, the player won’t see what we need them to see. Every
Uncharted game has presented literally thousands of these kinds of challenges for us
to overcome.
There are a couple of simple 5ps to know about here. We’ve found that players rarely
look outside the central part of the screen unless it’s to look at meta-‐informa5on like
the HUD, and that causes them to miss things that might be in plain sight, but that are
out in a corner.
Also, players are more a4en5ve to things in the lower half of the screen than the
upper half, because of something called ‘visual gravity’.
Bruce Block talks about these elements and many more in his excellent book, The
Visual Story, which should be required reading for every videogame developer, and I
also like this book, Composi6on from the great DK Eyewitness Art series, for some
simple 5ps about the way that ‘lines of force’ in a composi5on drive your eye around
a scene.
The takeaway for game designers here is that subtle details in the art of your game
don’t just make it more or less pre4y: they make it possible for the player to play your
game, at all.
Robin Hunicke, who works at thatgamecompany, told me that working on Journey
they created a new term for this: they called it ‘a4en5onal design’, and you can see it
at work in the opening moments of their game.
The way that the player’s a4en5on is drawn first to the character that they will play,
by the character’s isola5on in the frame, the sugges5on that the character is now
controllable, because it has risen to its feet. The establishment of first the hill as a
short-‐term goal for the player, through the use of a4en5on-‐grabbing detail and
anima5on, and then the mountain as a long-‐term goal for the player, by a change in
camera posi5on that frames the composi5on in such an impactul way: all these are
subtle ways of designing to direct the player’s a4en5on towards the mechanics and
goals of the game.
Robin told me that even these simple-‐seeming things took many hours of work to
accomplish, and I know from our work on the Uncharted series just how hard we
have to work to direct the player’s visual a4en5on in subtle ways like this.
Now Visual A4en5on is a lot weirder than you might think. We can only see a clear,
sharp image with a 5ny part of our re5nas – something called the fovea. It’s right in
the middle of the re5na, it’s about a millimeter across, and it’s responsible for a full
half of the signals that travel down your op5c nerve.
If you hold your arm out at arm’s length and look at your thumb, then the width of
your thumb is about the size of this 5ny piece of visual acuity you have.
Because of this, we can’t look at a scene in fixed steadiness, the way some other
animals like birds do. Instead, we build up a 3D picture of the world by zooming our
eyes around it with quick, simultaneous movements called saccades, punctuated by
periods of rest called fixa5ons.
Saccades are actually the fastest movements that the human body can make, and we
tend to fixate on clusters of detail, explaining a lot of the stuff that’s in the Eyewitness
Art Composi6on book.
You can see psychologist Jiří Lukavský‘s “Fovea5on Movie” which explains this
phenomenon, here:
www.jirilukavsky.info/foveation-movie
So knowing this, you’d think we would want to set up eye tracking equipment to be
able to see exactly what our players are fixa5ng on. And boy, would I like to.
I’ve always been fascinated by the kind of technology that can be used to see exactly
where on a screen someone is looking, and I’m itching to get my hands on some eye-‐
tracking kit, but it’s not something we’ve done at Naughty Dog.
Valve have had some experience using eye tracking as part of their playtes5ng
process -‐ in fact, they even rigged Portal 2 so that you could play it using your gaze to
aim the Portal Gun around the level.
This kind of technology is an amazing boon for paralyzed people, of course, and we’ll
be seeing a lot more of it in the future. It’ll become cheaper and more usable, and I
think that eye-‐tracking technology could well become standard equipment for game
developers, and maybe even game audiences, over the next ten years.
However, right now, eye-‐tracking technology is s5ll very expensive – around ten
thousand dollars a seat, I believe. And even if you could afford eye tracking, it might
not be desirable. In his excellent talk at GDC in 2009, Valve’s biometrics expert Mike
Ambinder men5oned a study that showed that, when people know that their eyes
are being tracked, they move them around somewhat differently.
So how else can we keep track of the player’s visual a4en5on, to make sure that the
a4en5onal design techniques that we’re trying to use are working? Well, of course
we study our player’s a4en5on all the 5me, when we’re using one of our most
important game design tools: namely, playtes5ng.
I’m sure everyone here knows about the importance of playtes5ng, although you’d be
surprised to hear how many game developers s5ll make do without it.
We playtest our games all the 5me on each other, as we’re developing them, of
course.
But the kind of playtest I’m talking about now is formal playtes5ng, with groups of ten
“Kleenex” playtesters from the general public who have never seen our game before,
and who we’ll only use to test it once.
When I joined Naughty Dog, to help finish Jak 3, we ran perhaps four or five playtests.
For Uncharted 3, we ran twenty-‐one tests over the course of the last six months of
development.
We run these tests in-‐house, in a dedicated playtest room. Each playtester’s sta5on
has a networked PlaySta5on 3 with the build of the game we’re tes5ng, headphones
and a networked DVR box that will capture video of the game to our network as they
play it.
Screens separate the sta5ons from each other, so that the players can’t see each
other’s games, even accidentally, and we ask them not to talk during the playtest.
Because we want to be as scien5fic as possible, we’re ruthless about not giving them
any help with what they’re playing. As the players play, the game records certain
informa5on about the gameplay session and posts it to a database on our network -‐
we call this our ‘metric data’.
At the end of the playtest, we get the playtesters to fill out a ques5onnaire about
their experience, and conduct an exit interview, which we record. These are useful to
help track the improvement in the players’ percep5ons of the game from test to test.
We always see a slow, gradual improvement, which helps us stay sane in the
knowledge that the game is geqng be4er, and we also get some interes5ng, if
anecdotal, game design perspec5ves from the exit interview.
Immediately aJer the playtest, we take a look at the metric data, which lets us
visualize how playable the game is.
One way we do this is by compiling a table of how many 5mes the players died in
each small chunk of the game, using the condi5onal formaqng in Excel to see when
the average and maximum deaths for each bit of game exceed certain thresholds.
This is very useful for aler5ng us to the parts of our game that are s5ll too hard.
But back to a4en5on. One way that we can track our players’ a4en5on without
expensive equipment is ridiculously simple: it’s just by watching them play – or by
watching their gameplay videos later. In a game like ours, it’s easy to see when
someone isn’t seeing something, or has forgo4en their goal.
When players repeatedly run right past the object they’re meant to be interac5ng
with, it’s a pre4y sure sign that they can’t see it. If they go up to it from 5me to 5me
and try to use it, but then leave it alone for a long 5me, you know they can see it but
they don’t think it’s important.
We make notes about these situa5ons as we see them arise, make fixes between
playtests, and then watch out for improvements in the next test. Just by doing this,
you can make an amazing difference to the way your players’ a4en5on moves over
the sensory surface of your game, and you can seriously improve the amount of fun
that players have with it.
But what about tougher problems, where the player’s a4en5on isn’t so easy to
observe? Well, you can use your metric data in some clever ways, to get a handle on
these kinds of issues.
The environments in Uncharted are visually very dense – thanks to our brilliant
ar5sts, there’s a lot going on in the picture of any random Uncharted screen shot. Like
we were just discussing, it’s easy for things that are important to gameplay to get lost
among all that visual informa5on.
Because of this, we faced a nasty recurring problem in the Uncharted games, one that
dated all the way back to the beginning of the series. We found that players would
oJen have a difficult 5me spoqng the edge grabs in the environments.
This was usually a disaster for us, because it would stop them climbing onwards to
the next part of the game. The problem was compounded by the fact that players
would also get distracted by things that looked like they were climbable, but weren’t.
The solu5on to this problem was quite brilliant, and it came from these three
handsome men: Lead Technical Ar5st Teagan Morrison, Lead Gameplay Programmer
Travis McIntosh and Gameplay Programmer Jaros Sinecky.
What we did was to set up a system that we only used during playtests, that would
record an XYZ coordinate every 5me the player pressed the jump bu4on and didn’t
end up jumping up to a ledge, but instead jumped up and down on the spot.
We wrote those coordinates to the metrics database on our network, and then when
the playtest was complete, exported this data back into the game and put a li4le red
sphere where every thwarted jump had taken place for every player in the test.
We called this our ‘bad jumps’ system. You could immediately see the bad jumps
clustered beneath objects that looked like edge grabs, but weren’t, and they told us
what we needed to fix.
In the days aJer each playtest, the environment ar5sts would go through their levels
with the ‘bad jumps’ turned on, changing the artwork of the things that our
playtesters thought had grabbable edges, to make them look less grabbable.
We were able to make a huge difference to the game by doing this, and we were
really sa5sfied by the results that we got from seqng up just this simple system.
We invented some other similar a4en5on-‐tracking metrics for Uncharted 3. How
oJen did players pick up a new gun, hold at it for a moment, and then drop it again in
favor of their old gun? That was easy to record in our metric data.
So was the data you can see here: the number of successful and failed a4empts that
players made to throw back grenades, instead of ignoring the throw-‐back mechanic.
This technique seems like it could be applied to widely different kinds of games, too -‐
I’m sure you can imagine them for whatever style of game you’re working on,
depending on its mechanics.
Before I wrap up this sec5on, I want to assure you about something that always
bothers me whenever I hear a talk about the psychological predictability of human
beings, and the way that we can use numerical measurements taken from our players
to make decisions about the crea5ve path that we should take with our games.
These are tools, like any other, and just like a knife, they can be used to peel a
delicious apple, or commit a horrible murder. It’s oJen good to be scien5fic in a
technical art like ours, but you shouldn’t go any further with these techniques than
your gut ins5ncts as a designer and as a crea5ve person tell you that you should go.
What you’re trying to do is to make sure that your game is landing with players in the
way that you intend it to land. You’re trying to get rid of the stupid oversights that
stop your game from being fun and interes5ng, while leaving in enough chewy stuff
that players want to sink their teeth in to your game, over and over again.
This also applies to my whole invoca5on of the concept of a4en5on. I’m not trying to
be reduc5ve, and explain things away in terms of psychology and brain science: that
would be a mistake. This is just one more lens that we can use to look at what we’re
doing when we’re designing a game.
So now let’s get to the topic I want to close on: the subject of ‘experien5al’
videogames, what they have to do with a4en5on, and why I think they’re important.
I’m sure that some of you have been wondering why I haven’t yet men5oned a
concept that seems obviously connected to a4en5on and the psychology of
videogames. I’m talking about the concept of ‘flow’.

Flow is, of course, the state of op5mal experience described by the psychology
professor Mike Csikszentmihalyi, and it’s been much discussed here at GDC over the
years, for the strong 5es that it has to games.
When we’re in a flow state, our a4en5on is held and is very highly focused, whether
we’re doing needlepoint, speed-‐typing or playing Triple Town. Our a4en5on is right
where it needs to be: on the subject ma4er we’re engaged with, helping us handle
the tasks at hand while our vigilance watches out for incoming tasks that we’re going
to have to deal with.
We lose some awareness of ourselves, of our bodies in par5cular, and we experience
strange 5me distor5on effects, as hours seemingly pass in minutes, or moments
stretch out, allowing us to act more rapidly than we otherwise could.

When we are in a flow state, we are riding the boundary between mastery -‐ where
things are easy -‐ and challenge -‐ where our abili5es are being tested. If the system’s
too hard, we’ll get frustrated, if it’s too easy, we lose interest. In either case, we drop
out of a flow state: the system has lost our a4en5on.

When we are in a flow state, we are ac5ng in a system as if it was a game,
manipula5ng the system towards certain goals.
However, we wouldn’t always want to call this system a game – surgeons are oJen
used as example cases of people who go into a flow state, and we agreed earlier that
some videogames like MinecraO don’t even have explicit goals, and yet millions of
people, myself included, go straight into a flow state when they pick MinecraO up.
In trying to solve the puzzle of how we can go into a flow state in non-‐game systems, I
finally understood how MinecraO holds my a4en5on so completely.

In the public debate about gamifica5on, we’ve talked a lot about extrinsic and
intrinsic goals.
We know from educa5onal and industrial psychology, going all the way back to the
1930s, that intrinsic goals are be4er mo5vators than extrinsic goals. It’s the playing of
a game for its own sake that holds our a4en5on, not the promise of some eventual
reward.

I realized that I create my own intrinsic goals in MinecraO, and that is what puts me
into a flow state. I make a game of MinecraO, by imagining a back-‐story where I need
to stay alive, to unlock the mysteries of my uniquely-‐generated world, and in order to
stay alive, I have to engage with the system around me, exploring the environment,
and using it to make the tools that aid me in my quest.
The goals set for themselves by other people who play inside the system that is
MinecraO might be quite different, like building a scale model of the USS Enterprise
or a working 16-‐bit processor.
What’s common for all of us is that we’ve had a profound experience of a flow state
that was only possible because of the goal-‐seqng aqtude that we brought to bear on
an open system.
(At GDC, I showed a por5on of the movie “Between Two Ci5es [MinecraJ 7.5 km
Railroad]”, uploaded to YouTube by NewEonOrchestra.)
www.youtube.com/watch?v=g-tSxazq2BA

This is what the Csikszentmihalyi calls ‘the Autotelic Experience’ (from the La5n auto
– self, and telos – goal).
Autotelic people are internally driven, and exhibit a sense of purpose and curiosity.
Because of their nature, they’re more autonomous and independent, and less easily
manipulated with threats or rewards. They love to explore systems, for the sake of
the exploring.

All human beings are autotelic, to some extent, but our genera5ons, the ones that
have grown up playing videogames, are very lucky, because we’ve been trained to be
more autotelic than any other genera5on of humanity before us.
The games that we’ve played have cul5vated our curiosity, our independence and our
ability to manipulate rich systems in goal-‐seqng ways.

I’m sure you or people you know have had deep, rich play experiences in Red Dead
Redemp6on, Far Cry 2 and Skyrim without following the main storyline of the game at
all, or by running pacifist play-‐throughs: this is testament to the power of the
autotelic experience.

As far as I know, Daniel Benmergui first coined the term ‘experien5al games’ on his
blog all the way back in 2007.
Broadly speaking, these are games that forgo explicit goals and tradi5onal game
compulsion loops in favor of systema5zed, open-‐ended, explora5ve experiences,
oJen in pursuit of an ar5s5c impact.

Early a4empts at these kinds of games and interac5ve art experiences frequently
struggled to hold people’s a4en5on for very long, compared to tradi5onal
videogames, because they had a tendency to rely on aesthe5cs and narra5ve – the
categories that are the least effec5ve at holding our a4en5on.
However, as game aesthe5cs and narra5ve elements have become more systemically
interes5ng, new opportuni5es have opened up for experien5al games because of this
autotelic drive we carry round with us. In my IndieCade talk last year I described the
way that Tale of Tales’ game The Graveyard…

…gave me the confidence to push through the experien5al “Peaceful Village”
sequence in Uncharted 2, even when some people on our team didn’t think it was
going to be a success.
The calm beauty of the scene, and the pleasure of the simple, systemic narra5ve
elements we sca4ered through it held people’s a4en5on for just long enough to
create a powerful moment at an important point in the game.
Players who were curious and explora5ve found plenty to reward their curiosity.
Players oriented towards the ul5mate goals of the game were able to move through
the scene quickly, and soaked up enough of the calm atmosphere to reset the
intensity of the game in prepara5on for the its final act.
You can find my IndieCade talk online if you’re interested to hear more about this.
www.g4tv.com/videos/55682/indiecade-2011-keynote/
www.youtube.com/watch?v=ZPju14BmOLY

Videogames are special, in their opportuni5es for non-‐linearity of meaning, but
they’re not unique in this regard. Philosophers have thought for years that so-‐called
linear ar5s5c forms offer similar opportuni5es.
In his 1962 book, The Open Work, the Italian author and scholar Umberto Eco, most
famous for his novel, The Name of the Rose, argued (and I’m quo5ng from Wikipedia
here, rather than Eco himself):

“…literary texts are fields of meaning, rather than strings of meaning (…)
“…they are understood as open, internally dynamic and psychologically engaged
fields.
“Literature which limits one's poten5al understanding to a single, unequivocal line,
the closed text, remains the least rewarding, while texts that are the most ac5ve
between mind and society and life (…) are the most lively and best”
en.wikipedia.org/wiki/Umberto_Eco

In other words: singularity of meaning is the enemy of art.
Great art in all of its forms is mul5plicitous, systemic and can have many meanings,
even contradictory ones, which is why the same piece of art can speak to us in
different ways at different 5mes in our lives.
If we can expect to find new things each 5me we come back to a work, then that in
itself is a draw for our a4en5on.

Amy Hennig, Naughty Dog’s Crea5ve Director and head writer on the Uncharted
series, talks about the very great value of leaving something to an audience’s
imagina5on, in terms of the storytelling in our games, so that the individual player
fills in the details for themselves, and the experience of the game truly becomes their
own, even before they have taken an ac5on in our interac5ve system.
To put it another way, if linear games are well wri4en, they’re not truly linear.

We’re learning more all the 5me about how to devise gameplay that expresses a
broad range of emo5on.
That’s crucially important for pushing games forward as an ar5s5c form, as Jonathan
Blow’s Braid demonstrated.

But I worry that we have a tendency to over-‐focus on compulsion loops and explicit
goals in the games we make, and we tend to look past the emo5onal value, and
ar5s5c value, brought to a game by things like nuanced aesthe5cs, the narra5ve and
the social, and the space for autotelic self-‐determina5on created by open systems.
In his GDC talk last year, An Apology for Roger Ebert, Brian Moriarty gave us the
defini5on, “Sublime art is the s5ll evoca5on of the inexpressible.” When we’re trying
to make games that succeed ar5s5cally, our ul5mate goal is to evoke ideas and
emo5on in the player about subjects that are difficult to express.
But as Brian suggested in his talk, and as Csikszentmihalyi himself says, the flow state
we enter when we’re caught up in the formal parts of a game is almost emo5onless,
as an essen5al part of its character. Our minds are just too busy when we’re dealing
with all the ra5onal tasks that a game presents, to allow in the subtlety of feeling that
we’re always chasing as ar5sts.
But I see room for both the flow of gameplay and reflec5ve experience in our games
if, from 5me to 5me in the progression of a game, we have the confidence to let the
player just drop out of their flow state.

As the player’s a4en5on begins to shiJ away from its laser focus on the goals of the
game, whether those goals are self-‐created or imposed by the designer, then we have
an opportunity to catch the player in just the right way with an affec5ng piece of
music, or a view something wonderful, or just a look from another character, and in
doing so, we can create one of those powerful moments that we cherish in our
favorite games.
Have you ever stopped to watch the sunset in MinecraO, or to admire a view in
Uncharted?
Think about how you are, in that moment, as your responsible, logical goal-‐oriented
mind gives way to the uncensored emo5on of your innermost self.

For me, this boundary, where the cool clarity of focused ra5onality meets the chaos
and immediacy of raw emo5on, crea5ng in us a lived experience that is uniquely ours
in both its intellectual and in5mate quali5es, is the richest and the most exci5ng
aspect of videogames as an expressive medium.
Game designers are oJen terrified to let gameplay’s iron grip on the player’s
a4en5on drop, for fear of losing that a4en5on forever.
But we can see from what we’ve looked at today, that there’s no need to panic: if the
player has come to trust you, they’ll stay with you, and we have so many tools that
we can use to get the player’s a4en5on back again in a meaningful way, that the
biggest challenge we face will be deciding which one to choose.

Ar5sts from Shakespeare to David Lynch to Fumito Ueda have understood how to
shape emo5on by guiding a4en5on while leaving room for personal experience, and
in doing so they have impacted our lives in countless ways.
The emo5onal results that Marc, Robin and Robert describe in MDA are the point of
games, no ma4er how we get there.

There’s no one right path with this. We developers are free to play in this domain, and
make our own discoveries about what works.
We hope we’ve shown with the Uncharted games, that the techniques of cinema can
make a great mix with videogames.
But we mustn’t lose sight of literary fic5on, of poetry, theater and music as vital forms
of self-‐expression that we’re only just beginning to synthesize effec5vely with the
crea5ve opportuni5es that videogames present.

There’s one piece of knowledge that I’m certain we all share.
It’s that all these things, new technologies and be4er development prac5ces, the
mechanisms and mysteries of the human mind, and the great powers of curiosity and
self-‐determina5on that we all have, are, in videogames, making a new kind of cultural
form possible; one that is ac5ve between mind and society and life.

That’s my talk. Thanks very much.

The Art of Game Design: A Deck of Lenses – Jesse Schell
Rules of Play: Game Design Fundamentals – Ka5e Salen and Eric Zimmerman
The Grasshopper: Games, Life and Utopia – Bernard Suits
The Shapes of Stories (on YouTube) – Kurt Vonnegut
Hamlet on the Holodeck and the “Joint A4en5onal Scene” – Janet Murray
Save the Cat! – Blake Snyder
MDA: A Formal Approach to Game Design – Hunicke, LeBlanc & Zubek
The Visual Story – Bruce Block Composi5on – DK Eyewitness Art
Flow: The Psychology of Op5mal Experience – Mihaly Csikszentmihalyi
An Apology for Roger Ebert – Brian Moriarty

Design Attention Not Immersion (Richard Lemarchand)

Uploaded by

Copyright:

Available Formats

Design Attention Not Immersion (Richard Lemarchand)

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Design Attention Not Immersion (Richard Lemarchand)

Uploaded by

Copyright:

Available Formats

A4en5on,

Not Immersion 7th March 2012

You can ﬁnd it on the internet at:

Here they are: poor old ‘immersion’ and ‘immersive’.

So how do I want us to talk instead?

Well, it’s really pre4y simple.

…like air traﬃc control, military surveillance and even lifeguarding.

…beauty, story and gameplay.

My second category starts out as ‘story’.

The developmental psychologist Michael Tomasello has wri4en about something

My last category is gameplay.

I’ll come back to this later.

This scheme needed a bit of preten5ous La5n, anyway.

Richard Lemarchand 100

Richard Lemarchand 101

Richard Lemarchand 102

Richard Lemarchand 103

Richard Lemarchand 104

Richard Lemarchand 105

Richard Lemarchand 106

Richard Lemarchand 107

Richard Lemarchand 108

Richard Lemarchand 109

Richard Lemarchand 110

Richard Lemarchand 111

Richard Lemarchand 112

Richard Lemarchand 113

Richard Lemarchand 114

Richard Lemarchand 115

Richard Lemarchand 116

Richard Lemarchand 117

Richard Lemarchand 118

Richard Lemarchand 119

Richard Lemarchand 120

Richard Lemarchand 121

Richard Lemarchand 122

That’s my talk. Thanks very much.

Richard Lemarchand 123

The Grasshopper: Games, Life and Utopia – Bernard Suits

The Shapes of Stories (on YouTube) – Kurt Vonnegut

Save the Cat! – Blake Snyder

The Visual Story – Bruce Block Composi5on – DK Eyewitness Art

Flow: The Psychology of Op5mal Experience – Mihaly Csikszentmihalyi

An Apology for Roger Ebert – Brian Moriarty

Richard Lemarchand 124

You might also like