Over the last thirty years we have seen significant advances in computer hardware we use to develop and play games. We are now able to simulate complex physics and light matter interactions, display believable animated characters as well as provide players with expansive worlds used as settings for designer stories and playgrounds for player own stories. All of that comes at a significant cost expressed in expanding team sizes, ballooning game budgets and lengthening production timelines. To stay competitive teams feel the pressure to focus on finer and finer details in game design, art and engineering. Would these effort pay off though? Will players pay attention, perceive and be motivated enough to continue playing? These are the questions we need to often ask ourselves since human players have very real limits to their attention, perception and motivation that may well make many of the investments irrelevant. To help leads and producers with the decision making this talk will discuss the following topics: - Limits of human working memory - Interactions of visuospatial sketchpad and phonological loop - Inattention blindness - Types of human memory and memory degradation - Perception of time passage - Overview of human visual system - Central vs. peripheral vision - Perception of shapes and shadows - Perception of illumination - Perception of movement - Frame rate and human visual system - Theories of human motivation - Mostly intrinsic vs. mostly extrinsic motivation factors - Cultural differences and Hofstede dimensions Current research on these topics will be reviewed and practical notes given on how to avoid common pitfalls as well as how to identify what is likely important and what is likely less so.

1. Understanding Human Player:
Attention, Perception and Motivation
Sergei Savchenko
Bethesda Game Studios

2. Sergei Savchenko
• Sr. Technical Lead at Bethesda Game Studios
• Also worked at:
• Warner Bros.
• EA
• 3DO
• Consoles, Handhelds and Mobiles
• Wrote a book on computer graphics years back

7. What does it have to do with game dev?
• Human perception is limited by attentional resources and speci
fi
cs of the
visual system
• Humans may also not be motivated enough to use these resources in the
fi
rst place…
• Creative intent and dev $$ may not be hitting the mark if the limits are
not respected

8. Working memory
• Very limited (~7 pieces of
information)
• Shared between different circuits
(e.g. visuospatial and
phonological)
• We cannot multitask well! (in a
limited manner when the tasks are
of different types)
Working
Memory
Visuospatial
Sketchpad
Phonological
Loop
Episodic
Buffer
Sensors &
Sensory memory
Longterm
memory

9. Inattentional Blindness
• Very signi
fi
cant events may be
unnoticed if attention is focused
elsewhere!
• How many features that we build are
Gorillas?
Image from Selective Attention Test, Simons and Chabris 1999

10. Chunking
• 123456789 - more than 7 digits but
easy to remember
• Several pieces of information may be
chunked into a single piece
• Prior knowledge reduces working
memory load
• But: 10 NPCs is a crowed (one piece of
information) and 12 NPCs is still a
crowd…
Working
Memory
Visuospatial
Sketchpad
Phonological
Loop
Episodic
Buffer
Sensors &
Sensory memory
Longterm
memory

11. Affordances
• Shapes or concepts with well known
function or meaning reduce load on
working memory
• These require less attention to
perceive

12. Forgetting Curve
• Longer term memory is fragile
• We only remember about 60% after
just 20 min
• Only ~30% after one day!
• Recently learned concepts do not
provide easy affordances
100
20
60
30
~20 min
1 day
Now
~30 days

13. The Gamer’s Brain
Solid overview of how player’s brain
works with the focus on UX design.
Dr. Celia Hodent

14. Image by Fibonacci, CC BY-CA 3.0

15. Image by Fibonacci, CC BY-CA 3.0

18. Image by António Miguel de Campos, Public Domain

19. Image by António Miguel de Campos, Public Domain

20. Image by Pinna, B, CC BY-SA 3.0

21. Image by Pinna, B, CC BY-SA 3.0

22. Image by Cmglee based on work by Kitaoka Akiyoshi, CC BY-SA 3.0

23. Image by Cmglee based on work by Kitaoka Akiyoshi, CC BY-SA 3.0

24. Human vision, while effective, is not accurate!

25. The Eye
• Incoming light is focused by a two
lens system of cornea and the lens
of variable curvature
• Light is projected on to the retina,
primarily focused in the foveal pit
• Fovea is depressed to minimize light
pollution and reduce light scattering
• There is a blind spot that we are
psychologically unaware of
Image by Holly Fischer CC BY 3.0

26. Vision Fidelity
• Detail vision is very narrow, central
and slow
• Peripheral vision is not sensitive to
detail but quite sensitive to motion
and change be it at course resolution
Image by Vanessa Ezekowitz BY-SA 3.0. Image by Zyxwv99 CC BY-SA 4.0

27. Cell Signals
On-
Horizontal
Cell
Bipolar
cells
On- Cells
Ganglion
cells
1.
2.
3.
4.

28. Receptive Fields
• Midget/parvocellulare pathways
(~90% of all RGCs)
• Achromatic and chromatic vision
• Slow temporal response
• Parasol/magnocellulare pathways
(~5% of all RGCs)
• Motion, change
• Fast temporal response
Midget system
Parasol system
Center illuminated

29. Oriented Edges
• Simple and complex cells are
wired to multiple retinal
receptive
fi
elds and respond to
oriented stimuli
• May respond to moving
oriented stimuli
• There may be more cells
responding to vertical and
horizontal stimuli!
Retina LGN Striate (V1)
Image by Mark Fairchild, CC BY-SA 3.0, Image by Pancrat, CC BY-SA 3.0, Image by Paul C. Foster, CC BY-SA 3.0

33. Color Vision
• We cannot see greenish reds or
yellowish blues
• Changes in greenish intensities
are easier to recognize
• Red and Blue may not be the best
for small foreground elements
• Colors are likely perceived by
much lower resolution compared
to edges
Edges
Blue/Yellow
Red/Green

34. Saccadic Movement
• Average saccade ~30ms
• functional blindness during the saccade
• Average
fi
xation (with micro-saccade drift)
~200-300ms
• Smooth pursuit movement to chase moving
targets
From Alfred Yarbus Eye Movement and Vision, 1967. Image by Lucs-Kho, Public Domain

35. Fixations
• Eye movement patterns are task
speci
fi
c
• Faces are almost always salient
(even in peripheral vision!)
• Feet position may also be salient
• Shape interiors are rarely
fi
xated
on…
From Alfred Yarbus Eye Movement and Vision, 1967. Image by Lucs-Kho, Public Domain

36. Image by Friedrich Frisch, CC BY-SA 4.0

37. Image by Friedrich Frisch, CC BY-SA 4.0

38. Lighting from Above
• Shape perception is trained on natural
lighting stimuli (hence lighting from above
perceptual bias)
• Lighting from above bias is not absolute
and easily compensated by perspective
and shadows clues

39. Image by Rob Bogaerts Image manipulation: Phonebox, Public Domain

40. Image by Rob Bogaerts Image manipulation: Phonebox, Public Domain

41. Shadows
• Shadows are generally not interpreted as
shapes
• Shadows provide strong positional clues
(sometimes stronger than perspective
clues)
• Shadows edges are nearly always softer
compared to object edges

43. Shadows and Illumination
Direction
• Inconsistencies in shadows direction/
length and illumination direction are not
always salient
• Humans in experiments may not pay
attention to inconsistencies as large as
45°

44. Shadows and Illumination
Direction
• Inconsistencies in shadows direction/
length and illumination direction are not
always salient
• Humans in experiments may not pay
attention to inconsistencies as large as
45°

45. Perception of Global Illumination
• Scenes are likely segmented in our
perception into areas where
illumination is perceived as mostly
the same
• There is evidence that people
mostly perceive in
fl
uence from one
or two light directions per object
• Diffuse to diffuse interactions (e.g.
small color bleeding) may be less
perceptually salient
Images by Barahag, CC BY-SA 4.0

47. Image by Cmglee based on work by Kitaoka Akiyoshi, CC BY-SA 3.0

48. Peripheral Drift Illusion
• Patterns with luminosity change may
be interpreted as motion in
peripheral vision likely due to eye
movement (including micro-saccades)
• Motion is perceived as proceeding
from dark to light areas
Image by Herbert and Faubert, CC BY-CA 3.0

49. Images by Mlechowicz, CC BY-CA 3.0

50. Orientation Sensitive
• Similarly to special circuitry related to
face recognition (Fusiform Gyrus) we
posses dedicated brain areas to
recognize biological movement (STS -
Superior Temporal Sulcus)
Images by Mlechowicz, CC BY-CA 3.0

51. Frame Rate
• Primates have multiple parallel vision systems
with different temporal behavior:
• Fine Detail, Color (Parvo?)
• General shape and motion (Magno?)
• Threat/
fl
ocking (Konio?)
• Day/night cycle (Konio, photosensitive
RGCs?)
• Attention shift eye movement, Pupillary
Light Re
fl
ex, Color and lightness constancy
• We know that:
• Detail vision, slow
• Rough shape vision/motion perception, fast
Midget system
Parasol system
Center illuminated

52. Flicker Fusion Rate
• Flicker fusion threshold depends on retinal
illumination level and stimuli size
• Generally fusion happens somewhere
between low 20hz and just under 60hz
• But, high intensity
fl
icker can be felt (not
necessarily perceived) even at 300+Hz
• LEDs in HDR TVs…
-3 2
-2 -1 0 1
0
10
20
30
40
50
Retinal Illumination
Log Trolands
3
60
Cycles per
seconds
Adapted from Hecht and Smith, Intermittent stimulation by light, 1936
4 5 6
19°
6°
2°
0.3°
Stimuli Size
Visual Field Deg

53. Flicker Fusion Eccentricity
• Central vision is less sensitive to
fl
icker
(slower midget RGCs?)
• Mid periphery appears the most sensitive
(faster parasol RGCs?) 0 100
20 40 60 80
20
30
40
50
60
70
Flicker
Hz
Eccentricity
Deg
Adapted from Rovamo & Raninen, Critical flicker frequency and M-scaling of stimulus size and retinal illuminance, 1984

54. Visual Perception From a Computer
Graphics Perspective
Extensive review of visual perception from
computer graphics perspective
William Thompson, Jeanine Kelly Stefanucci, Sarah Creem-Regehr, Roland Fleming

55. Theories of Motivation

56. Hierarchy of Needs (A. Maslow 1943)
Images by Androidmarsexpress, CC BY-SA 4.0

57. E.R.G. (Clayton Alderefer, 1969)
Existence
Needs
Relatedness
Needs
Growth
Needs

58. Theory of Needs (D. McClelland, 1960s)
Achievement
Needs
Power
Needs
Affiliation
Needs

59. Self Determination Theory (E. Deci and R. Ryan, 1985)
Intrinsic Extrinsic
Competence
Autonomy
Relatedness
Regulation
Introjection
Identification
Integrated
regulation
Controlling:
Autonomous:

60. In Practical Terms
• Competence: Goals clarity, feedback, accomplishments,
fl
ow
• Autonomy: Choices with consequences, openness and customizations
• Relatedness: Coopetition, visible prior achievement indicators
• External Regulation: Reward and punishment, timing of rewards
• Introjection: Impact of friends and In
fl
uencers…
• Identi
fi
cation: Game brand and company brand and values

61. In positive psychology, flow, also known as
the zone, is the mental state of operation
in which a person performing an activity is
fully immersed in a feeling of energized
focus, full involvement, and enjoyment in
the process of the activity. In essence, flow
is characterized by complete absorption in
what one does, and a resulting loss in
one's sense of space and time.
-Wikipedia

62. Flow
The psychology of optimal experiences
Dr. Mihaly Csikszentmihalyi

63. National/Cultural Differences
• JRPG vs. WRGP?
• Geert Hofstede, Dutch
researcher who worked for
IBM formulated dimensions
outlining cultural differences
• Importantly these are not
individual traits but re
fl
ective
of prevalent societal attitudes
Image by Piotrus, CC BY-SA 4.0

64. Thanks!
Sergei Savchenko
sergei.savchenko@bethesdastudios.com