The researchers elected to create their own 3D game to minimize
the effects of player experience while maintaining the basic experience of common modern games. The researchers chose not
to modify an existing game because there may have been
inconsistencies between players who have experienced the game
previously. This is especially important when factoring the
addition of new game elements, which may have been too
prominent for players familiar with the unmodified game
environment. A custom game also offers fine control over player
experience and the ability to provide all research participants with
a relatively new experience.
The researchers elected to create a car racing game for the study.
The car racing game type offers a few distinct benefits from other
common game experiences. In car racing games, the game goal is
clear to a very large audience – be the first to cross the finish line.
The game rules are equally clear, as steering a car is a commonly
practiced action. The game experience can be well controlled, as
all players follow the same basic path, unlike non-linear game
experiences. As a combination of simulation and sport, car racing
games also commonly have some in game advertising in them.
Racing games are also designed to be played by a wide
demographic, containing little objectionable content and affording
for a wide range of play abilities.
The researchers conducted a fundamental survey of standard
digital game types and found that other common play experiences
offered challenges that were likely be a detriment to the study.
According to the NPD Groups 2010 study of bestselling video
games, racing games rank 5th of the 13 ranked categories [10].
The researchers found the top two ranking categories, sports
games and family entertainment, were too broad for this study.
Sports games in particular required substantial player investment
in learning rules and controls. Family entertainment experiences
were inappropriate for study simply because few of these games
employ any in game advertising, as much of the in game
advertising is targeted at more specific audiences. The third and
fourth ranking categories, shooters and action games, offered less
clear goals, more complex controls and were subject to
inconsistent player experiences.
For the experiment, the researchers constructed a 3D race car
game, figure 1, using the Unity3D game environment. The game
employs conventional physics and environment cues common to
an average racing game. The game includes one player controlled
red car and two competitor silver cars, all of the same make and
model. Players control their car with the computer keyboard’s
arrow keys.
The researchers also elected to create custom advertising, to
control for the variety of previous experience study participants
may have had with specific brands or products. The researchers
chose to advertise common consumer product goods that players
were likely to understand quickly. The products were a laundry
detergent, toothpaste, soda - commonly used by a wide array of
consumers. The products were advertised via in-game billboards.
billboard was designed against the same basic template. They each
contained a single image of the product and a simple slogan that
incorporates the product’s use. Each billboard was designed to
use a single, distinct accent color drawn from the product’s image.
The product billboards are demonstrated in figure 2.
The researchers expected billboard location to effect retention. As
a control, three tracks were constructed. Billboard location was
rotated for each of the constructed tracks. The billboard that was
first in track 1, was second in track 2, and third in track 3 for
example. Other than changes to billboard order, all other track
properties remainded the same.
By rotating billboard location the researchers hoped to reveal
patterns in retention based on exposure time and visibility. Each
of the billboards was placed on the right-hand side of the track.
Each billboard was a uniform size. The track was also lined with
a bordering wall to prevent players from falling too far off course.
No other signs or indicators were provided in the game
environment.
The tracks were shaped as a fairly simple indented elipse
illustrated in figure 3. The tracks was designed to offer a
comparable experience to a beginning level track in a commercial
game. The environment was completely flat, offering no major
topography, as it was expected that billboard visibility could be
effected by rate of travel over hills.
23 volunteers were recruited for the study. Participants were asked
to complete a pre and post survey outlining 11 factors about their
experience with video gaming and 8 factors about their experience
in the game. Once each participant completed the pre-game
survey, they were asked to play the study game for 3 minutes.
Participants were only asked to play the game, they were never
asked to pay attention to the in game advertising.
During the play session, the researchers recorded the number of
times players crashed and the number of laps they completed.
Crashes were defined as any moment the player clearly lost
control of the car and sharply diminished speed by hitting other
cars or parts of the race course. After 3 minutes, gameplay was
stopped and the players were asked to complete the post-game
survey.
The research was conducted on 3 Windows XP machines running
with 20” flat panel monitors. Players were provided with a pair
of JVC over-ear headphones to hear game audio. The study was
executed in a focus group lab designed for evaluating user
experience. The participants played individually and were not
allowed to watch other players or preview the game before they
played. Each game playing station occurred in a single person
kiosk, preventing other participants from witnessing other play
sessions. Participants were equally distributed between each of
the 3 available tracks.
14 Male participants and 9 female participants were recorded.
They ranged in age from 19 to 46 with a mean age of 22.4. 74%
owned any type of gaming console, and 43% owned a portable
gaming console. 96% of the participants played a video or
computer game within the last month. 26% played games
everyday, and another 26% played at least 2-3 times a week. In
general the participant group was fairly comfortable with digital
games and all participants were used to using a computer on a
daily basis.
The participant group was also fairly proportionate to game
players in the United States. Our participant ratio of 61% male,
39% female is nearly identical to the national ratio of 60%
females, 40% female [11] game players. With 74% of the
participants owning gaming consoles, the study group was slightly higher than the national average of 67% [11] for console
ownership.
The researchers recorded 2 primary gameplay characteristics, laps
completed and player crashes, to investigate the relationship
between player in-game performance and their retention of in
game advertising. The game was designed to allow an
advertisement to appear every 10 seconds of gameplay, requiring
a player to complete a lap in about 30 seconds to see each of the 3
advertisements. In the experiment’s 3 minute play session, the
mean number of laps completed by participants was 6.0. 73.8%
of participants completed between 5-7 laps. One participant was
able to complete only one lap. One participant completed 10 laps,
which was the most laps completed by any participant.
The game was designed to minimize crashes by providing easy
turns and competing cars that held their racing line to provide
predictable behavior. Crashes were dictated by that game’s rigid
body physics system. In general crashes were initiated when a
player car created a head-on collision with a non-player car or
wall in the game environment. The mean number of crashes
experienced was 4.39, with 73.8% of participants experiencing
between 2-5 crashes. One participant only crashed once. Two
participants crashed nine times, providing the most number of
crashes experienced by any participants.
An important goal of this experiment was to evaluate variables
that might influence the quality of in-game advertising and
branding. To do this, the authors noted the total number of brands
that each participant was able to recall. There was a negative
correlation (r = -.33) between the number of crashes experienced
by each participant and the number of brands recalled. There was
a positive correlation(r = .36) between the number of laps
completed and the number of brands recalled.
The researchers’ findings suggest that poor performance, as
measured by the number of crashes, prevents retention of in game
advertising. From the data, it seems poor performance may have
distracted participants from focusing on the in-game
advertisements. Conversely, the findings also suggest that
improved performance, as measured by the number of laps,
increase retention of in game advertising. This is perhaps because
more laps, meant increased opportunity to notice the in-game
advertisements.
As evidence, one female participant, who did not notice any of the
billboard advertisements, crashed nine times. This participant tied
another participant for the most crashes and completed five laps,
which was less than the average participant. This participant plays
PC or video games rarely, about once a month. Referring to the
billboards, she wrote in the post-game survey, "I realized they
were there but didn't feel like I had time to look at them. They
weren't in my direct line of vision."
Similarly, another female participant who did not notice any of
the ads crashed five times, which was more than the average
participant, and completed five laps of the game, which was less
than the average participant. She rarely plays PC or video games,
less than once a month, but plays games on her phone 2-3 times
per week. She wrote, "I am honestly so bad at race type games
that I did not r