Conference PaperPDF Available

Die-r Consequences: Player Experience and the Design of Failure through Respawning Mechanics

Authors:

Abstract and Figures

In games, failure that leads to death is a trope that players are all too familiar with. We were motivated by this to explore how altering the consequences of death on player progress affected aspects of the player experience. Specifically, our research investigated the relationship of death and respawn-ing mechanics-precisely the location of respawn points-to player experience (PX) constructs, such as mastery, challenge, autonomy, curiosity, and immersion. We developed a simple 2D platformer game that only differed in respawn point locations: the start of the game (permadeath), the start of a level, the last reached checkpoint, and the last manually saved point. We report findings from a study with 72 participants that indicated modifying a respawn mechanic can lead to varying effects on PX and that different mechanics may be more effective for specific types of players (challenge-and goal-oriented). We then discuss the implications for targeted game design and opportunities for further research into death and respawning mechanics.
Content may be subject to copyright.
Die-r Consequences: Player Experience and the
Design of Failure through Respawning Mechanics
Marjorie Ann M. Cuerdo, Anika Mahajan, and Edward F. Melcer
Alternative Learning Technologies and Games Lab, Computational Media Department
University of California, Santa Cruz
Santa Cruz, CA, USA
{mcuerdo, anmahaja, eddie.melcer}@ucsc.edu
Abstract—In games, failure that leads to death is a trope that
players are all too familiar with. We were motivated by this
to explore how altering the consequences of death on player
progress affected aspects of the player experience. Specifically,
our research investigated the relationship of death and respawn-
ing mechanics – precisely the location of respawn points – to
player experience (PX) constructs, such as mastery, challenge,
autonomy, curiosity, and immersion. We developed a simple 2D
platformer game that only differed in respawn point locations:
the start of the game (permadeath), the start of a level, the
last reached checkpoint, and the last manually saved point. We
report findings from a study with 72 participants that indicated
modifying a respawn mechanic can lead to varying effects on PX
and that different mechanics may be more effective for specific
types of players (challenge- and goal-oriented). We then discuss
the implications for targeted game design and opportunities for
further research into death and respawning mechanics.
Index Terms—Game design, Player experience, Failure, In-
game death, Respawning; Player traits, Platformers
I. INTRODUCTION
Death is commonly ingrained in the player experience as
a way to represent failure [1]–[4]. While there is much work
that examines the mechanics of player death in games through
the lens of difficulty – dynamic difficulty adjustment [5],
[6], challenge design [7]–[9], and challenge modeling [10]
– there are other aspects of the player experience (PX) that
have similarly crucial roles in how a game is perceived. More
specifically, games are typically evaluated in a general sense,
such as whether they are fun, hard, or have flow; however,
it is often difficult to address the direct connection of those
perceptions to specific game design mechanics. Therefore,
it can be helpful to break the player experience down into
narrower areas. Challenge is one example of a functional
aspect of PX, whereas there are also psychosocial aspects such
as meaning, mastery, immersion, autonomy, and curiosity [11].
As death is largely unavoidable when playing games, game
designers and researchers should work to better understand
how players are affected by it.
In this paper, we conducted a between-subjects study with
72 participants where we modified where players respawned
after dying in a simple 2D side-scrolling platformer game,
based on respawn point locations from the Death and Re-
birth Taxonomy [12], [13]. In respective test conditions, the
respawn points were located in: the very start of the game
("permadeath"-like), the start of a level, the last reached
checkpoint, or the last point where the player manually saved
the game state. We used the Player Experience Inventory (PXI)
[11] to examine how altering the consequences of death on
player progress affects challenge and psychosocial aspects of
PX. We also examined whether players’ affinity for challenge
and goals in games (i.e., their traits) were related to their
experience of those PX constructs.
Our results showed that there were significant differences
among the respawn point location groups. Specifically, the use
of checkpoints related to lower PX ratings for autonomy and
curiosity as opposed to using savepoints and respawn to start of
game (permadeath). Players’ final death counts also negatively
related to most of the measured PX constructs (mastery, auton-
omy, curiosity, and challenge), meaning that the more times
they died, the less they experienced those dimensions of PX.
However, immersion was positively related, meaning that the
more times players died, the more immersed they felt. Lastly,
higher scores for players’ challenge orientation traits positively
related to higher PX ratings for mastery and immersion, while
their goal orientation scores positively related to autonomy.
Our results indicate that even modifying just one respawning
mechanic in the same platformer game leads to differences in
player experiences.
II. BACKGROU ND
A. Player Death and Failure
Deadly mistakes are recoverable in video games but have
much graver final consequences outside of that virtual reality.
McAllister & Ruggill [14] argued that games use this human
understanding of "mortality salience" or death awareness as
their "deepest mechanic". Similarly, Juul [1] argued that failure
in games leads to feelings of inadequacy that players feel they
must overcome. However, failures in games are not of one
nature to players as there are also "positive failures" [15].
For example, through the lens of queer game studies, the
experience of failure could actually be seeked out instead
of avoided, due to some dissatisfaction with abiding by the
status quo [16]. Aytemiz [17] differentiated player failures
into either in-loop (expected difficulties within game loop,
such as failing to solve a puzzle) and out-of-loop (unexpected
difficulties outside of game loop, such as accessibility issues).
When players encounter in-loop failures repeatedly by dying,
the motivation for gameplay is brought to life.
978-1-6654-3886-5/21/$31.00 © 2021 IEEE
Despite death being the way of life in games, there is
surprisingly a limited amount of academic work on the subject
of its mechanics. There has been interest in death primarily
from the game studies perspectives, such as regarding its rep-
resentation and relationship to human experiences of mortality
[1]–[4], [18], [19]. However, death and respawning mechanics
themselves are lesser examined. One notable exception is
the use of "permadeath" mechanics—the permanent in-game
death of a playable character [20]—where there is interest in
studying why players are drawn to these high-risk mechanics
[20]–[23]. This genre of games have also gained commercial
popularity in recent years as demonstrated in games such as
Hades,Dark Souls,Rust, and DayZ.
More broadly, the Death and Rebirth Taxonomy [12], [13]
was developed as a tool to identify elements related to the
experience of dying and respawning in games. We used this
taxonomy to focus our study on modifying the locations of
respawn points after player death, as this was a way to vary
the consequences of death.
B. Player Experience and Failure
While many player experience surveys exist such as PENS
[24] and GEQ [25], Abeele et al. [11] argued that they tend
to focus more on overall experiences, such as enjoyment, that
are difficult to pin down to smaller game design elements.
Secondly, they offer different perspectives on what constitutes
a ’good’ player experience" [11]. The Player Experience In-
ventory (PXI) was then developed to measure both functional
and psychosocial aspects of player experience on one scale.
The breadth of concepts that PXI measures at once enables
game designers and researchers to more clearly understand
how the selection of game design elements relate to certain
player experiences. Therefore, the PXI was particularly useful
in our case, as we were interested in how specific respawning
elements affect relevant PXI constructs, such as mastery,
challenge, autonomy, curiosity, and immersion. For clarity,
these were the following definitions given for our selected
constructs: (a) mastery is a "sense of competence and skill-
fulness derived from playing the game", (b) challenge is "the
extent to which the challenges in the game match the player’s
skill level" (not the same as difficulty level), (c) autonomy
is "a sense of autonomy and freedom to play the game as
desired", (d) curiosity is "a sense of interest and curiosity the
game arouses in the player", and (e) immersion is "a sense
of absorption and immersion experienced by the player" [26].
These were relevant as they addressed players’ self-perception
of their skills and interest in the game as they face variations
of setback punishments (i.e. respawning to different locations).
Mastery-oriented individuals are often related to having bet-
ter resilience to experiences of failure, such as working harder
to find solutions to problems, as opposed to those helpless-
oriented who more easily discouraged [27], [28]. Craig Ander-
son has examined the relationship between mastery orientation
and failure-related behaviors in video games [15], [29]–[31].
Challenge and difficulty levels affect the intensity of those
in-game failures and the skill level required to manage them
[32]. As attribution theory explains that people tend to relate
events to specific causes [33], it’s crucial to examine whether
players experience autonomy to understand what factors they
attribute their failures to (e.g. their skills, the game’s design,
etc.). Lastly, games are supposed to be engaging regardless of
difficulty level. Therefore, measuring curiosity and immersion
can inform whether players feel compelled to keep playing
despite experiencing failure. These constructs are conceptually
related to the PENS definition of presence [11], which is often
measured as a crucial component of player enjoyment [24].
III. MET HO DS
A. Study Goals
The goal of this study is to use the Death and Rebirth
Taxonomy [12], [13] to explore how modifying respawn point
locations after in-game death affects player experience. We
tested four respawn point location conditions (independent
variable): (1) respawn to start of game (permadeath), (2)
respawn to start of level, (3) respawn to checkpoint, and (4)
respawn to savepoint (Fig. 1 shows these respawn groups
on a game progression timeline). This study examines the
relationship of those conditions to the player experience of
mastery, challenge, autonomy, curiosity, and immersion (de-
pendent variables). We also explored whether player death
count and challenge and goal player orientation trait scores
significantly related to those PX constructs.
We assumed that altering respawn locations would be a
simple but effective way of varying the degrees of punishment
after in-game death. Juul described this type of consequence
as a "setback punishment" [32], where the player needs to
replay parts of a game. The variation of punishment for player
failure affects the perception of a game’s difficulty and flow.
However, Juul also found that the desires of players are often
contradictory, as they simultaneously want to win (i.e. game
should be easy) and to be challenged (i.e. game should be
hard). Achieving this tricky balance in game design is a
challenge in itself; therefore, we explored the nuances of that
dynamic by examining other parts of the player experience in
addition to challenge.
We assumed that a small punishment for failure with
respawning to regularly-set checkpoints—the most popular of
respawning mechanics found in a previous study on platformer
games [13]—would immerse players the most, leading to
higher self-perception of skills (mastery) and motivation to
keep exploring the game (curiosity). In contrast, we assumed
that the most punishment for failure with the permadeath-like
respawning to the start of the game would break immersion
most, leading to lower self-perception of skills (mastery) and
motivation to keep exploring (curiosity).
As for other setback punishment variations, we assumed
that giving the player agency to save the game whenever they
wanted would afford the most autonomy, and that having the
player respawn to the start of levels was a better balance
for difficulty (checkpoints would be too easy and permadeath
would be too hard). When used in the context of games,
the conceptual differences between immersion and flow are
Fig. 1: A timeline depicting where players will respawn upon dying in the game, depending on the respawn point location
type: checkpoint, savepoint (manual), start of level, and start of game (permadeath).
still being debated, but it has been found that common
measurements observe the same phenomenon [34]; therefore,
we believe that our study also contributes to the understanding
of how failure design relates to flow.
We detail our hypotheses more specifically in the following.
Our main hypothesis was that there will be significant
differences among the four different respawn point location
conditions in regards to player experience constructs. We
further developed this notion into the following hypotheses
to be more specific in our observation:
Compared to other respawn point location conditions, players
in:
Respawn to checkpoint will experience: highest mastery
PX (H1a), immersion PX (H1b), and curiosity PX (H1c).
Respawn to start of game (permadeath) will experi-
ence: lowest mastery PX (H2a), immersion PX (H2b),
curiosity PX (H2c), autonomy PX (H2d), and challenge
PX (H2e).
Respawn to savepoint will experience: highest autonomy
PX (H3).
Respawn to start of level will experience: highest
challenge PX (H4).
Regarding player death counts:
Higher death counts will be significantly negatively
related to all measured PX constructs, meaning lower:
mastery PX (H5a), challenge PX (H5b), autonomy PX
(H5c), curiosity (H5d), and immersion (H5e).
Regarding player orientation traits:
Challenge orientation trait scores will be significantly
positively related to mastery PX (H6a), immersion PX
(H6b), and challenge PX (H6c).
Goal orientation trait scores will be significantly posi-
tively related to mastery PX (H7a), immersion PX (H7b),
autonomy PX (H7c), curiosity PX (H7d), and challenge
PX (H7e).
B. "Jumpy" Platformer Game Design
To investigate our hypotheses, we conducted a between-
subjects study. We created a simple platformer game Jumpy
in four versions for each respawn point location condition: (1)
respawn to start of game (permadeath), (2) respawn to start of
level, (3) respawn to checkpoint, and (4) respawn to savepoint.
All versions of the game employed the same mechanics and
only differed where the player respawned after they died in
the game.
We intentionally designed it so that other identified Death
and Rebirth Taxonomy components (death conditions, player
progress changes, aesthetics, and obstacles) were uniform to
isolate the potential effects of specifically modifying respawn
locations. The game had a total of five levels that increased
in difficulty. Implementing conventions from the platformer
genre, the player moves across the level from left to right
until they collide with a treasure chest, which represented the
end of a level.
As the only mechanics in Jumpy are to avoid environmental
obstacles (move left, move right, jump, collect coins, and get
hurt), we implemented a more forgiving death condition out of
health, as opposed to instant death. The player started with five
hearts and could earn up to three bonus hearts by collecting
coins. Each coin was worth 10 points and every 50 points
earned one bonus heart (max total health of eight hearts).
Any player collision with an enemy or environmental obstacle
resulted in the loss of a heart. The player died whenever they
lost all their hearts (hence out of health) or fell into the water
or spiky pits.
We also standardized player progress – both number of
hearts and points – to save only up to the last reached respawn
point location and end of a level (e.g. when beating a level, the
current number of hearts and points are saved). I.e., all player
progress from the last reached respawn point is lost upon death
(e.g. in the respawn to start of game (permadeath) condition,
all points and/or hearts earned are lost when you die before
beating the entire game). The consequences for player death
then ranged from low-risk (checkpoints and/or savepoints) to
high-risk (start of game).
Furthermore, the levels were identical across conditions
and designed with two types of environmental obstacles [12],
[13] that hurt the player on collision: (1) static enemies
and environmental objects, which stayed in place, and (2)
automated enemies, which patrolled in consistent movement
patterns. Additionally, to prevent the potential bias of aesthetic
representations of death in this study, player death (failure)
simply triggered a very short sequence of events where a glitch
sound played, the player’s character faded out, and abruptly
cut to the player being dropped to the last reached respawn
point location.
Fig. 2: A screencap of Jumpy. Player progress stats (hearts
and points) were displayed on the game UI at all times.
C. Measurements
Participants completed surveys before (pre-test) and after
(post-test) playing the game.
In the pre-test survey, data was collected for demographics
and player orientation trait scores through the Trait Model
of Game Playing Preferences [35]. Tondello et al.’s player
orientation traits included aesthetic, narrative, goal, and chal-
lenge orientation; our study focused on challenge and goal
orientation scores.
The game recorded final player progress statistics; our study
focused on the total player death count.
In the post-test survey, we used the Player Experience
Inventory (PXI) [11] to measure functional and psychosocial
PX constructs. Our study focused on mastery, immersion,
autonomy, curiosity, and challenge. Abeele et al. [11] estab-
lished the construct validity of the PXI using both exploratory
and confirmatory analysis. Items across both surveys were
measured on a 7-point Likert scale, ranging from 1-Strongly
disagree to 7-Strongly agree.
D. Participant Recruitment
Participants were recruited through university students and
social media sites (e.g. Twitter, Reddit, Facebook, Discord,
and Slack). A total of 72 participants (age ranged from 18-44
years old; broken down into 18-24 years old group (43.5%
of participants), 25-34 years old group (52.7%), and 35-44
years old group (4.2%)), completed the study fully online.
The breakdown of gender was the following: 37 participants
identified as female, 32 as male, and three as non-binary.
Daily gaming frequency habits data was also collected, with
22.22% of participants playing less than one hour daily,
34.72% playing 1-2 hours daily, 34.72% playing 3-4 hours
daily, 4.2% playing 5-6 hours daily, and 4.2% playing 7+
hours daily. All participants participated voluntarily, with only
eligible university student participants receiving class credit.
E. Procedure
When participants clicked the invitation link, they were
randomly assigned to one of the four test conditions: (1)
respawn to start of game (permadeath), (2) respawn to start
of level, (3) respawn to checkpoint, and (4) respawn to
savepoint. Participants first took the pre-test survey regarding
demographics and player orientation traits. They were then
given up to 15 minutes to play a version of Jumpy, the
platformer game. They weren’t given information as to the
death and respawning mechanics in their version. They were
simply given the game controls (moving and jumping) and
scoring rules – every 10 points earns a heart and beating the
entire game within a certain time period earned a bronze,
silver, or gold medal. If they finished the game faster than
15 minutes or didn’t finish on time, the game stopped and
their final gameplay statistics were displayed. Simple stats
were displayed such as their total completion time, total score,
total number of deaths, and earned medal. Then, they were
automatically moved to the post-test survey where they took
the PXI [11].
IV. RES ULTS
A. Statistics
We were interested in exploring how modifying the location
of respawn point types related to PX constructs such as
mastery, challenge, autonomy, curiosity, and immersion. With
an alpha = 0.05 and power = 0.80, conducting an a priori
power analysis for an ANOVA with effect size = 0.4 using
G*Power [36] resulted in the projected sample size of 68
participants with 17 participants in each test condition. Firstly,
we conducted normality tests and found that we didn’t have
normal distribution. Data transformation techniques (square
root and log10) also didn’t normalize the data. Therefore, we
decided to use the Kruskal-Wallis H test, or "one-way ANOVA
on ranks", an alternative non-parametric method to analyze our
data. Additionally, we were also interested in whether death
count and players’ orientation traits related to PX constructs.
We used Spearman’s Rank-Order Correlation to analyze those
relationships. An alpha level of 0.05 was used for all statistical
tests. See Table I for a table of statistical tests results.
B. Respawn Point Location Types and Player Experience (PX)
Constructs
Significant differences of moderate effects were found in
the medians among the respawn point location type groups
for PX constructs of autonomy (𝜒2(3, N = 72) = 9.757, p
=0.021, 𝜂2= 0.112) with mean ranks scores of 36.89 for
respawn to start of level, 42.86 for respawn to start of game,
23.86 for respawn to checkpoint, and 42.39 for respawn to
savepoint, and curiosity (𝜒2(3, N = 72) = 11.230, p = 0.011,
𝜂2= 0.134) with mean ranks scores of 35.64 for respawn to
start of level, 48.22 for respawn to start of game, 25.14 for
respawn to checkpoint, and 37.00 for respawn to savepoint.
We then conducted post-hoc tests adjusted with Bonferroni
correction to evaluate pairwise comparisons among the four
TABLE I: Player Experience Constructs in Relation to Death Counts, Respawn Types, and Player Trait Scores
Mastery Challenge Autonomy Curiosity Immersion
Total Player Death Count
Spearman’s Corr.
p = 0.037
𝑟𝑠=-0.367
Spearman’s Corr.
p = 0.027
𝑟𝑠=-0.261
Spearman’s Corr.
p = 0.023
𝑟𝑠=-0.268
Spearman’s Corr.
p = 0.032
𝑟𝑠=-0.253
Spearman’s Corr.
p = 0.017
𝑟𝑠=0.554
Respawn Point Location Types
(Checkpoint, Savepoint, Level, Permadeath)
Kruskal-Wallis H
p = 0.212
Kruskal-Wallis H
p = 0.592
Kruskal-Wallis H
p = 0.021
𝜂2=0.112
Kruskal-Wallis H
p = 0.011
𝜂2=0.134
Kruskal-Wallis H
p = 0.642
Challenge Orientation Trait Score
Spearman’s Corr.
p = 0.048
𝑟𝑠=0.234
Spearman’s Corr.
p = 0.542
𝑟𝑠=0.073
Spearman’s Corr.
p = 0.120
𝑟𝑠=0.185
Spearman’s Corr.
p = 0.887
𝑟𝑠=0.017
Spearman’s Corr.
p = 0.788
𝑟𝑠=-0.032
Goal Orientation Trait Score
Spearman’s Corr.
p = 0.683
𝑟𝑠=-0.049
Spearman’s Corr.
p = 0.322
𝑟𝑠=-0.118
Spearman’s Corr.
p = 0.038
𝑟𝑠=0.245
Spearman’s Corr.
p = 0.069
𝑟𝑠=0.216
Spearman’s Corr.
p = 0.138
𝑟𝑠=0.177
groups. The results of these tests indicated the following
significant differences:
For autonomy PX, the respawn to start of game (per-
madeath) group slightly scored higher than the respawn
to checkpoint group (p = 0.037).
For autonomy PX, the respawn to savepoint group also
slightly scored higher than the respawn to checkpoint
group (p = 0.046).
For curiosity PX, the respawn to start of game group
significantly scored higher than the respawn to checkpoint
group (p = 0.005).
No significant differences were found among the respawn
point location groups for mastery, immersion, and challenge.
C. Death Counts and PX Constructs
Significant correlations of varying effects were found be-
tween player death counts and all the measured PX constructs
(mastery, immersion, autonomy, curiosity, and challenge). We
detail the respective significant results:
For mastery PX, death counts had a weak negative
correlation across all respawn point location groups (𝑟𝑠
=-0.367, p = 0.037). However, the respawn to start of
level group in particular had a strong negative correlation
with death counts (𝑟𝑠=-0.693, p = 0.001).
For challenge PX, death counts also had a weak negative
correlation across all respawn point location groups (𝑟𝑠=
-0.261, p = 0.027). However, the respawn to checkpoint
in particular had a strong negative correlation with death
counts (𝑟𝑠=-0.610, p = 0.007).
For autonomy PX, death counts had a weak negative
correlation across all respawn point location groups (𝑟𝑠
=-0.268, p = 0.023).
For curiosity PX, death counts had a weak negative
correlation across all respawn point location groups (𝑟𝑠
=-0.253, p = 0.032).
For immersion PX, death counts did not have significant
correlations overall, but did show a moderate positive
correlation for the respawn to start of game group (𝑟𝑠
=0.554, p = 0.017).
Overall, it appears that examining results for respective
respawn point location groups yielded stronger correlations,
reinforcing their differences.
D. Player Orientation Traits and PX Constructs
Examining results for respective respawn point location
groups similarly yielded stronger correlations made when an-
alyzing player orientation traits and PX constructs. Significant
correlations of varying effects were found between player
orientation traits of challenge and goal and mastery, autonomy,
and immersion. We detail the respective significant results:
For challenge orientation trait,mastery PX had a weak
positive correlation across all respawn point location
groups (𝑟𝑠=0.234, p = 0.048). However, challenge
orientation trait scores in the respawn to start of level
group in particular had a strong positive correlation with
mastery (𝑟𝑠=0.674, p = 0.002).
Regarding immersion PX, the respawn to savepoint group
in particular had a moderate positive correlation with
challenge orientation trait score (𝑟𝑠=0.534, p = 0.022).
For goal orientation trait,autonomy PX had a weak
negative correlation across all respawn point location
groups (𝑟𝑠=0.245, p = 0.038).
V. DISCUSSION
We will discuss the observed effects of modifying the
location of respawn points on each player experience (PX)
construct respectively for clarity and their implications for
game design for specific types of players.
A. Autonomy: Transparency & Goal-Oriented Players
Our assumptions for effects on autonomy were partially
supported. We hypothesized that players would experience the
most autonomy in the respawn to savepoint condition (H3) and
the least in respawn to start of game (permadeath) condition
(H2d). Recall that permadeath actually had highest autonomy
(mean rank of 42.86) with savepoint following closely after
(mean rank of 42.39). We found the high autonomy ratings
in the savepoint group intuitive, as players had free will to
save their current progress at any point in the game. This
also accompanied our findings that players scored autonomy
higher the less they died, and may have had some impact on
the permadeath group’s higher perceived autonomy since they
experienced the least deaths (mean of 25.06).
However, we initially expected the permadeath group to
experience the least autonomy, because dying in that condition
led to the greatest loss (all player progress), so we found
it interesting that respawn to checkpoint actually scored the
worst (mean rank of 23.86). We theorize that respawn to
checkpoints scored the least autonomy because players weren’t
involved in the decision as to where they respawned after dying
in the game. Though placements of respawn point locations
may seem intuitive to game designer(s), those decisions may
appear arbitrary or a frustratingly "bad" decision from the
player’s perspective. When designing the game for our study,
we selected checkpoints around stretches of the game that
could be more difficult to complete (e.g. a checkpoint right
before a long jump that required precise timing or a checkpoint
right after beating a challenging area). Regardless, the player
may not have wanted their progress to be automatically and
unexpectedly saved at a checkpoint. Despite the consequences
of death being the greatest in the permadeath condition,
players knew exactly what to expect whenever they died.
Additionally, a player’s goal orientation trait score – in-
dicating how much they enjoyed completing game goals –
showed a slight relationship to their autonomy score (H7c).
This suggests that if one were to target goal-oriented players,
the game should afford a high degree of autonomy. A potential
method to accomplish this is to pay careful attention to
relaying as much information about the consequences of death
to the player. Though our game did not explicitly state what
happens when a player dies, in the savepoint condition, the
controls for saving the game were displayed as part of the
start screen. In the permadeath condition, the consistent reset
to zero points, initial five hearts for health, and relocation to
the very start of the game was explicit and obvious. These
factors could have contributed to the experience of highest
autonomy in that condition. We suggest that more work could
examine this phenomenon deeper, as something like a specific
study that focused on the presence (and lack) of transparency
around death and respawning mechanics could have an effect
on player autonomy.
B. Mastery: Frequency of Failure and Achievements &
Challenge-Oriented Players
Though this study did not yield significant differences
among the respawn point location groups in mastery (H1a,
H2a,H7a), we did find that mastery significantly related
to death count (H5a) and challenge orientation trait scores
(H6a) —particularly in the respawn to start of level condition.
Specifically, as we expected, players scored mastery lower
the more times they died. We argue this indicates game
designers should try to minimize the occurrences of persistent
unconquerable failure if they want to maximize their players’
self-perception of mastery.
However, this experience could vary depending on players’
orientation traits. Our findings showed that in the respawn
to start of level group, players who were more challenge-
oriented scored mastery higher. Challenge-oriented players
prefer difficult challenges [35]. Therefore, this suggests that if
one wanted to target challenge-oriented players, implementing
middle-ground consequences for failure such as respawning
to start of levels could better higher mastery. We hypothesize
that designing a level that is too easy simply could backfire
for challenge-oriented players, as they may perceive their
in-game skills to be superficial due to the game’s lack of
difficulty. In our game, failing/dying in the respawn to start
of level condition meant that the player didn’t lose all of their
player progress, yet still had to live with the periodically-saved
consequences of their past actions, whether those were good
or bad performances in previous levels. Beating a level seemed
to be a fairer assessment of ability compared to the low-risk,
high-reward situation in respawn to checkpoints and savepoints
and extreme-risk, no-reward situation in permadeath.
1) Preventing Learned Helplessness in Educational Games:
Game design that affords mastery could help engage players
that are less mastery-oriented (i.e. prone to learned helpless-
ness). This knowledge can also be particularly helpful in the
case of educational game design, where failure experiences
have been found to promote learning [31]. Future work could
study how setback punishment, such as in the form of respawn
points and player progress changes, could impact the effective-
ness of educational games in different subject areas.
C. Curiosity: Revealing Less Means More
We initially expected that players would rate curiosity high-
est in the respawn to checkpoint condition (H1c) and the least
in start of game (permadeath) (H2c). However, our findings
found the reverse to be true. We assumed that players would
be overwhelmed with frustration in having to completely start
over repeatedly every time they died, leading to less curiosity
or motivation to finish the game. Analyzing our findings
showed it is apparent that designing specifically for curiosity
is based on how much is revealed to the player over time.
Players were least curious when respawning at checkpoints,
because it was more likely for them to save their progress
in a level and therefore see more of it faster. This contrasts
players who respawned to the start of game (permadeath) who
likely saw less of the levels, having to play more slowly and
carefully to avoid death, due to the higher-risk consequences
of proceeding in the level without caution. Our findings did
show that the more times a player died, the lower they scored
their curiosity and permadeath had the least amount of deaths
(mean rank of 21.19). Consequently, we suggest that game
designers pay closer attention to what is revealed over time to
their players to maximize a sense of curiosity in their game.
D. Challenge: Death Counts Affect Perception of Difficulty
To reiterate, a higher challenge score in the PXI [11] meant
players perceived the game’s difficulty to be appropriate (i.e.
match their perceived skill level), not that they perceived
the game to be the most difficult. We expected respawn
point location groups to demonstrate differences in regards
to challenge PX scores (H2e,H4,H6c,H7e) but did not find
any. However, higher death counts did significantly relate to
challenge PX scores (H5b), meaning that the more times a
player died, the more they felt that that the game did not
have an appropriate difficulty level (i.e. unbalanced difficulty).
This observation was particularly strongest in the respawn
to checkpoint condition, which makes sense as players in it
had the highest death counts (mean rank of 50.78, compared
to 21.19 in permadeath). Additionally, Arias & Larsson [37]
previously found that players were more accepting of difficult
gameplay when they felt that they had more influence in the
game. The perceived high level of difficulty would then be
justified by a greater sense of autonomy in the eyes of the
player. As previously discussed, the checkpoints group also
experienced the least autonomy, which may have affected their
perception of the game’s challenge level.
Another possible factor is that similar to autonomy, the
presence (or lack of) explicit internal or external information
about the game’s difficulty affects the perception of it. A game
like Jumpy did not state its intended challenge level (e.g.
difficulty selection screen) nor had reviews of it online that
players had as a point of reference. Our findings do indicate
that if game designers were to create an intentionally difficult
game, they may need to be intentional with their failure design
choices as implementing checkpoints may make the game feel
unbalanced. Overall, we do not necessarily want to state that
respawn point locations have absolutely no effect on perceived
challenge. Rather, we call for more research to be done on
these nuances and how other aspects of PX and game design
(i.e. beyond simply altering one mechanic, such as balanc-
ing player progress changes or differing representations of
death/failure) influence the perception of difficulty in games.
E. Immersion: Raise the Stakes of Player Actions
Counter to our initial expectations, no significant differ-
ences were found among the respawn point location groups
for immersion. We assumed that immersion would be most
present for players that respawned to checkpoints because it
would afford a more continuous experience for players (H1b),
and least present for players that respawned to the start of
game (permadeath) because it would afford a more disjointed
experience (H2b).
However, it was within the respawn to start of game
(permadeath) group where we observed that the more times a
player died, the more immersed they actually felt. Instead of
creating a disjointed player experience, it appears that players
were more immersed playing the high stakes extreme-risk,
no-reward game version. These findings were supported by
the literature surrounding permadeath. Copcic et al. [20]
summarized the shared sentiment of permadeath scholars that
the finality of dying in games gives more excitement and
meaning to in-game death and player actions. Interestingly,
we also found that the more challenge-oriented a player
was, the higher they scored immersion (H6b). Within this
context, it is more obvious to see the connection to the rising
popularity of permadeath mechanics in roguelikes, RPGs,
and other game genres. Scholars have argued that dying can
lead to greater player satisfaction/enjoyment [22], [38]–[40].
The sunk cost fallacy [41] could also be relevant here, as
players want to see some reward worthy of the time they
spent playing the game. With the permadeath condition, they
lose the reward (e.g. satisfaction of beating the game) every
time they die, so it could affect their engagement (immersion)
to persist past failure. Therefore, our findings indicate that
if game designers wanted their challenge-oriented players
to experience a higher degree of immersion, they should
raise the stakes in gameplay to afford more active zen focus
compared to passive/casual attitudes.
Overall, our findings clearly indicate that modifying
the location of respawn points can affect respective aspects
of the player experience, as opposed to simply measuring
whether a game is fun, hard, or has flow. We believe this
calls for more research into how game design elements –
especially relating to functional/systematic death mechanics
– can be more intentionally used to create specific player
experiences tailored to particular types of players.
F. Limitations and Future Work
When designing the study, we anticipated facing difficulties
with participant recruitment during the coronavirus pandemic,
as we depended on remote online participation that required
at least 20 minutes of a volunteer’s time. It is possible that our
results could’ve trended towards more significant results with
more participants. Additionally, quantitative data from surveys
is useful but still only one type of tool to tell the story of
player experience. A mixed-methods approach incorporating
qualitative methods – such as obtaining live player reactions
to failure and/or recorded in-game behaviors – would be useful
to accompany survey data, and this is work that we hope
continues in the future.
Regardless, our findings show the relevance of studying
the relationship of death and respawning mechanics with
functional and psychosocial aspects of player experience. We
hope that our work can be used to motivate other games
researchers and designers to experiment with designing player
failure, as well as provide a starting point for further studies on
the experience of failure (or lack thereof) in other game genres
(e.g. RPGs, narrative-based games), modes (e.g. cooperative
versus competitive multiplayer), and platforms (e.g. console,
mobile, VR/AR).
VI. CONCLUSION
In this paper, we explored the effects of modifying the
location of respawn points in a platformer game on the
player experience (PX). Upon dying in the game, players were
respawned to one of the following locations: start of the game
(permadeath), start of the level, checkpoint, and savepoint.
Altering those conditions were tested for their effects on PX
constructs, such as mastery, challenge, autonomy, curiosity,
and immersion. We also studied the relationship of player
death counts and player orientation traits – challenge and goal
– with those PX constructs.
We found that there were significant differences among
the respawn point location groups. Players who respawned to
checkpoints typically experienced less autonomy and curiosity
compared to players that respawned to start of game (per-
madeath) and those who respawned to savepoint. Player death
counts also had significant relationships with all measured PX
constructs. Additionally, players’ challenge orientation trait
scores related to their experience of mastery and immersion,
whereas their goal orientation trait scores related to their
experience of autonomy. These findings suggest that mod-
ifying death and respawning mechanics has the ability to
affect respective aspects of the player experience. Our findings
indicate that more work can be done to further explore how to
tailor experiences of failure towards specific types of players
in various contexts such as entertainment and/or education
(serious games).
REFERENCES
[1] J. Juul, The art of failure: An essay on the pain of playing video games.
MIT press, 2013.
[2] L. Klastrup, “Death matters: understanding gameworld experiences,
in Proceedings of the 2006 ACM SIGCHI international conference on
Advances in computer entertainment technology, 2006, pp. 29–es.
[3] S. Mukherjee, “’remembering how you died’: Memory, death and
temporality in videogames,” 2009.
[4] E. Flynn-Jones, “Death smiles: A study of the function, nature and player
experience of in-game death,” Unpublished PhD Thesis: University of
Wales, 2013.
[5] R. Hunicke, “The case for dynamic difficulty adjustment in games,
in Proceedings of the 2005 ACM SIGCHI International Conference on
Advances in computer entertainment technology, 2005, pp. 429–433.
[6] A. Denisova and P. Cairns, “Adaptation in digital games: the effect
of challenge adjustment on player performance and experience,” in
Proceedings of the 2015 Annual Symposium on Computer-Human In-
teraction in Play, 2015, pp. 97–101.
[7] M. Brandse, “The shape of challenge,” in International Conference of
Design, User Experience, and Usability. Springer, 2017, pp. 362–376.
[8] A. Sarkar and S. Cooper, “Transforming game difficulty curves using
function composition,” in Proceedings of the 2019 CHI Conference on
Human Factors in Computing Systems, 2019, pp. 1–7.
[9] R. R. Wehbe, E. D. Mekler, M. Schaekermann, E. Lank, and L. E.
Nacke, “Testing incremental difficulty design in platformer games,”
in Proceedings of the 2017 CHI Conference on Human Factors in
Computing Systems, 2017, pp. 5109–5113.
[10] N. Sorenson, P. Pasquier, and S. DiPaola, “A generic approach to
challenge modeling for the procedural creation of video game levels,
IEEE Transactions on Computational Intelligence and AI in Games,
vol. 3, no. 3, pp. 229–244, 2011.
[11] V. V. Abeele, K. Spiel, L. Nacke, D. Johnson, and K. Gerling, “Devel-
opment and validation of the player experience inventory: A scale to
measure player experiences at the level of functional and psychosocial
consequences,” International Journal of Human-Computer Studies, vol.
135, p. 102370, 2020.
[12] M. A. M. Cuerdo and E. F. Melcer, “I’ll be back": A taxonomy of
death and rebirth in platformer video games,” in Extended Abstracts of
the 2020 CHI Conference on Human Factors in Computing Systems,
2020, pp. 1–13.
[13] E. F. Melcer and M. A. M. Cuerdo, “Death and rebirth in plat-
former games,” in Game User Experience And Player-Centered Design.
Springer, 2020, pp. 265–293.
[14] K. S. McAllister and J. E. Ruggill, “Playing to death.” American journal
of play, vol. 11, no. 1, pp. 85–103, 2018.
[15] C. G. Anderson, “Hits, quits, and retries-player response to failure
in a challenging video game,” in International Conference on the
Foundations of Digital Games, 2020, pp. 1–7.
[16] B. Ruberg and A. Shaw, Queer game studies. U of Minnesota Press,
2017.
[17] B. Aytemiz and A. M. Smith, “A diagnostic taxonomy of failure in
videogames,” in International Conference on the Foundations of Digital
Games, 2020, pp. 1–11.
[18] D. Melnic and V. Melnic, “Saved games and respawn timers: The
dilemma of representing death in video games,” University of Bucharest
Review: Literary and Cultural Studies Series, vol. 7, no. 07, p. 2018,
2018.
[19] M. Neto, “A hero’s death: Human mortality and video games,”
University of California, Santa Cruz.[sd]. https://www. academia.
edu/10622358/A_Heros_Death_Human_Morta lity_and_ Video_Games.
[20] A. Copcic, S. McKenzie, and M. Hobbs, “Permadeath: A review of
literature,” in 2013 IEEE International Games Innovation Conference
(IGIC). IEEE, 2013, pp. 40–47.
[21] M. Carter, M. Gibbs, and G. Wadley, “Death and dying in dayz,”
in Proceedings of The 9th Australasian Conference on Interactive
Entertainment: Matters of Life and Death, 2013, pp. 1–6.
[22] S. Petralito, F. Brühlmann, G. Iten, E. D. Mekler, and K. Opwis, “A good
reason to die: how avatar death and high challenges enable positive
experiences,” in Proceedings of the 2017 CHI Conference on Human
Factors in Computing Systems, 2017, pp. 5087–5097.
[23] F. Allison, M. Carter, and M. Gibbs, “Good frustrations: the paradoxical
pleasure of fearing death in dayz,” in Proceedings of the Annual
Meeting of the Australian Special Interest Group for Computer Human
Interaction, 2015, pp. 119–123.
[24] R. M. Ryan, C. S. Rigby, and A. Przybylski, “The motivational pull
of video games: A self-determination theory approach,” Motivation and
emotion, vol. 30, no. 4, pp. 344–360, 2006.
[25] W. A. IJsselsteijn, Y. A. de Kort, and K. Poels, “The game experience
questionnaire,” Eindhoven: Technische Universiteit Eindhoven, vol. 46,
no. 1, 2013.
[26] V. Vanden Abeele, L. E. Nacke, E. D. Mekler, and D. Johnson,
“Design and preliminary validation of the player experience inventory,
in Proceedings of the 2016 Annual Symposium on Computer-Human
Interaction in Play Companion Extended Abstracts, 2016, pp. 335–341.
[27] C. S. Dweck and N. D. Reppucci, “Learned helplessness and rein-
forcement responsibility in children.” Journal of personality and social
psychology, vol. 25, no. 1, p. 109, 1973.
[28] C. S. Dweck and E. L. Leggett, “A social-cognitive approach to
motivation and personality.” Psychological review, vol. 95, no. 2, p.
256, 1988.
[29] C. G. Anderson, K. Campbell, and C. Steinkuehler, “Building per-
sistence through failure: the role of challenge in video games,” in
Proceedings of the 14th International Conference on the Foundations
of Digital Games, 2019, pp. 1–6.
[30] C. Anderson, “Game on: Mastery orientation through the lens of a
challenging video game.” in CogSci, 2020.
[31] C. G. Anderson, J. Dalsen, V. Kumar, M. Berland, and C. Steinkuehler,
“Failing up: How failure in a game environment promotes learning
through discourse,” Thinking Skills and Creativity, vol. 30, pp. 135–144,
2018.
[32] J. Juul, “Fear of failing? the many meanings of difficulty in video
games,” The video game theory reader, vol. 2, no. 237-252, 2009.
[33] H. H. Kelley and J. L. Michela, “Attribution theory and research,” Annual
review of psychology, vol. 31, no. 1, pp. 457–501, 1980.
[34] L. Michailidis, E. Balaguer-Ballester, and X. He, “Flow and immersion
in video games: The aftermath of a conceptual challenge,” Frontiers in
Psychology, vol. 9, p. 1682, 2018.
[35] G. F. Tondello, K. Arrambide, G. Ribeiro, A. J.-l. Cen, and L. E.
Nacke, ““i don’t fit into a single type”: A trait model and scale of
game playing preferences,” in IFIP Conference on Human-Computer
Interaction. Springer, 2019, pp. 375–395.
[36] F. Faul, E. Erdfelder, A. Buchner, and A.-G. Lang, “Statistical power
analyses using g* power 3.1: Tests for correlation and regression
analyses,” Behavior research methods, vol. 41, no. 4, pp. 1149–1160,
2009.
[37] J. Larsson and A. Amigo Arias, “Fun with death and failure: An
exploration of player experiences in a decentralized open world rpg,
2014.
[38] W. Van den Hoogen, K. Poels, W. IJsselsteijn, and Y. De Kort, “Between
challenge and defeat: Repeated player-death and game enjoyment,
Media Psychology, vol. 15, no. 4, pp. 443–459, 2012.
[39] M. Carter and F. Allison, “Fear, loss and meaningful play: Permadeath in
dayz,” Journal of Gaming & Virtual Worlds, vol. 9, no. 2, pp. 143–158,
2017.
[40] M. J. B. Wilson, “Playing with permadeath,” in 2020 IEEE Games,
Multimedia, Animation and Multiple Realities Conference (GMAX).
IEEE, 2020, pp. 1–5.
[41] D. Friedman, K. Pommerenke, R. Lukose, G. Milam, and B. A. Hu-
berman, “Searching for the sunk cost fallacy,Experimental Economics,
vol. 10, no. 1, pp. 79–104, 2007.
... As an example, one participant said, "It was really defeating to get a bad ending and being sent back to the beginning instead of just looping to a separate point or having a save state." Considering that this was also commonly perceived as an impediment to participants' autonomy, we believe that implementing a back button or save state feature in future iterations of the game would likely considerably improve enjoyment of the game-as is successfully employed in other genres such as platformers Cuerdo and Melcer (2020), Melcer and Cuerdo, (2020), Cuerdo et al. (2021). However, care should be taken in this implementation since it has been shown in other game-based learning contexts that players are more likely to think carefully about how to solve a problem when there are greater consequences for failure (e.g., the amount of time required to make more attempts) Mann et al. (2009), Melcer and, Villareale et al. (2020). ...
Article
Full-text available
Choice-based interactive storytelling games such as Academical, our responsible conduct of research training game, show great promise as a novel way of providing efficacious ethics training. However, much work remains to determine what factors of such games contribute to their advantages over traditional text-based training tools, especially if we hope to further improve their enjoyment, engagement and efficacy. In this article, we present a case study exploring how the motivational factors of Self-Determination Theory (SDT) underlie players’ perceived most and least enjoyable experiences arising from the design of Academical. Specifically, we discuss how certain elements of Academical’s design influence different SDT factors and subsequently player experience, as well as how such elements can be changed to further improve the game. Furthermore, our work highlights potential limitations of existing conceptualizations for the relatedness factor of SDT—discussing ways that it can be extended to properly understand player enjoyment within single-player educational interactive narrative games.
Chapter
Full-text available
Failure is a central aspect of almost every game experience, driving player perceptions of difficulty and impacting core game user experience concepts such as flow. At the heart of failure in many game genres is player death. While techniques such as dynamic difficulty adjustment have addressed tweaking game parameters to control the frequency of player death occurrence, there is a surprisingly limited amount of research examining how games handle what happens when a player actually dies. We posit that this is a rich, underexplored space with significant implications for player experience and related techniques. This chapter presents our exploration into the space of player death and rebirth through the creation of a generalized taxonomy of death in platformer games—one of the genres that features player death and respawning most heavily. In order to create this taxonomy, we collected and catalogued examples of death and respawning mechanics from 62 recent platformer games released on the digital distribution platform Steam after January 2018. Games selected varied equally across positive, mixed, and negative overall reviews in order to provide a broader range of mechanics, both good and bad. We observed gameplays of each individual game and noted the processes of death and rebirth, respectively. A grounded theory approach was then employed to develop the taxonomy of game death and respawning, resulting in five notable dimensions: (1) obstacles, (2) death conditions, (3) aesthetics, (4) changes to player progress, and (5) respawn locations. Finally, we discuss how the different dimensions and mechanics highlighted in our taxonomy have implications for key aspects of player experience, as well as how they could be used to improve the effectiveness of related techniques such as dynamic difficulty adjustment.
Conference Paper
Full-text available
Failure, often represented through death, is a central aspect of almost every video game. In-game death can drive player perceptions of difficulty and greatly impact the core player experience; however, there is surprisingly limited amounts of research examining how games actually handle this occurrence. We posit that this is a rich, underexplored space with significant implications for player experience and the design of many games. This paper presents our initial exploration into the space of player death and rebirth through the creation of a generalized taxonomy developed from 62 recent platformer games. Our taxonomy consists of five key dimensions: 1) obstacles, 2) death conditions, 3) aesthetics, 4) changes to player progress, and 5) respawn locations.
Article
Full-text available
Games User Research (GUR) focuses on measuring, analysing and understanding player experiences to optimise game designs. Hence, GUR experts aim to understand how specific game design choices are experienced by players, and how these lead to specific emotional responses. An instrument, providing such actionable insight into player experience, specifically designed by and for GUR was thus far lacking. To address this gap, the Player Experience Inventory (PXI) was developed, drawing on Means-End theory and measuring player experience both at the level of Functional Consequences, (i.e., the immediate experiences as a direct result of game design choices, such as audiovisual appeal or ease-of-control) and at the level of Psychosocial Consequences, (i.e., the second-order emotional experiences, such as immersion or mastery). Initial construct and item development was conducted in two iterations with 64 GUR experts. Next, the scale was validated and evaluated over five studies and populations, totalling 529 participants. Results support the theorized structure of the scale and provide evidence for both discriminant and convergent validity. Results also show that the scale performs well over different sample sizes and studies, supporting configural invariance. Hence, the PXI provides a reliable and theoretically sound tool for researchers to measure player experience and investigate how game design choices are linked to emotional responses.
Conference Paper
Full-text available
Some video games create challenging environments that presume that players will often fail as part of developing the skills and knowledge they need to progress through the game. Psychology has a rich history of studying the ways that individuals react towards failure, including mastery orientation. Mastery oriented individuals are characterized by positive reactions to failure, such as renewed effort, heightened affect, and positive, affirming language. However, the challenging environments of which these video games are comprised have not yet been investigated fully in their potential impact on how players react to failure. In an initial exploration, we surveyed 928 undergraduates at a major university in the Pacific United States on their gameplay experience and attitudes towards challenge and failure. Analyses partially replicate previous findings and show that players who are more attracted to challenge in video games report higher mastery orientation. While this suggests that challenge in video games has some association with positive reactions to failure, further analyses are warranted to understand how challenge in video games influence player reactions and the impact this may have beyond of the game.
Conference Paper
Full-text available
Player typology models classify different player motivations and behaviours. These models are necessary to design personalized games or to target specific audiences. However, many models lack validation and standard measurement instruments. Additionally, they rely on type theories, which split players into separate categories. Yet, personality research has lately favoured trait theories, which recognize that people's preferences are composed of a sum of different characteristics. Given these shortcomings of existing models, we developed a player traits model built on a detailed review and synthesis of the extant literature, which introduces five player traits: aesthetic orientation, narrative orientation, goal orientation, social orientation, and challenge orientation. Furthermore, we created and validated a 25-item measurement scale for the five player traits. This scale outputs a player profile, which describes participants' preferences for different game elements and game playing styles. Finally, we demonstrate that this is the first validated player preferences model and how it serves as an actionable tool for personalized game design.
Chapter
Player typology models classify different player motivations and behaviours. These models are necessary to design personalized games or to target specific audiences. However, many models lack validation and standard measurement instruments. Additionally, they rely on type theories, which split players into separate categories. Yet, personality research has lately favoured trait theories, which recognize that people’s preferences are composed of a sum of different characteristics. Given these shortcomings of existing models, we developed a player traits model built on a detailed review and synthesis of the extant literature, which introduces five player traits: aesthetic orientation, narrative orientation, goal orientation, social orientation, and challenge orientation. Furthermore, we created and validated a 25-item measurement scale for the five player traits. This scale outputs a player profile, which describes participants’ preferences for different game elements and game playing styles. Finally, we demonstrate that this is the first validated player preferences model and how it serves as an actionable tool for personalized game design.
Conference Paper
Player engagement within a game is often influenced by its difficulty curve: the pace at which in-game challenges become harder. Thus, finding an optimal difficulty curve is important. In this paper, we present a flexible and formal approach to transforming game difficulty curves by leveraging function composition. This allows us to describe changes to difficulty curves, such as making them "smoother", in a more precise way. In an experiment with 400 players, we used function composition to modify the existing difficulty curve of the puzzle game Paradox to generate new curves. We found that transforming difficulty curves in this way impacted player engagement, including the number of levels completed and the estimated skill needed to complete those levels, as well as perceived competence. Further, we found some transformed curves dominated others with respect to engagement, indicating that different design goals can be traded-off by considering a subset of curves.