How to analyse therapeutic games:

The Player / Game / Therapy Model

Stephanie Mader? , Stephane Natkin, and Guillaume Levieux

C.E.D.R.I.C. / C.N.A.M.,
292 Rue Saint Martin, 75003 Paris, France
{stephanie.mader,
stephane.natkin,
guillaume.levieux}@cnam.fr
http://cedric.cnam.fr

Abstract. In this paper, we present a new model to analyse therapeutic

games. The goal of the model is to describe and analyse the relations
between the three aspects of a therapeutic game: the player, the game,
and the therapy. The model is intended to game designers. It is a tool to
improve the communication between health experts and game designers,
and to evaluate the game design coherency of therapeutic games. It also
helps to analyse existing games to discover relevant features. The model
is built with respect to existing serious game definitions and taxonomies,
medical definitions, motivation theory, and game theory. We describe
how the model was used to design le village aux oiseaux, a therapeutic
game which goal is to train people with attention disabilities. In the
last section, we present the results of analysis done with our model and
discuss the model limits.

Keywords: video games, evaluation, analysis, model, game design, se-

rious games, therapeutic

1 Introduction
Game design is the process by which a game designer creates a game, to be
encountered by a player, from which meaningful play emerges [29]. Salen and
Zimmermans definition emphasizes that a game can not be dissociated from its
player. Neither the player and its relation to the game, nor games fundamentals
like gameplay, feedbacks, or goals are stressed in existing definitions of serious
game. In this paper, we present a new analysis model for therapeutic games.
This model is intended to game designer. They can use the model to evaluate
the coherency of their design regarding the relation that bounds the three aspects
of a therapeutic game: the therapy, the game, and the player. The goal of the
model is also to facilitate the communication between game designers and health
?
Stephanie Mader gratefully acknowledges the financial support of the Swiss National
Science Foundation (SNF). This is the author version Final editor version published
here: http://link.springer.com/chapter/10.1007/978-3-642-33542-61 7
2

experts. Finally, the model can be used to analyse existing therapeutic games to
better understand why some of them are efficient while others are not.
First, we present how we built the model with respect to existing serious
game definitions and taxonomies, medical definitions, motivation theories and
game theories. Then, we present our model and explain how it was used during
the design of le village aux oiseaux. After that, we present interesting results of
analyses made with the model. Finally, we discuss our work and present the next
steps of our research.

2 Background

2.1 Serious Games

According to Sawyer and Smith, the definition of serious game has evolved [30].
Old definitions were narrow, they restricted serious games to certain game genres
(i.e. simulation-based game) and application fields (e.g. government, education).
New definitions are more general, they include all game genres and application
fields. Although these definitions are numerous [8, 7, 35, 11], we have identified
two main common aspects: a serious game has a purpose beyond entertainment,
and the main objective of this combination (serious and game) is to use the at-
tractive shapes of the game to serve the serious purpose. Many other authors also
consider that enhancing the users motivation is the main advantage of serious
games over other media [6, 2, 10, 26]. However, other aspects of serious games
are still under discussion. For our research, we use the definition of Guardiola et
al.: A serious game is a rule-based formal system with a variable and quantifi-
able outcome, where different outcomes are assigned different values, the player
exerts effort in order to influence the outcome, the player feels attached to the
outcome. But a serious game is combined with a defined real life objective.1 [11].
This definition is based upon the classical game definition of Juul [14]. Accord-
ing to Guardiola et al., the main difference between a video game and a serious
game is that the designer of a serious game try to have an impact on real life.
Thus, the real life consequences, that are optional and negotiable as stated by
Juul, become more controlled by the designer in a serious game. If this serious
purpose is what differentiate a video game and a serious game, according to ex-
isting taxonomies, the serious purpose is also what differentiate a serious game
from another one.
Sawyer and Smith propose several taxonomies for serious games [30]. Based
on market and purpose, their taxonomy split serious games into categories like
Games for Education or Game for Health. Then, for each of these categories,
Sawyer and Smith propose a specific taxonomy. The taxonomy of Games for
1
un jeu utile est un systeme dynamique formel dont le comportement, delimite par
des regles, produit des consequences variables et des effets quantifiables. Le joueur
doit avoir la sensation que ses actions influencent de facon controlee le comportement
du jeu. Il doit etre emotionnellement attache aux resultats observes. Mais le jeu utile
est associe a un objectif defini de la vie reelle
 3

Health has two parameters: function (preventative, therapeutic, assessment,

educational, informatics) and public (personal, professional practice, research
/ academia, public health). Djaouti and Alvarez built their own classification
upon Sawyer and Smiths taxonomies [1, 7]. Djaouti and Alvarez classification of
Games for Health is a four-parameter model: gameplay, purpose, function, and
public. Function and public parameters are similar to Sawyer and Smiths, except
that therapeutic becomes a subitem of the broader function care. The purpose
only contains three items: broadcast a message (which can be informative, per-
suasive, etc.), provide a training, enhance data sharing. Djaouti and Alvarezs
work provides information on game features. Their gameplay parameter makes
the difference between Ludus and Pada based on the presence or not of explicit
goals. Moreover, they work on a general serious game classification which uses
their gameplay bricks taxonomy along with game genres and themes keywords
[21] [8].
With the help of these taxonomies, we identified that therapeutic games
are serious games which are classified as Games for Health. Their therapeutic
function is what makes them different from other Games for Health. Although we
did not found a classification specific to therapeutic games, the classification of
Alvarez and Djaouti helps to differentiate therapeutic games. Such classification
is useful. First, it deepens our understanding of what therapeutic games can
achieve. Second, it helps to find similar therapeutic games. Finally, it makes
possible to delimit the scope of a therapeutic game during the design steps.
Thus, we use this classification to study therapeutic games in the next section.

2.2 Therapeutic games

McGraw-Hill Concise Dictionary of Modern Medicine2 proposes this definition

of a therapy: A general term for any form of management of a particular con-
dition; treatment intended and expected to alleviate a disease or disorder; any
technique of recovery, which may be medical, psychiatric, or psychological. The
dictionary also proposes a list of more than two hundreds different therapies.
Some of them are very general (e.g. Physical therapy), while others are very spe-
cific (e.g. Nicotine replacement therapy). But game designers are seldom medical
experts. Thus, it would not be useful for them to classify therapy from a med-
ical standpoint. However, we have to acknowledge the wide scope of possible
therapeutic protocol and define a classification useful for game designers.
We have analysed existing therapeutic games and identified that, while some
of them are directly therapeutic, others are indirectly therapeutic. For example,
the game Brick a break described in Burke et al. is a direct therapeutic games
[3, 4]. The game activity is the rehabilitation protocol. Brick a break improves
the particular condition of having a loss of motor control in upper limb after a
stroke. Equally, playing Virtual Iraq is what directly alleviate or suppress the
post traumatic stress of soldiers [27]. Also, le village aux oiseaux aims to reduce
2

c 2002 by The McGraw-Hill Companies, Inc.
4

the effects due to the disease Alzheimer by training the patients attentional
network.
On contrary, indirect therapies are in themselves not sufficient to improve the
particular conditions of a patient. Instead, they improve the patient behaviour,
mood, or observance. For example, RE-Mission is a game for patients having a
cancer [15]. The game provides knowledge on the disease, its treatment, as well
as the means to alleviate the adverse effects. As a result, the patients become
more observant with their therapy protocol (i.e. they take more regularly their
oral chemotherapy, along with drugs to alleviate the adverse effects). This change
in the patients behaviour make the main therapy more efficient.
Other indirect therapeutic games do not target the patient, but the care
giver work, the scientific research or the patients relatives. For instance, Snow-
World is a virtual reality game that try to overload the attentional network
of highly-burned patients. The objective is to alleviate the patients pain dur-
ing the bandage change in order to facilitates the care givers work [12]. FoldIt
and Phylo are games that help the scientific research to advance on therapeu-
tics by using the computing power of mankind [17] [16]. These games aims the
gathering of knowledge that could improve the development of therapy. Finally,
other therapeutic games target the patients relatives. For instance, Elude helps
the patients relatives to understand what it means to be depressive [31]. These
games have an indirect therapeutic effect by engaging the relatives in the recov-
ery process.
We have classified these therapeutic games according with Alvarez and Djaoutis
public and purpose parameters.

Name Public Purpose

Brick a break Patient / Direct Provide a training
Virtual Iraq Patient / Direct Provide a training
Le Village aux Oiseaux Patient / Direct Provide a training
RE-Mission Patient / Indirect Broadcast a message
FoldIt / Phylo Everyone / Indirect Improve data sharing
Elude Patients relatives / Indirect Broadcast a message

Fig. 1. Classification of some therapeutic games

These different therapeutic uses of games do not require the same design
methods. For example, games like RE-Mission or Elude are, from a game design
standpoint, similar to Games for Education. They can be designed with existing
methods, see [24] or [34] for examples. Also, FoldIt and Phylo are defined as
Games for Research, they require a specific design process which is different from,
for example, the game design process of le village aux oiseaux or Virtual Iraq.
Also, we did not classify SnowWorld. The game has a direct therapeutic effect
(i.e. it alleviates temporary the patients pain) and an indirect therapeutic effect
(i.e. it facilitates the care givers work). But, none of these effects are precisely
 5

described with the existing items of Djaouti and Alvarezs purpose parameter.
As our objective is to improve the design efficiency, we prefer to treat games like
SnowWorld as special case, instead of describing them imprecisely. In fact, the
most relevant purpose item is the broadcast of a persuasive message. But the
design of persuasive game seems not to be the same as the design required to
create a game like SnowWorld. The analyses of these therapeutic games reveal
that they require very different design process. Thus, we decided to limit the
scope of the first version of our model to therapeutic games which provide a
training that improve directly the patient condition.
Most of these therapeutic games have been evaluated and have a proven
medical efficiency. The fact they are played could indicate that they are good
games, but they can also be played only for their therapeutic efficiency (i.e.
without pleasure). As we advocate that therapeutic games have to be good
games, we study why games are motivating in the next section.

2.3 Game
As motivation is the most expected effect of a game in the context of a therapy,
the model emphasizes motivational aspects of games. First, we compare play
and therapeutic activities regarding motivation theory. Then, we examine game
theory with regards to their relation to the player motivation.
Ryan and Deci classify motivation quality as follows: amotivation (i.e. when
there is no motivation), extrinsic motivation (i.e. when the activity is done to
obtain a separable outcome), and intrinsic motivation (i.e. when the activity is
done for the activity itself) [28]. They split extrinsic motivation into four sub-
categories, depending on the level of internalization (i.e. the process of taking
in a value or a regulation) and integration (i.e. transformation of the value or
regulation to fully integrate the sense of self) ones made of the activity. Also,
Ryan and Deci stated that intrinsic motivation exists in the nexus between a
person and a task .
However, Csikszentmihalyis flow theory explains that some activities can
more easily elicit intrinsic motivation [5]. Because games are designed to be
pleasurable and interesting activities, games are very powerful to create and
foster motivation of their players. Generally, games are played for the sake of
playing them. On the contrary, therapeutic activities are designed to maximise
their efficiency. Patients engage in therapy for an identified separable outcome:
health recovery. Ryan and Deci describe a similar issue in education: Given
that many of the educational activities prescribed in schools are not designed to
be intrinsically interesting, a central question concerns how to motivate students
to value and self-regulate such activities, and without external pressure, to carry
them out on their own. But, according to Ryan and Deci, an activity started
for a separable outcome can become a self-determined activity if the person finds
the activity sufficiently interesting. We emphasize this and recommend to make
a therapeutic game as interesting and fun as possible to smooth out the medical
aspect. Doing so should maximise the patients motivation. As a result, we have
now to understand which game features enhance the player motivation.
6

There exists numerous models of what makes a good video game [23, 19,
33, 18, 25, 6]. These models encompass many aspect of video game enjoyment:
for instance having an adapted level of challenge, as well as fostering social
connectedness or having a high level of fantasy. In this paper, we propose a model
for game designers, and moreover, designers of therapeutic games. We will thus
extract the very common and core features of these models, and especially the
features related to the very basis of the game designers work.
Two features seems to be present in many models and are fundamental to
the game designers work: challenge and variability. Challenge is common to all
of the pre-cited models. Malone defines challenges as the fact that the player is
uncertain to reach his goals. This is why games have a certain difficulty level:
the outcome of the game depends on the player inputs, and not any input may
lead the player to his goal. To analyse the challenge level of a game, we may
consider different kinds of difficulty that a game may rely on. For instance,
Levieux proposes to distinguish between sensitive, logical and motor difficulty
to better analyse the challenge level of a task [20]. These kinds of difficulty are
particularly relevant for therapeutic games. For instance, patients having low
motor abilities can have standard logical abilities.
Variability is common to [23, 33, 18, 6]. In his model, Malone presents vari-
ability as the curiosity level. It may be sensitive curiosity, that is, the variability
of the representation, or the cognitive curiosity, which corresponds to how much
information the player can learn when playing a game. Also Vorderer et al ex-
plains that a game should propose many choices to the player [33], Ralph Koster
that we enjoy learning patterns in games [18], and Denis and Jouvelot that game
designers must derail the gameplay to create alternatives for the player [6].
Challenge and variability aspects are tightly related to the design of a thera-
peutic game. If we want to keep the player motivated, we have to inform him on
his progression towards the goal. Another basis is to adapt the challenge level, if
we dont want to create anxiety or boredom, or even let the player reach a flow
state [5]. And if we want to keep the player motivated in the long run, we must
provide enough variability. It is not enough to give the player the same task with
an adapted difficulty level, we must also make him learn new patterns, gather
new informations, explore the consequences of making different choices.

2.4 Le village aux oiseaux

The model presented in this paper has been built during the design of Le vil-
lage aux oiseaux. Le village au oiseaux is born from a partnership between four
companies (Tekneo, Seaside Agency, SpirOps, Neofactory) and two research lab-
oratories (CEDRIC3 , INSERM 4 ). Le village au oiseaux targets patients suffering
from the Alzheimer disease. The game stimulates the patients attention to slow
down the cognitive decline due to the disease. Le village au oiseaux is designed to
be a full-featured game providing a game world, a story and different gameplay
3
Center for Research in Computer Science and Telecommunications
4
National Institute for Health and Medical Research
 7

activities. The player takes the role of a photograph that will help the inhabi-
tants of a little town to avoid their towns destruction. The players mission is
to take pictures of rare birds. After the model presentation, we explain how the
model has proven useful during the design process of le village aux oiseaux.

3 Therapeutic Game Model

As seen in section 2.2, a therapy targets a particular condition which can be

a disease, a disorder or a function to recover. Thus, the context of use and
the patients condition are mandatory parameters to define a therapeutic game.
A game is not therapeutic de facto, the therapeutic function emerges from the
relation between the player and the game. To illustrate this fact, a good example
is Active Music Therapy which uses music as a mediation mean [2]. Most of the
time, playing music is not a therapeutic activity, but within the specific context
of psychotherapy, playing music becomes therapeutic.
Moreover, games are designed to please a specific audience. In consequence, a
strong understanding of the therapy, the player, the game, and their underlying
relations is necessary to design a therapeutic games. This is the main reason
why our analysis model is built around these three aspects: player, game, and
therapy. In the following section, we explain the scope of the model and our
global goals. Then, we study what is important for each aspects of our analysis
model. As these aspects have been thought to be analysed quite independently,
the relation between each aspect can be examined to find design inconsistencies.
Our analysis model is a list of items to look at for each aspect. To help the
game designer to understand which information are relevant, the items proposes
examples or questions the analysis has to answer.

3.1 Model Objectives

As seen in section 2.2, we decided to limit the scope of our model to therapeutic
games which are directly therapeutic, target the patient, and provide a training
which can be cognitive or motor. Our model goal is two-fold. First, with our
model, we can analyse existing therapeutic games to find their interesting fea-
tures. Second, the model should facilitate the game designers work during the
design process. At the beginning, the model tells the game designer which infor-
mation he has to gather, especially medical information. Knowing which infor-
mation is necessary should improve the communication between game designers
and health experts. Then, along the design and development of a therapeutic
game, the model helps to examine the coherency of the game design choices. For
example, an as-fast-as-you-can scoring system is not a coherent design choice
for a game that train fine motor control. In this context, such scoring system
could even be harmful for the patient. The model helps to reveal these kind of
design inconsistencies. More broadly, the objective of the model is to find the
therapeutic constraints. By doing so, the game designer is aware of which game
features he has to design very carefully (i.e. which part of the game design is
8

strongly constrained by the therapy). Also, the game designer knows on which
game features he can be more creative to provide a pleasurable and interesting
game.

3.2 Therapy

As stated before, the therapeutic function emerges from the relation between the
player and the therapy. As a result, the most important element of a therapy is
what the therapy can improve. Also, the definition of therapy states intended and
expected. As a therapy is a medical practice we advocate that this expectation
is based on a scientific evaluation of efficiency. However, scientific evaluation is
not possible from the very start of the development (i.e. during the conceptual
step). During the development, evaluation can be done on prototypes but will
generally not be sufficient. When the development is finished, long-term benefits
evaluations can start. These facts reveal that being clear-cut on scientific proof
of efficiency could be harmful to our model, it would exclude not yet finished
therapeutic games. As our model is intended to help the game designer during the
design process, we want to propose a smoother approach. So, our model defines
in-development game as te be evaluated. However, we want to know if these games
are founded on existing scientific evidences (e.g. similar positive effects proven
in similar games). Nevertheless, a minimal assessment of efficiency is necessary.
Games that have been negatively or not evaluated after their completion are
thus determined as not therapeutic by our model.

Expected short-term therapeutic value: e.g. burning calories, improv-

ing attentional network
Expected long-term therapeutic value: e.g. definitive recovery, life-long
supporting therapy
Scientific proof of efficiency: e.g. long-term effects demonstrated, to be
evaluated, discussed
Scientific references: Do similar useful games features have been scientif-
ically evaluated in other games?

3.3 Player

Understanding the player is an important aspect of game design. Thus, numer-

ous studies, either academical or industrial, have been conduced to analyse the
player and understand what he likes while playing. But, as stated by Ijsselsteijn
et al. [13], while information are numerous about usual targets of video games,
there is a lack of information on specific population (e.g. seniors). Although
there exists numerous usability guidelines for these population, these guidelines
are not informative on what, for example, seniors enjoy while playing or which
game genres they prefer. As a result, our player model is limited to demographic
information and health condition. For the game designer, the most important
is to understand what the target is able to do and learn regarding its health
 9

condition. Demographic information serves to know player preferences and in-

terests, as well as knowing which usability guidelines to follow. Also, we advice
game designer to create personas 5 of their targeted players. It gives guidance
while designing by determining the extreme profiles: a senior who never used a
computer, while another one is fond of new technologies. Using these extreme
profiles can make the game designer aware of the wide range of different indi-
viduals the game should be able to entertain. It also reveals potential usability
issues. Finally, we believe that game designers has to gather more information
on the targets interests if these information are not available. For example, the
game designer can conduce early playtests with similar games as we explained
in another article [22].

Age range
Gender
Particular conditions: e.g. early states of the Alzheimer disease, motor
control loss in upper limb
Abilities: Regarding their age and particular conditions, which are the
knowledge and abilities (i.e. motor control and cognitive functions) of the
player? Are they likely to present other particular conditions?

3.4 Game

As seen in section 2.3, numerous approach exists on what makes a good game,
but we have identified that challenge and variability were very common and im-
portant. Thus, we propose to analyse the gameplay of therapeutic games from
the challenge (goals, feedbacks, scores, and difficulty) and variability perspec-
tives. Also, information about the input and output systems are relevant ele-
ments while designing a game. As some games are recognised to provide positive
side-effects, the model asks if the game has features which are known to cre-
ate positive side-effects (e.g. some games are known to improve the mood or
self-esteem of their players).

Input system: How does the player interact with the game?
Output system: How does the game convey information?
Goals: Are there appealing goals? Are they short, mid or long-term goals?
Feedbacks: Which means are used to communicate with the player? Are
they informative on the player performance or progression?
Score: What does the score mean? (e.g. player performance, player pro-
gression, health improvement) Is the score informative on the progression
towards mid or long-term therapeutic goals?
Difficulty: How is the difficulty level chosen? (e.g. adaptive, manually cho-
sen by the player, manually chosen by another person) If adaptive, how does
it work? Which parameters of the game are modified by the difficulty level?
5
A persona is a description of an imaginary person which represents an entire group.
10

Variability: Does the game propose enough variability? (e.g. the player is
always doing the same sequence of actions, the player learns regularly to
master new patterns, the player can choose his own path within the game,
the player has to create new strategies to progress)
Usability: What are the minimal abilities and knowledge necessary to play
the game? Does the game features tutorials and explanations?
Expected positive side-effect: What can the game provide to the player
that is not part of the therapy itself?
Reported serious uses: Have the game or analysed features been used for
another serious purpose?

3.5 Underlying relations

The most important part of the analysis process is to validate the design co-
herency. This coherency is examined by evaluating the relation between each
aspects of a therapeutic game: the player, the game, and the therapy. Moreover,
each aspect of our model has been thought to be independent. So, the model
is useful to evaluate an entertainment game: finding its motivational features,
existing serious uses, and possible positive effects.
Player / Therapy: Does the player have a particular condition that can
be improved by the therapy?
Game / Therapy: What is the context of play? (e.g. at home, with a
therapist) What is the place of the game within this context? Which game
features are therapeutic? (e.g. the gameplay) Which game features are only
motivational means?
Player / Game: Is the player able to play the game? Is the game enjoyable
for the player? Is the game safe for the patients health?

3.6 Le village aux oiseaux

Le village aux oiseaux has been inspired by a scientific study of Green and Bave-
lier. This study proves that action games, particularly first person shooter (FPS),
improve the attention network of their players [9]. Le village aux oiseaux uses
these intrinsic effects of FPS to smooth the cognitive decline of patients suffering
from the Alzheimers disease. At the beginning of the project, we had informa-
tion about the game genre and the player demographic information. However,
with the model, we knew that we lacked information about our players abilities,
as well as which features of FPS games are effectively training the attention net-
work. Here, to illustrate how our model can work while designing a therapeutic
game, we explain two main design decisions the model helped us to make.
First, the analysis of the player-game relation makes clear from the very
start that it would not be possible for seniors to play a standard FPS. These
games require a strong ability to move in a virtual 3D space. We assumed that
starting a therapeutic game by having to learn a difficult ability would alienate
our players. So, we decided that Le village aux oiseaux would be a rail shooter
 11

(i.e. the player movement are automated and he can focus on the aim and shot
activity), and that the player will aim with a natural controller (i.e. similar to
the Nintendos wiimote). However, we had no data to prove that seniors may
enjoy a game were they have to aim and shoot with a natural controller. In
consequence, we conduced early playtests with similar games to be sure that
seniors will enjoy the game we were designing. These playtests and results have
been detailed in another article [22].
Second, as the gameplay was the therapeutic part of the game, it was really
important to know better what was therapeutic in it. This aspect required to
exchange with health experts. To sum up, the main concept that health experts
make us understand was that if the player can use his memory (i.e. instead of
processing the game environment) the therapeutic gameplay fails at training the
attention. But, challenge and scoring system of rail shooters are strongly tied
to the memorisation of the game sequences. The long-term players objective
is to perform each sequence perfectly. We changed this aspect of the game to
make it therapeutic. This is why le village aux oiseaux features adaptive levels,
in which the birds act differently each time. We knew that this change had
consequences on the long-term motivation provided by the game. With adaptive
levels, the player cant dedicate himself to perform perfectly the game sequences.
So, to compensate the loss of these long-term goals and strengthen the patients
motivation, we had to design other features (e.g. storyline, collection of pictures,
rare birds to discover).
In conclusion, the model helped us to find design inconsistencies. It also
revealed the consequences of each game design change and helped us to design
more efficiently.

4 Discussion

So far, we used our model to analyse five entertainment games and therapeutic
games. From these analyses, we draw some early conclusions on the model use-
fulness. First, the model can be used to determine if a particular use of a game
is therapeutic or not. For instance, Brain Age has some proven effects on the
brain (e.g. memory improvement). But Brain Age is not therapeutic, because
it does not target any particular medical conditions. Equally, the game Ico has
been designed for entertainment. Within the specific context of a psychotherapy,
Ico can be used as a mediation tool with children having a particular psychol-
ogy condition [32]. This specific use of Ico is therapeutic. With our model, we
can analyse Ico has an entertainment game, but we can also analyse the specific
therapeutic use of ICO. Finally, using the model is efficient to find similarities
and differences between an entertainment game and a therapeutic games. For
example, Dance Dance Revolution and Dance Town feature the same gameplay.
Dance Dance Revolution (DDR) is an entertainment rhythm game where the
player has to step on buttons according to arrow patterns displayed on a screen.
DDR targets adolescents and young adults. On the contrary, Dance Town targets
seniors to improve their coordination and balance. We have found that Dance
12

Town features three main adaptations: more protection around the dance pad,
songs and visuals that please the seniors, and easier arrow patterns (both in
rythm and complexity). These adaptations aims to make the game playable for
seniors, but also pleasurable by featuring, for instance, songs they like. In con-
clusion, the effect of improving coordination and balance is already present in
the original DDR, game designers of Dance Town have mainly adapted DDR for
the specific target.
However, we have also found weaknesses in our model. First, the model asks
the game designer to write full-text description for each items. This is very useful
and usable for game designers, but not very efficient to conduce automated meta-
analyses. Also, we limited the quantity of items in each aspects trying to keep
only essential ones. We have now to conduce more analyses to be sure that these
items are sufficient. We think that analysing more games like Virtual Iraq could
reveal other essential game features (e.g. environment, story).
The next step of our research will be the development of a tool. This tool has
the objective to facilitate the use of our model. Also, it will help to structure data
in order to conduce meta-analyses later. After that, we will analyse numerous
existing therapeutic games to better understand how they are designed and how
the relations between the player, the game, and the therapy work. This analysis
model and the building of more knowledge is the first step of our research, our
final objective is to provide methods to design therapeutic games more efficiently.

5 Conclusion

In this paper, we have presented how we built an analysis model for therapeutic
games. The model proposes to analyse separately each aspect of a therapeutic
games: the player, the game, and the therapy. The objective of the model is to
help a game designer while designing a therapeutic games: gathering information
with medical experts, examining the game design coherency by analysing the
relations between each aspect, evaluating the game efficiency regarding the two
objectives of a therapeutic game (i.e. being therapeutic and motivating), and
analysing similar games to find their interesting features and how they work
regarding therapeutic activity.
During the design of le village aux oiseaux, the model has proven useful.
It helped us to make design decision and predict consequences of game design
changes. Also, by conducing some analyses, we have found that our model was
able to deepen our knowledge on therapeutic games. Separating the three as-
pects of therapeutic games (i.e. player, game, and therapy), to then analyse their
relations has been powerful to understand how therapeutic games work. Also,
we have used our model to make comparison between similar games to better
understand the effects of certain games features. However, we need a tool to im-
prove the analysis model usability and to obtain structured data on therapeutic
games. The development of this tool is the next step of our research, then we will
deepen even more our knowledge on therapeutic games. Finally, our objective is
to build design methods alongside this analysis model.
 13

References
1. Alvarez, J., Djaouti, D.: Une taxinomie des serious games dedies au secteur de
la sante. Revue de lElectricite et de lElectronique, Societe de lElectricite, de
lElectronique et des Technologies de lInformation et de la Communication (SEE)
11 (2008) 91102
2. Benveniste, S., Jouvelot, P., Lecourt, E., Michel, R.: Designing wiimprovisation for
mediation in group music therapy with children suffering from behavioral disorders.
In: Proceedings of the 8th International Conference on Interaction Design and
Children. IDC 09, New York, NY, USA, ACM (2009) 1826
3. Burke, J.W., McNeill, M.D.J., Charles, D.K., Morrow, P.J., Crosbie, J.H., Mc-
Donough, S.M.: Optimising engagement for stroke rehabilitation using serious
games. Vis. Comput. 25 (October 2009) 10851099
4. Burke, J.W., McNeill, M.D.J., Charles, D.K., Morrow, P.J., Crosbie, J.H., Mc-
Donough, S.M.: Augmented reality games for upper-limb stroke rehabilitation. In:
Proceedings of the 2010 Second International Conference on Games and Virtual
Worlds for Serious Applications. VS-GAMES 10, Washington, DC, USA, IEEE
Computer Society (2010) 7578
5. Csikszentmihalyi, M.: Flow: The Psychology of Optimal Experience. Harper Peren-
nial (March 1991)
6. Denis, G., Jouvelot, P.: Motivation-driven educational game design: Applying
best practices to music education. In: Proc. Int. Conf. on Advances in Computer
Entertainment Technology, ACM Press (2005, ACM Press (2005) 462465
7. Djaouti, D.: Serious Game Design Considerations theoriques et techniques sur la
creation de jeux video a vocation utilitaire. PhD thesis, Universite Toulouse III
Paul Sabatier (UT3 Paul Sabatier) (2011)
8. Djaouti, D., Alvarez, J., Ghassempouri, R., Jessel, J.P., Methel, G.: Towards a
classification of video games. In: Artificial and Ambient Intelligence convention
(Artificial Societies for Ambient Intelligence). (2007)
9. Green, C.S., Bavelier, D.: Action video game modifies visual selective attention.
Nature 423(6939) (May 2003) 5347
10. Greitzer, F.L., Kuchar, O.A., Huston, K.: Cognitive science implications for en-
hancing training effectiveness in a serious gaming context. J. Educ. Resour. Com-
put. 7 (November 2007)
11. Guardiola, E., Natkin, S., Soriano, D., Loarer, E., Vrignaud, P.: Du jeu utile au
jeu serieux (serious game). Hermes 62 (2012) 8793 in press.
12. Hoffman, H.G., Doctor, J.N., Patterson, D.R., Carrougher, G.J., Furness, T.A.:
Virtual reality as an adjunctive pain control during burn wound care in adolescent
patients. Pain (85) (2000) 305309
13. Ijsselsteijn, W., Nap, H.H., de Kort, Y., Poels, K.: Digital game design for elderly
users. In: Proceedings of the 2007 conference on Future Play. Future Play 07, New
York, NY, USA, ACM (2007) 1722
14. Juul, J. In: The Game, the Player, the World: Looking for a Heart of Gameness.
Volume 120. Utrecht University (2003) 3045
15. Kato, P.M., Cole, S.W., Bradlyn, A.S., Pollock, B.H.: A video game improves
behavioral outcomes in adolescents and young adults with cancer: A randomized
trial. Pediatrics 122(2) (2008) e305e317
16. Kawrykow, A., Roumanis, G., Kam, A., Kwak, D., Leung, C., Wu, C., Zarour, E.,
Sarmenta, L., Blanchette, M., Waldispuhl, J., players, P.: Phylo: A citizen science
approach for improving multiple sequence alignment. PLoS ONE 7(3) (03 2012)
e31362
14

17. Khatib, F., Cooper, S., Tyka, M., Xu, K., Makedon, I., Popovic, Z., Baker, D.,
Players, F.: Algorithm discovery by protein folding game players. Proceedings of
the National Academy of Sciences 108(47) (2011) 1894918953
18. Koster, R.: A Theory of Fun for Game Design. Paraglyph Press, Scottsdale,
Arizona (2005)
19. Lazzaro, N.: Why we play games: Four keys to more emotion without story. In:
Game Developers Conference. (March 2004)
20. Levieux, G.: Mesure de la difficulte dans les jeux video. PhD thesis, Conservatoire
National des Arts et Metiers (2011)
21. Ludoscience: Serious game classification. http://serious.gameclassification.com
Last access 2011/06/08.
22. Mader, S., Dupire, J., Guardiola, E., Natkin, S.: Conception de jeux therapeutiques
pour seniors: lexemple du village aux oiseaux. In: Handicap 2012, Paris (June
2012)
23. Malone, T.W.: What makes things fun to learn? heuristics for designing instruc-
tional computer games. In: SIGSMALL 80: Proceedings of the 3rd ACM SIGS-
MALL symposium and the first SIGPC symposium on Small systems, New York,
NY, USA, ACM (1980) 162169
24. Marne, B., Huynh Kim Bang, B., Labat, J.M.: Articuler motivation et appren-
tissage grace aux facettes du jeu serieux. In: Environnements Informatiques pour
lApprentissage Humain. (5 2011) 6980 isbn: 978-2-87325-061-4.
25. McGonigal, J.: Reality is broken. Penguin Books,, New York (2011)
26. Prensky, M.: Digital game-based learning. Comput. Entertain. 1 (October 2003)
2121
27. Rizzo, A., Difede, J., Rothbaum, B., Reger, G., Spitalnick, J., Cukor, J., Mclay, R.,
et al.: Development and early evaluation of the virtual iraq/afghanistan exposure
therapy system for combat-related ptsd. Annals of the New York Academy of
Sciences 1208(1) (2010) 114125
28. Ryan, R.M., Deci, E.L.: Intrinsic and extrinsic motivations: Classic definitions and
new directions. Contemporary Educational Psychology 25(1) (2000) 54 67
29. Salen, K., Zimmerman, E.: Rules of Play : Game Design Fundamentals. The MIT
Press (October 2003)
30. Sawyer, B., Smith, P.: Serious game taxonomy. In: The Serious Games Summit @
GDC. (2008)
31. Singapore MIT GAMBIT Game Lab: Elude.
http://gambit.mit.edu/loadgame/elude.php Last access 2012/03/29.
32. Stora, M.: Histoire dun atelier jeu video : ico , un conte de fee interactif pour
des enfants en manque dinteractions. http://www.omnsh.org/spip.php?article84
(2008) Last access 2011/06/08.
33. Vorderer, P., Hartmann, T., Klimmt, C.: Explaining the enjoyment of playing video
games: the role of competition. In: ICEC 03: Proceedings of the second interna-
tional conference on Entertainment computing, Pittsburgh, PA, USA, Carnegie
Mellon University (2003) 19
34. Wei, T., Li, Y.: Design of educational game: A literature review. In Pan, Z., Cheok,
A., Muller, W., Zhang, X., Wong, K., eds.: Transactions on Edutainment IV. Vol-
ume 6250 of Lecture Notes in Computer Science. Springer Berlin / Heidelberg
(2010) 266276
35. Zyda, M.: From visual simulation to virtual reality to games. Computer 38(9)
(September 2005) 2532