leaving room for improvisation: towards a design approach ... · pdf fileleaving room for...

Leaving Room for Improvisation: Towards a Design Approach for Open-ended Play

Linda de Valk, Tilde Bekker & Berry Eggen Department of Industrial Design

Eindhoven University of Technology P.O. Box 513, 5600 MB Eindhoven

The Netherlands +31 40 2474537

{l.c.t.d.valk, m.m.bekker, j.h.eggen}@tue.nl

ABSTRACT Open-ended play with interactive objects provides children with the freedom to construct their own rules, goals and meaning. Instead of games with strict rules, open-ended play designs offer interaction opportunities as a trigger for creating personalized games. The process of developing these designs differs from designs with predefined use. This paper presents the further development of a design approach on how to design for open-ended play. We give an overview of related work and analyze eight existing open-ended play designs. Next, interviews with design students are discussed that illustrate the process of developing open-ended play designs. As a conclusion, we describe our design approach for open-ended play, including an improved definition of open-ended play, an overview of which design parameters have to be considered and advice for tailoring a design process to consider these parameters.

Categories and Subject Descriptors H.5.2 [User Interfaces]: Interaction styles (e.g., commands, menus, forms, direct manipulation), User-centered design.

General Terms Design, Human Factors.

Keywords Open-ended Play, Design approach, Children.

1. INTRODUCTION A dozen of colorful, interactive devices lie in a playroom. Two girls are rolling the devices over the floor to each other. When the devices roll, they change color. The girls try to guess the next color: “Blue! Red!” Another girl joins them, asking the girls what they are doing. They explain their little game to her. The new girl adds another rule: “If you’re wrong, it’s my turn.” On the other side of the room, two boys are shooting at each other with their devices. “You’ve been hit!” screams one of them, pointing at the other boy’s device that has turned into his color.

This scenario is based on the use of the ColorFlares [3]; play

devices that change color when rolled and that can transmit color when shaken and then pointed to another device. The scenario illustrates open-ended play: children can interpret the interactive play objects and create their own games with them. Designing for open-ended play is a rather difficult design goal as the design should deliberately leave room for interpretation by the players. For a couple of years now, several international researchers have been working on designing for open-ended play, including us. In discussions with colleagues and with others, we have realized that it is a challenge to design for play that is less predefined and to communicate how to deal with this process to others. Therefore, the aim of this paper is to explore how to design for open-ended play by positioning open-ended play in relation to previous work, comparing several existing designs and investigating the design process through reflection on practice. Developing this design approach is a work in progress.

This work is related to our research in the field of playful interactions. Play is defined as an intrinsically motivated activity with no direct benefit or goal that is situated outside of daily life [7, 14]. Salen & Zimmerman [21] define play as “free movement within a more rigid structure”. An important form of children’s play is so-called ‘free play’. Free play is described as play that is not initiated by somebody else (e.g. adults) but that provides children with the freedom to choose themselves what to do and how to do this, when to stop and to try something new [22]. Children are creators of their own play, using whatever toys or (natural) props they prefer. Free play is considered an important learning experience for children. Through free play, children develop various skills and increase their understanding of the world around them [22]. Children engage mostly in outdoor free play in the yard at home, at their own street and at local parks or playgrounds [25].

Open-ended play is inspired by the notion of free play and the opportunities it provides to children to create their own play. Designs for open-ended play can be toys, props or large-scale playgrounds. For example, FeetUp [19] is a design that consists of lights embedded in children’s shoes that react on the movement of the children. It uses a simple rules system so that children can create their own challenges and goals with it. With traditional toys, children also vary the rules and goals and the meaning of objects during play. For instance, pretending a stick to be a magical sword or adding extra rules to an existing game of hide-and-seek. We focus on intelligent play objects that are designed to incorporate and support the variations in this attachment of meaning.

In previous research, we have gained some experience with designing for open-ended play. Doing several design projects and having Industrial Design students applying this approach has led

Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than ACM must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from [email protected]. IDC '13, June 24 - 27 2013, New York, NY, USA Copyright 2013 ACM 978-1-4503-1918-8/13/06…$15.00.

92

Full Papers IDC 2013, New York, NY, USA

to insights into the design process. For example, we have discussed open-ended play as a design value for designing playful interactions [3], evaluated the influence of multimodal output on open-ended play [13] and explored how to design for specific playful experiences in open-ended play environments [24]. Elaborating on this work, the next step is to develop a more clearly defined design approach for open-ended play. In order to achieve this, three complementary activities were carried out. Firstly (a), we grounded open-ended play in literature on play and games aiming at a better positioning of open-ended play. Secondly (b), we analyzed eight existing open-ended play designs, developed at several research groups around the world, to examine how an open-ended play approach influences the design parameters of the play objects. Thirdly (c), we interviewed design students at our department about their projects on open-ended play. While the design analysis focuses on parameters of the final designs, the interviews describe the process towards that final design. Although this paper presents the three activities in succession, these activities did not occur in a chronological order, but rather parallel and simultaneously. The results of the three activities add up to a model illustrating our design approach for open-ended play.

2. POSITIONING OPEN-ENDED PLAY IN RELATION TO FREE PLAY AND GAMES Several scholars have discussed variations of play with a component of freedom and flexibility. They represent a variety of backgrounds including game design, interactive art and sociology. They look at play using different lenses; some discuss their work in relation to interactive objects, while others focus on play in general. This is represented by the definitions and characteristics they offer. We will discuss these variations below.

In his work on play and games, Callois [5] makes a distinction between two opposite forms of play: ludus and paidia. He describes ludus as structured, rule-bound and goal-directed play and paidia as unstructured and spontaneous improvisational play. Ludus refers to solving a particular problem for personal satisfaction and concerns games as chess, hide-and-seek, sports games and bets. These games are all very common for children. Children make all kinds of bets with themselves or with others, for instance: holding their breath or trying to stare longest without blinking. Rules are very important; either invented or part of the official game, they should always be obeyed. Normally ludus leads to a development of a special skill, a certain mastery over an action or product or the discovery of a satisfactory solution. Paidia does not have such a clear outcome. As examples of paidia Callois mentions spinning around to get dizzy, somersaults, scribbling, or a child laughing at its rattle. Paidia is related to the need for disturbance or tumult, expressing itself in activities as holding up a queue, disturbing the work and play of others and endlessly cutting up paper with a scissor.

Bateman [1] describes paidia as an amusing, creative and chaotic activity, but also short-lived as it soon becomes a game (and thus ludus). As an example of a paidia game he mentions Sink; a natural play activity played close to a large body of water. It consists of throwing something that floats into the water and trying to sink it by throwing another thing at it. Bateman [1] sees play as a journey from paidia to ludus and vice versa. For instance, when temporarily escaping from the rules of ludus, one can move back to paidia. To support paidia, no complex learning should be involved and players have to be encouraged to experiment.

Nachmanovitch [17] mentions free play as an activity of spontaneous free improvisation. He identifies two types of free play, namely Lila and Bricoleur. Lila means divine play and is a state in which players enjoy the simplest of things. Bricoleur (or Bricolage) is a state in which players spontaneously improvise with what the environment offers them.

As mentioned before, Salen & Zimmerman discuss play as a rigid structure with free movement [21]. Costello & Edmonds [7] continue on this definition by making a distinction between directed and emergent play. They state that designs for a play experience should create a balance between directing play (rigid structure) and offering opportunities for play to emerge through the activities of the player (free movement) [7].

In their work on free play in open-ended interactive art environments, Morrison et al. define free play as “non-narrative, non-competitive, and without logical ending point” [16]. Play is kept alive as the ending point continually evolves. This relates to Carse’s [6] work on finite and infinite games. The aim of a finite game is winning the game, while an infinite game is played only for the purpose of continuing the play. Morrison et al. mention the followings two aspects of free play. First, it is constantly rejuvenated as it is co-constructed and co-authored by its participants. They are in control of discovering what the design has to offer them. Secondly, it is free from any predetermined order of meaning. People construct their own meaning and invent their own interpretations from interacting with the design [16].

Table 1 illustrates the differences between games and free play, as described by the researchers discussed before.

Table 1. Properties of Games and Free Play

Games (ludus) Free play (paidia)

Structure Chaotic

Finite, with (end) goals Infinite, no logical ending point

Fixed (game) rules Improvisation, spontaneity

Predefined Own construction of meaning

Challenge, competition Sensation of play, Expression

In this table, the difference between games and free play appears to be rather distinct, while in fact there is a large grey space in between. Open-ended play can be positioned somewhere in this grey space. At the start, it mostly resembles free play, as little is predefined. But it can move either way: when players come up with rules and goals, they move towards games while when they play for the sensation of it, they move more towards free play. This also holds for free play itself. Children are encouraged to use their own initiative to play spontaneously without, or with little, structure. This kind of play can eventually lead to a game with rules (ludus). Free play differs from open-ended play, as it stresses the freedom children have to play with whatever whenever they want. Open-ended play somewhat restricts children in their free play as it offers objects with design intentions. The challenge for open-ended play is to develop designs that are specific and easy to understand but also general enough to encourage imagination and creativity in how to use them.

Open-ended play can be seen as a form of improvisation; children create their own play spontaneously and without preparation. Open-ended play designs actively encourage a child’s imagination

93


by not determining game rules or a strict process. Just as in improvisation, children in open-ended play are influenced by their everyday experiences. They might have seen a television show with characters they want to act out, just as musicians are inspired by a tune they heard that day. Children already improvise often in play by creating their own rules and meanings. Even if children play rule-based games, they create their own rules on top of them. Bruce [4] describes twelve important features of play, including: children make up the rules as they play, to get a feeling of control. For instance, they play tag, but you cannot be tagged if you stand on one leg. Vygotsky [26] claims there is no play without rules. In his view, a child acts according to the existing meaning of things. Rules can be instructed by adults or arise among children themselves. The children’s mental model of the rules does not always correspond with the system’s conceptual model. For instance, they assume that a certain sound feedback in an adventure game means they are going the wrong way, while it actually means for the system that they found a special treasure. As Gray [11] describes it, rules of play are mental concepts of which the understanding must be shared amongst the players. These shared mental models are also of importance in improvisation [9]. Shared mental models are individual assumptions that are similar for a group; the feeling of “being on the same page” [9]. In improvisation, consensus on these assumptions is essential to lead to a meaningful result. The same counts for open-ended play: when a person is not playing by the (improvised) rules, this can destroy the game that is currently played.

Summary: Open-ended play can be positioned in the grey space between games and free play. At the start, open-ended play mostly resembles free play and its properties as described in Table 1. Children can continue to play for the sensation of it or move in the direction of games as game rules and goals are developed.

3. ANALYZING EXISTING DESIGNS FOR OPEN-ENDED PLAY In the previous section, the design space of open-ended play has been positioned in relation to games and free play. To further explore this design space, we use this knowledge from literature to take a closer look at a number of interactive designs for open-ended play. These designs illustrate how open-ended play has been interpreted by various design researchers so far. We analyzed eight of these interactive designs. This selection was made based on a literature search on interactive open-ended play designs. Three of the designs were developed by our faculty; the other five are from research groups around the world. These eight designs are: Statue [20], FeetUp [19], Morel [15], jogo [8], Interactive Pathway [23], LEDtube [3], ColorFlare [3] and FlowSteps [24]. We will first describe the designs briefly and then analyze them in comparison to each other.

3.1 Designs for open-ended play Statue [20] is a fanny pack that reacts on children’s movements by audio-visual feedback. The system only reacts on vigorous movements in the Y- and Z-axis. The fanny pack encourages children to play: “Statue stimulates children to play games related with being a statue or moving without being noticed, which are commonly played by children, according to our ethnographical observations.” FeetUp [19] lets children play with gravity. It consists of pressure sensors and lights embedded in children’s shoes. As soon as the shoes are not in contact with the floor or any other surface (i.e.

children are jumping) lights on the shoes start to flash. “The system incorporates a simple rules system so that children have the opportunity to create their own system of rules, challenges and goals based on the basic, but consistent information they receive.”

Morels [15] are soft, cylindrical objects that can be kicked and thrown like balls. When two Morels are in each other’s vicinity, this is communicated by sound feedback. While other Morels are nearby, players can ‘charge them up’ by holding and squeezing their Morel. The other Morels will now respond with sound feedback, communicating that they are being charged up. When the charging comes to a maximum, the charger can squeeze once more to launch up the other Morels. “Morel encourages the improvising of new games and behavior by not defining game rules on its own, but by providing players the ability to know the existence of other Morels in the vicinity, and to remotely make other Morels launch up in the air.”

Jogo [8] is a music generator consisting of a round tabletop and colored ping-pong balls. The play surface consists of four circles of sixteen holes dividing the surface into sixteen steps, each one representing one sixteenth note in a measure of music. The balls have five different colors which each represent a different sample. These balls can be placed in the holes on the surface. Their position and color are detected and the corresponding sounds are played. “Simplicity and ambiguity in the design aims to inspire and motivate free play.”

The Interactive Pathway [23] consists of two separate wooden paths with pressure-sensitive mats that control motors next to the mats. When children step on a mat, the corresponding motor starts to spin. In this way, the children are accompanied by the motors spinning one by one while they walk on the pathway. “The diverse play patterns observed resulted from a simple design which enabled open-ended play patterns instead of predetermining any one or two activities.”

Figure 1. Eight designs which were analyzed. From top to bottom, from left to right: Statue [20], FeetUp [19],

Morel [15], jogo [8], Interactive Pathway [23], LEDtube [3], ColorFlare [3] and FlowSteps [24].

. .

94


LEDtube [3] is a cylindrical shape emitting light at each end. It reacts to children’s behavior by changing the color of its light (red, green or blue). Two versions were created: one changes color when it is rolled and the other when it is shaken. The LEDtube, “which had fairly limited interaction possibilities, was designed to explore whether players enjoy interacting with open-ended play objects.”

ColorFlare [3] is based on the LEDtube, but with more interaction possibilities and more diverse output. While rolling the ColorFlare, it changes color. When shaken, it will start to flash. In flash mode, the ColorFlare can transmit its color to another ColorFlare in range. “We increased the number of states to which the children can allocate meaning and thus support more diverse games, which may increase the chance that children enjoy playing with it over time.”

FlowSteps [24] are interactive mats that react with light on children’s movements. Two different colors of lights (red and blue) are used with different interaction behavior. Children can step on a light to catch it or move it around by stepping along on other mats. “Children can decide themselves to play with the red or with the blue light. By giving meaning to the interaction opportunities, they can create their own game rules.”

3.2 Analysis of design parameters In order to explore the design decisions that play a role in designing for open-ended play, we analyzed the eight designs in reference to several design parameters. For this, we looked at the typical HCI model, which consists of a user (human) and an object (computer) who interact with each other through input and output devices [12]. In open-ended play, designs can consist of multiple objects and sometimes these objects can also communicate with each other. Moreover, multiple people can interact with the objects simultaneously and communicate with each other while involved in social play. Based on an informal comparison of the designs, the following three parameters were found to describe important differences between the designs: output modalities, communication between objects and ownership of the objects. The context of play also adds formal game elements to the picture. We selected three formal game elements described by Fullerton et al. [10]: players, procedures and rules. Below, the different parameters will be briefly discussed. After that, we will discuss the analysis of the designs on these parameters.

Players concerns if the design is intended for single-user or multi-user purpose. The players can have different roles and perform different interaction patterns as cooperative play or competition [10]. Procedures describe the required methods of interaction. A design can require a sequence of steps (multiple actions) existing of multiple rules. For instance, ColorFlare incorporates a sequence: for transmitting a color, players first need to shake it so that it starts flashing before they can point it towards another ColorFlare to transmit their color. Rules are closely related to procedures and explain the allowed actions and how the design responds to that. Rules describe the interaction of the design which can consist of simple, cause-and-effect rules or more complex, multiple interaction rules. For example, FeetUp has one simple rule: if the shoes are not in contact with a surface, the lights on the shoes start to flash. Morel on the other hand have more complex rules, e.g. holding and squeezing a Morel charges it up and lets other Morels respond with sound feedback. Modalities are part of the feedback from the design. It can consist of one or multiple modalities. For example, Statue contains two modalities

(audio and visual) while jogo has one (audio). When the design exists of multiple objects, Communication relates to if these objects can communicate with and influence each other. Ownership concerns if the designed objects are personal or shared.

Tables 2a-2c give an overview of the design analysis. We divided the six parameters into three pairs to create a comprehensible overview.

Table 2a. Design analysis on players (vertical) and ownership (horizontal)

Personal object Shared object

Singe player - -

Multi player 1, 2, 3, 6, 7 4, 5, 8

Table 2b. Design analysis on interaction rules (vertical) and procedures (horizontal)

No steps Sequence

Simple 1, 2, 5, 6 -

More complex 4 3, 7, 8

Table 2c. Design analysis on communication between objects (vertical) and modalities (horizontal)

1 modality >1 modalities

No communication 2, 4, 5, 6 1

Communication 7, 8 3

Please note: the numbers in Tables 2a-2c correspond with the order of the designs as they are described before: Statue is 1,

FeetUp is 2, Morel is 3, jogo is 4, Interactive Pathway is 5, LEDtube is 6, ColorFlare is 7 and FlowSteps is 8.

As the tables show, the designs are not similar in their design parameters. Table 2a classifies the designs related to players (single or multi) and ownership (personal or shared objects). All designs are multi-player. The majority consists of personal objects that are either attached to the body (Statue, FeetUp) or tangibles (Morel, LEDtube, ColorFlare). The shared objects consist of tabletops (jogo) and objects on the floor (Interactive Pathway, FlowSteps). In Table 2b the designs are described in terms of their procedures (no steps or sequence) and rules (simple or more complex). The most ‘simple’ designs have no steps and only cause-and-effect rules (Statue, FeetUp, Interactive Pathway, LEDtube). In between is a more complex design (as it has multiple rules) but with no steps (jogo). On the other end are three designs that require a sequence of steps in their interaction and consist of more complex rules (Morel, ColorFlare, FlowSteps). This asks more from the players if they want to understand all the rules, but this also offers them more opportunities to play around with these rules. Table 2c lists the parameters of output modalities (one or more) and communication between objects. Most designs include only one modality. Most of them have no communication between objects (FeetUp, jogo, Interative Pathway and LEDtube) while others do (ColorFlare, FlowSteps). The designs that have more than one modality are divided into no communication (Statue) and communication (Morel). More modalities provide

95


more opportunities for interpretation, i.e. more properties that players can attach meaning to and involve in their game play.

Summarizing these tables shows that in the design space of open-ended play there is still a lot of freedom to move around. The described parameters show how different design decisions can lead to a variety of open-ended play designs. Statue, FeetUp, Interactive Pathway and LEDtube are designs with very simple, one dimensional interaction possibilities. ColorFlare and jogo already offer more variety in their interaction possibilities but no or only a simple sequence to go through. Morel and FlowSteps have interaction possibilities that go through several steps, e.g. first action A and then action B for reaction C. The tables are not meant as an assessment tool, i.e. one design is not better than the other. This depends largely on other factors as the goal of the project, the context and the target group. Summary: When designing for open-ended play, different design parameters should be considered, including: players, procedures, rules, modalities, communication and ownership. These parameters influence design decisions that can lead to various open-ended play designs. The design analysis further grounds open-ended play, but remains rather theoretical with a focus on the final design. The next step is to investigate designing for open-ended play in practice. In the following section we will examine the design process by reflecting on the interviews we held with students working on designing for open-ended play.

4. INTERVIEWING STUDENTS ABOUT DESIGN PROCESS This section describes the interviews held with Industrial Design students on their design process for open-ended play. At the time of these projects, the grounding and design analysis of open-ended play was still under development. In this section, we will first discuss the methodology. Then, the project cases will be briefly described. After that, we will present the results on the definition of open-ended play and the design process.

4.1 Interview method Six recent Master student projects were selected (see the next paragraph for a brief description of all six projects). All projects were done at the “Playful Interactions” group within the Department of Industrial Design, Eindhoven University of Technology. At this department, students perform design and design research projects. Most of these projects are based on a design brief provided by the department or (in cooperation with) real-world clients and stakeholders. Students have a lot of freedom to give their own twist to the project. For instance, students can choose a specific target group or form of play to design for. All students have their own style and vision and their interpretation is key in our educational system (see also [2]). Most of the projects were coached by one of the authors. As our approach for open-ended play is constantly growing and developing, we already provided the students of the projects described in this paper with some preliminary insights and gave them various stimuli for their process. For example, the differences between games and open-ended play were discussed and literature on play and age groups was distributed.

The projects were examined as follows. First, the project reports were reviewed, with special attention to the process and design rationale. Then, the students were interviewed individually by one

of the authors. In these interviews, the focus was on their goals, vision and approach for the start, middle and end of the project. Our aim was to explore how their design process for open-ended play developed during the project. A ‘standard’ interaction design process usually goes iteratively through phases of establishing requirements, designing alternatives, prototyping and evaluating [18]. Preferably, this process is combined with a user-centered approach: users are involved throughout the design process [18]. We were interested to find out how the students started the project and how they determined their aim of the project; what changed during the project and why; what result the project had and how satisfied they were with this result; and what they would have done differently if they would do the project again. Questions asked included: ‘What was the design brief at the start of the project?’ and ‘What went well and what didn’t go well in your process concerning open-ended play?’

All interviews were audio recorded and transcribed afterwards. Summaries of the interviews were communicated to the other authors and served as input for discussion sessions.

4.2 Project cases Six Master student projects were selected as design cases (see Table 3 and Figure 2 on the next page for an overview of the projects). One project (Bababa) was a short project involving two students. Another project (Wondrous Imagination) was an individual graduation project. These two projects were initiated by the students themselves. The other four projects were all individual projects; one design project (CoCones) and three design research projects (Shuffle, Cooil and Push/pedal car). These projects were defined by the faculty in a design brief. This design brief made the overall goal of the project and its connections to previous work and current (research) projects explicit, but it left opportunities for the students to find their own direction of interest. This paragraph briefly describes each of these projects, in their chronological order.

Shuffle (2010) was a design research project by Koen Verbruggen. This project focused on stimulating social play among children. The design Shuffle offers a personal device to children with which they can exchange colors of light with other players. The children need to negotiate with each other to get the desired color or help another with achieving this.

Cooil (2011) was a design research project by Stephan Hoes. The goal of this project was to explore how multimodal feedback can contribute to social play. A design called Cooil was developed that demonstrates different output opportunities when playing together than when playing alone.

CoCones (2011) was a design project by Patrick Leijte. His goal was to increase the value of respect among high school students. The project’s focus was on designing objects to be used during gym class at high school. The design CoCones was developed: cone shaped objects that react with light feedback. The teacher can decide which games to play with the objects. For instance, when students run towards a cone and shake hands, the lights of the cone change to their group’s color.

Push/pedal car (2011) was a design research project by Martijn Kors. The goal of the project was to investigate how cooperative play can be supported by adding a system to an existing design. Push/pedal cars were chosen as an existing design and a system with tokens was added that indicated when children should change position or should find a new child to ride around with.

96


Bababa (2012) was a short project of approximately four weeks by Chris Gruijters and Gijs Houdijk. This was a learning activity initiated by the students themselves. The goal of the project was to make the old-fashioned ball pit more attractive by adding sound. Interactive balls were developed that reacted on movement with sound feedback. Ten of these balls were put into a ball pit together with normal balls. Young children played with the ball pit and tried to discover all the interactive balls.

Wondrous Imagination (2012) was the graduation project of Gijs Houdijk. This project was initiated by the student himself, building on his previous projects concerning playful interactions. The goal was to create an immersive world that triggers children’s fantasy. A carpet with abstract interactive objects was developed offering children the opportunity to direct their play with tools and puppets instead of being an active actor themselves.

Table 3. Overview of the six student projects

4.3 Results interviews In this section, the results of the student interviews are discussed concerning the definition of open-ended play and the design process.

4.3.1 Defining open-ended play In discussions on the notion of open-ended play with students, it became clear that open-ended play has less degrees of freedom than free play, but is still far removed from games. When technology using sensors and actuators is integrated in the design of play objects, this leads to a kind of structure. Certain calibration values and action-reaction behaviors need to be set beforehand. Playing with these objects is not as open as non-interactive free play objects, but a lot less strict than games with fixed rules. In discussion with the students, the following definition of open-ended play was formed:

“Open-ended play is play without predefined (game) rules in which players can attach meaning to the design properties and the interactions themselves while playing. Its goal is to trigger a player’s creativity by leaving room for interpretation.”

Actually, claiming that there should be no predefined rules, already defines a rule. But not all rules are similar. The students distinguished game rules and interaction rules. Game rules are rules related to the actual game play; for instance, how to reach a goal or how to gain extra points. These game rules relate to the fixed rules in Table 1. Interaction rules, on the other hand, are

rules for interacting with the design as input-output relations and affordances; for instance, shaking an object emits a certain sound. These interaction rules are similar to the rules described as design parameter in 3.2. In open-ended play, game rules may be left open, while some interaction rules have to be defined by the designer. For instance, the Shuffle is designed to be a personal, individual object that children can hold in their hands. The U-shape of the design encourages children to place their Shuffle against another child’s Shuffle, which activates the exchange of colors (interaction rule) that children can interpret as the goal being to collect all different colors (game rule).

Adjusting the aspects of openness (ambiguity) and complexity is a key action when designing for open-ended play. Overall it is important that there is some ambiguity in output and rules so that players can create their own goals and follow their own steps to accomplish these goals. On the one hand, the design should not be too open. At one point it will be nothing anymore, no real design. On the other hand, too many interaction rules will make it too complex. Some restrictions need to be set and some aspects need to be defined. Making these decisions may feel like a risk as it looks so conclusive, but it is necessary to be able to claim anything with the design, especially in a design research project. Open-ended play should focus on finding a balance between spontaneity and structure. A design that is too structured prevents children from being spontaneous, but without any structure children are not able to be spontaneous. This refers back to the definition of play as “free movement within a more rigid structure” [21], finding a balance between directing play and emergent play [7].

Context is another important factor that can help in positioning the design but also limits it. For instance, when play occurs in an indoor room, players cannot throw balls or run as far as they would do on an outside playground. CoCones is an example of a project that uses the context as part of its structure. It was specifically designed to be used during gym class and uses the

Project Weeks (Year)

Initiative

Shuffle Design research

10 (2010) Department

Cooil Design research


CoCones Design research


Push/pedal car Design research


Bababa Design activity

4 (2012) Students

Wondrous Imagination

Graduation 24 (2012) Student

Figure 2. The six selected student projects. From top to bottom, from left to right: Bababa, Shuffle, CoCones,

Wondrous Imagination, Push/pedal car and Cooil. . .

97


shape of cones, a familiar object in this context. Bababa is also very context dependent. If the interactive balls were not part of a ball pit, but solitary objects to play with on the school yard for instance, this would lead to different forms of play. Instead of trying to discover the interactive balls in the collection of balls in the ball pit, the interactivity might be used while throwing or rolling the balls to each other. Wondrous Imagination showed that a design can also create its own context. The play mat served as a context for fantasy stories to develop upon.

Summary: Designers can predetermine various interaction rules and let users come up with their own game rules. Moreover, the relation between ambiguity (as important element of open-ended play) and complexity (as a choice in design parameters) should be chosen well. A balance has to be found between structure and spontaneity. Also, context needs to be taken into considerations to support and restrict the design.

4.3.2 Design process This section describes the design process for open-ended play. The interviews with the students illustrated that designing for open-ended play asks for a somewhat adjusted process. Students indicated that, as aspects are deliberately left open, it took longer for the design to become more specific. Therefore, especially the first design iteration took more time than in previous projects in which design decisions could be made more easily on the ground of theories or standards. With open-ended play, design decisions can be based on for instance user profiles, observations or a designer’s vision. Students stressed that it is extremely important to make these decisions; otherwise the design remains too broad – which can be a pitfall particularly for open-ended play.

Defining the design space For the students, the project usually started with further defining the design space. This included making decisions on context (e.g. outdoor playground), user group (e.g. 10-12 years old boys) and the overall goal of the design (e.g. stimulating physical play or encouraging social interaction). Most students did not have a particular play scenario in their mind at the start of the project. Usually this arose after some observations or a first explorative study. Some of the students worked with a specific higher goal, related to their vision as a designer. For instance, one student wanted to stimulate social behavior as he observed the society becoming less social nowadays (Shuffle). Another student wanted his design to lead towards more respect among high school students (CoCones). Such a value was mainly used as an inspiration; the short duration of the student project did not support real validation. The students mentioned it also helped them in their thinking process: instead of focusing on details of the design or its interactions, designing for a certain value helped them in creating the bigger picture in the project. Designing for open-ended play particularly meant choosing a certain framing and exploring what happens within that frame. This frame helped in establishing an environment in which interactions could occur. For instance, the Bababa project focused on developing an interactive ball pit. Their frame was thus a ball pit with a bunch of same-sized balls with different colors that children around the age of four could play with. Then, a designer’s role is to decide on adding and removing certain aspects within that frame by tweaking the corresponding parameters such as ownership and output modalities. For example, the amount of balls can influence if it is considered a personal or a shared object (ownership) and each ball can react with only sounds, or also lights or vibrations (output modalities).

Summary: Designers should early on in the project define the design space by shaping a frame with clear boundaries related to user values, age, context and forms of play. This is especially important for open-ended play, as a design with no restrictions becomes difficult for players to interpret.

What (not) to design When designing for open-ended play, this must not be considered as ‘just designing something’ and assuming that children can come up with their own games anyway. It is easy to assume that some aspects are not worked out but that children will do the interpretation themselves. Designing for open-ended play means taking a risk, as one of the students put it. As a designer you do not know at the start of the project what the outcome will be. You have some assumptions, but these assumptions can turn out to be wrong. Open-endedness is about making conscious choices in the design process on what is going to be designed and what is going to be left open for the children to interpret while interacting with the design. The age of the children can influence these decisions. The Wondrous Imagination project showed that young children (around the age of four) are not capable of dealing with very abstract play objects yet. They needed less abstract objects and an introducing story in order to get drawn into fantasy play.

Throughout the process, the design is approached both top-down and bottom-up. The design moves from a high level intention, through the dialogue between the user and the system in the middle level towards the tangible actions in the lowest level. Open-ended play is then triggered at the low level (by sensors, actuators and affordances) and will move up from there. Students made design decisions based on various aspects. Some of them were related to the students’ own development while others were more practical decisions, e.g. choosing a technology that is currently available. Making these design decisions occurred at different moments in the design process. For instance, in the Push/pedal car project the decision to use existing pedal cars was made early on in the project. Children are familiar with these cars and know how to use them; e.g. sitting on them and riding them or pushing others children. After first observations, the design decisions was made to add tokens with different colors to the design, but to leave open if these tokens were personal or shared objects.

Designers often choose to integrate several functions in order to support various forms of play. But this can make the design too complex. For example, in the Shuffle project the final design offered so many interaction rules that it became too complex for children to understand and use all of them. Besides that, the children also came up with a lot of self-invented game rules. Reflecting on the design, the student mentioned that one or two interaction rules would already have been sufficient for diverse play to occur.

Summary: Designers should make clear decisions on what to design and what to leave open for the player’s interpretation, creating a balance between openness and complexity. The design parameters as described in the design analysis can be used as guidance for design decisions, e.g. number of output modalities and whether to provide communication between objects.

User confrontations After making the first design decisions, it is crucial to quickly try out the design concept with users from the target group. This can be done best by confronting user with a first (interactive) prototype. By doing that, it becomes clear what actually happens

98


when users from the target group start interacting with the design. As a designer for open-ended play, you have certain assumptions but its openness makes it hard to predict. Children can come up with games and rules that were not expected beforehand. This makes designing for open-ended play unpredictable, but this is also its strength. It is exciting and, as one student mentioned: you don’t have to worry about restrictions. When such confrontations are done early in the design process, the insights can be inspirational for next iterations and improve the overall design concept. In order to be able to do these kind of confrontations, designers need to quickly make some design decisions. The real design part mostly happens after a first, quick iteration. When children have been playing with the design, one can determine which properties should be intensified for the next iteration. Then, it is crucial, as in the ‘standard’ interaction design process, to repeat such user confrontations frequently in the remainder of the design process. To get most out of these confrontations, students indicated that it is important to give children a chance to get used to your design and the setting. If children don't start to play with the design themselves, a short introduction can help. This might feel unnatural, but as soon as they start interacting with the design, it is still valuable to watch how they interact and play with it. In the Wondrous Imagination project, three iterations were performed, each ending with a user evaluation. The design changed from an abstract white carpet with simple objects to a visually richer carpet showing a top-view of a natural environment with objects that children could interact with in various ways. The final design really evoked fantasy play, which was the aim of project. The reflections moments helped in developing a stronger design that fitted this aim.

Summary: Designers are encouraged to confront users of their target group with their design early and frequently to validate the design and gather valuable insights for further design iterations. For open-ended play, these user confrontations are even more crucial as the creation of meaning is unpredictable and only happens during play. Doing early user testing is necessary in order to find out what happens with it.

5. DESIGN APPROACH FOR OPEN-ENDED PLAY In this paper, we have described three activities that contribute to an improved understanding of how to design for open-ended play. In this section, we will summarize the results in a discussion of our design approach for open-ended play. This approach is illustrated by the model in Figure 3 (see below).

This model illustrates the iterative design process for open-ended play. The first activity (a) grounded open-ended play in literature and positioned it in between games and free play. This clarified the design space of open-ended play. In the second activity (b), the analysis of eight existing designs, we looked at different design parameters and how they influenced design decisions. These design decisions illustrate how designers for open-ended play can vary with such parameters in order to create a playful design that fits their project goal and context. The third activity (c) consisted of interviewing students who performed a project on open-ended play on their design process. These interviews made clear how others experience designing for open-ended play. This led to an updated definition and a clarification of what the strength, pitfalls and challenges of open-ended play are. For instance, open-ended play gives control to the players to create their own play. But when the design offers too many opportunities for improvisation, it might confuse the players and lead to discomfort or even bring the play experience to an end. Moreover, designing for open-ended play asks for different actions throughout the design process. At the start designers should frame the design space and its openness. During the process, designers have to explicitly decide what properties to design and what properties to leave open for the interpretations of the players. Furthermore, open-ended play asks for an explorative design process with an emphasis on early prototyping and frequent user confrontations, even more so than in a standard interaction design process. These actions are represented in the model by four separate blocks. The model also indicates how the two other activities (positioning and design analysis) relate to the design process. The positioning of open-ended play supports framing the design space, as it gives some first direction into designing for open-ended play. It underlines certain characteristics of open-ended play and provides a starting point. This positioning stays of importance in the next block of choosing what to design. The design parameters of the design analysis serve as input for making design decisions in the next block. These parameters should be considered when creating ideas and concepts and developing prototypes. We believe this model can support designers and design researchers in their work on designing for open-ended play as it can help in setting up and carrying out projects in this research area. For us, the model already has value in explaining designing for open-ended play to (new) students. By presenting the different steps together with the definition and design parameters, we provide students with pointers for their design process.

6. DISCUSSION Both the analysis and the interviews made clear that designing for open-ended play is still a rather undefined process. This was further underlined in the discussions between the authors. For instance, we had to search for the right terminology and an appropriate method of analysis. Although the design analysis covers designs from several international researchers, the interviews were done with students from only our own department. There is an undeniable correlation between our research and what these students are taught. On the other hand, interviewing them helped in further understanding the design approach and preparing for new student projects. The summaries might also be relevant for other design projects, but we would like to stress its ‘specific’ importance for open-ended play. By not predefining its use and functionalities in detail, designing for open-ended play asks for some steps (as early user confrontations and defining the design space) to be even more important than in a traditional user-centered design process.

Figure 3. Model of open-ended play. . .

99


By no means we see the results presented here as a final piece of work. To further support our understanding of how to design for open-ended play, we are planning to continue with this design approach and use the knowledge from this paper in our future work. For instance, we currently used six design parameters in the design analysis. By examining the Tables 2a-2c and the first insights, it became clear that we can extend this work by including other parameters as the type of modality or the fixed or flexible location of objects. Also, it opens up possibilities for new designs that focus on one player with multiple objects or combining diverse objects in one design. Furthermore, a new group of students is currently working on projects about open-ended play. Compared to the student projects described in this paper, we gave the current students more guidance in the shape of workshops, presentations and feedback sessions. We intend to interview these students throughout the project, instead of only at the end. Besides that, we are also considering gaining information on the design processes followed by other international researchers. Moreover, within the focus of open-ended play we are looking into designing for specific forms of play (e.g. fantasy or competitive play) and a variety of age groups to explore age issues further. Eventually, this will lead to an extension of our knowledge and applicable results as design tools or methods. We hope this paper opens up a discussion on designing for open-ended play and inspires other design researchers to, together with us, further develop this area and sharpen its design approach and definition.

7. ACKNOWLEDGMENTS The authors would like to thank the students Chris Gruijters, Stephan Hoes, Gijs Houdijk, Martijn Kors, Patrick Leijte and Koen Verbruggen for their work and reflections and our colleagues for their input and feedback in discussions on open-ended play. This research was funded within the Creative Industry Scientific Programme (CRISP). CRISP is supported by the Dutch Ministry of Education, Culture and Science.

8. REFERENCES [1] Bateman, C. 2005. The anarchy of paidia.

http://onlyagame.typepad.com/only_a_game/2005/12/the_anarchy_of__1.html.

[2] Bekker, T., Hummels, C., Nemeth, S. and Mendels, P. 2010. Redefining toys, games and entertainment products by teaching about playful interactions. International Journal Arts and Technology. 3, 1, 17-35.

[3] Bekker, M., Sturm, J., and Eggen, B. 2009. Designing playful interactions for social interaction and physical play. Personal and Ubiquitous Computing. 14, 5, 385-396. DOI= http://dx.doi.org/10.1007/s00779-009-0264-1.

[4] Bruce, T. 2001. Learning through play: babies, toddlers and the foundation years. David Fulton Publishers, London.

[5] Caillois, R. 1961. Man, Play and Games. Free Press of Glencoe, New York.

[6] Carse, J.P. 1987. Finite and infinite games: A vision of life as play and possibility. Ballantine Books.

[7] Costello, B. M. and Edmonds, E.A. 2009. Directed and emergent play. In Proceedings of the 7th ACM conference on Creativity and Cognition (Berkeley, California, USA, October 26-30, 2009). C&C’09. ACM, New York, NY, 107-116. DOI=http://dx.doi.org/10.1145/1640233.1640252.

[8] Creighton, E. 2010. jogo: an explorative design for free play. In Proceedings of the 9th International Conference on Interaction Design and Children (Barcelona, Spain, June 9-12, 2010). IDC’10. ACM, New York, NY, 178-181. DOI= http://dx.doi.org/10.1145/1810543.1810565.

[9] Fuller, D. and Magerko, B. 2011. Shared mental models in improvisational theatre. In Proceedings of the 8th ACM conference on Creativity and Cognition (Atlanta, Georgia, USA, November 3-6, 2011). C&C’11. ACM Press, New York, NY, 269-278. DOI=http://dx.doi.org/10.1145/2069618.2069663.

[10] Fullerton, T., Swain, C. and Hoffman, S. 2004. Game Design Workshop: Designing, Prototyping and Playtesting Games. CMP Books, San Francisco.

[11] Gray, P. 2009. Play as a foundation for hunter-gatherer social existence. American Journal of Play. 1, 4, 476-522.

[12] Hewett, Baecker, Card, Carey, Gasen, Mantei, Perlman, Strong and Verplank. 1996. ACM SIGCHI Curricula for Human-Computer Interaction. Chapter 2: Human-Computer Interaction. Retrieved in January 2013. http://sigchi.org/cdg/cdg2.html.

[13] Hopma, E., Sturm J., and Bekker, T. 2009. Interactive play objects: the influence of multimodal output on open-ended play. In Proceedings of 3rd International Conference on Intelligent Technologies for Interactive Entertainment (Amsterdam, The Netherlands, June 22-24, 2009). Intetain‟09. Springer, Berlin, 78-89. DOI= http://dx.doi.org/10.1007/978-3-642-02315-6_8.

[14] Huizinga, J. 1955. Homo Ludens: A Study of the Play Element in Culture. Beacon Press, Boston.

[15] Iguchi, K. and Inakage, M. 2006. Morel: remotely launchable outdoor playthings. In Proceedings of the 2006 ACM SIGCHI international conference on Advances in computer entertainment technology (Hollywood, California, USA, June 14-16, 2006). ACE’06. ACM, New York , NY. DOI= http://dx.doi.org/10.1145/1178823.1178866.

[16] Morrison, A., Viller, S., Mitchell, P. 2011. Building sensitizing terms to understand free-play in open-ended interactive art environments. In Proceedings of SIGCHI Conference of Human Factors in Computing Systems (Vancouver, Canada, 7-12 May, 2011). CHI’11. ACM, New York, NY, 2335-2344. DOI=http://dx.doi.org/10.1145/1978942.1979285.

[17] Nachmanovitch, S. 1990. Free play: improvisation in life and art. J.P. Tarcher, Los Angeles.

[18] Rogers. Y., Sharp, H. and Preece, J. 2011. Interaction design: beyond human-computer interaction. 3rd edition. John Wiley & Sons, Chichester, UK.

[19] Rosales, A., Arroyo, E., and Blat, J. 2011. FeetUp: A Playful Accessory to Practice Social Skills through Free-Play Experiences. In Proceedings of the 13th IFIP TC13 Conference on Human-Computer Interaction (Lisbon, Portugal, September 5-9, 2011). INTERACT’11. Springer, Berlin, 37-44. DOI= http://dx.doi.org/10.1007/978-3-642-23765-2_3.

[20] Rosales, A., Arroyo, E. and Blat, J. 2011. Evocative experiences in the design of objects to encourage free-play. In Proceedings of AMI 2011 Workshops (Amsterdam, The

100


Netherlands, November 16-18, 2011). Springer, Berlin, 229-232. DOI= http://dx.doi.org/10.1007/978-3-642-31479-7_39.

[21] Salen, K. and Zimmerman, E. 2003. Rules of play: game design fundamentals. Cambridge, MA: The MIT Press.

[22] Santer, J., Griffiths, C. and Goodall, D. 2007. Free play in early childhood: a literature review. National Children’s Bureau, England.

[23] Seitinger, S., Sylvan, E., Zuckerman, O., Popovic, M. and Zuckerman, O. 2006. A new playground experience: going digital? In Extended Abstracts on Human Factors in Computing Systems (Montréal, Québec, Canada, April 22-27, 2006). CHI’06. ACM, New York, NY, 303-308. DOI= http://dx.doi.org/10.1145/1125451.1125520.

[24] Valk, L. de, Rijnbout, P., Bekker, T., Eggen, B., Graaf, M. de and Schouten, B. 2012. Designing for playful experiences in open-ended intelligent play environments. In Proceedings of IADIS International Conference Game and Entertainment Technologies (Lisbon, Portugal, 18-20 July, 2012). GET 2012. 3-10.

[25] Veitch, J., Salmon, J., Ball, K. 2010. Individual, social and physical environmental correlates of children’s active free play: a cross-sectional study. International Journal of Behavioral Nutrition and Physical Activity. 7, 11. DOI= http://dx.doi.org/10.1186/1479-5868-7-11.

[26] Vygotsky, L.S. 1967. Play and its role in the mental development of the child. Soviet Psychology. 5, 3, 6-18. http://www.marxists.org/archive/vygotsky/works/1933/play.htm.

101


leaving room for improvisation: towards a design approach ... · pdf fileleaving room for...

Documents