A Case Study of Exploding Places, a Mobile ... - ACM Digital Library

3 downloads 0 Views 548KB Size Report
Rachel Jacobs, Matt Watkins, Robin. Shackford. Active Ingredient. 100 Mansfield Road. Nottingham NG1 1HD. {rachel, matt, robin}@active-ingredient.co.uk.
A Case Study of Exploding Places, a Mobile Locationbased Game Martin Flintham, Chris Greenhalgh, Tom Lodge, Alan Chamberlain, Mark Paxton

Rachel Jacobs, Matt Watkins, Robin Shackford

Horizon Digital Economy Research University of Nottingham Nottingham, NG7 2TU

Active Ingredient 100 Mansfield Road Nottingham NG1 1HD

{mdf, cmg, txl, azc, mcp}@cs.nott.ac.uk

{rachel, matt, robin}@active-ingredient.co.uk

ABSTRACT In this paper we present a case study of a new mobile locationbased game based on the creation of virtual communities against a backdrop of historical content. We describe the design, implementation and public pilot of the game, and present a quantitative and qualitative analysis of the use of the game’s features. We discuss how and why the combination of these features affected players’ engagement with the game. We conclude by reflecting on the design tensions resulting from the artists’ ‘evocative’ approach.

Categories and Subject Descriptors H.5.m [Information Interfaces and Presentation (e.g., HCI)]: Miscellaneous; J.5 [Arts and Humanities]: performing arts.

General Terms Design, Human Factors.

Keywords Mobile, location-based, game, historical, time-line, community, engagement

1. INTRODUCTION A broad range of mobile interactive and location-based experiences have been developed, ranging from simple games to artistic performances and cultural or historical experiences. In many cases these target pedestrians exploring a city whilst engaging with a mobile device, and together they form a growing and significant area of research. Some of these experiences have taken the form of pervasive games that involve direct translations of existing game mechanics to more pervasive structures. Can You See Me Now? [1] connects players in an online recreation of a physical space with professional performers in the real city in a fast-paced game of tag, while Day of the Figurines [5] presents a text adventure game that unfolds over a period of a month in an attempt to be interwoven into the everyday activities of its players. Such games also include commercial offerings for mobile phones (such as Bot 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. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Short presentation, ACE’2011 – Lisbon, Portugal Copyright 2011ACM 978-1-4503-0827-4/11/11

Fighters! [10]) and increasingly for modern smart phones [9]. Other location-based experiences present participants with the opportunity to engage with a location-based story as they explore the city, either as part of an interactive art performance such as Uncle Roy All Around You [2], or a growing collection of personal stories and anecdotes scattered throughout the city by previous participants. Finally, there are a number of experiences that present a cultural or historical tour of a city space, ranging from browser-like interactions with a context aware mobile system such as Guide [3], to rich media-based interactions in applications such as Riot! [8] Many of these experiences have incorporated novel designs, and novel forms of interaction and game play. Complementing these, a significant amount of research has focused on creating supporting tools for artists, designers and non-technical domain experts to understand and engage with this emerging field, including tools such as MediaScapes [7] which fostered a significant community of users. In this paper we present a case study of creating and deploying a new mobile experience that specifically addresses the challenges of designing a meaningful experience that combines elements of historical narrative, game mechanics and user generated content. We begin by describing the game and the way that it combines these elements, before analyzing its pilot deployment.

2. THE GAME Exploding Places (a multiple reference to the pilot venue, Woolwich and its historic arsenal, and a virtual community game mechanic), explores how people define their own urban landscape in terms of territory, history and community, using a multiplayer mobile location-based game. We present an overview of the players’ experience and describe the three mechanics of the game in more detail. Up to 25 players at a time are booked in for a specific “performance” of the game, which lasts for about one hour. The participants for each game gather at the hosting venue, where they are each loaned a mobile phone – in the pilot, a Nexus One Android phone, running a custom game client – and are given a short introduction to the game by a performer in historical costume. The group is led outside into a nearby square and the game begins. Players are instructed to explore the city streets, and play the game for the following hour. As the players roam the streets of the area, the phone displays a map (see figure 1) on which each player creates their own virtual community of residents, using their physical location to populate the local area with new community members. The mobile phone indicates the passing years, and gives glimpses of historical events that affected the area. These events are also revealed through their effect on the new community, leading to births and deaths of

community members. These effects plus interactions with other players’ communities encourage players to move and expand their own community.

The War is Over! Communities across Woolwich celebrate as they hear the news that the war is over! All communities gain 2 health, wealth and participation points. A flash flood in Plumstead.The ravene lake on Plumstead Common has flooded and overflowed in torrents down into Roydene Road flooding many houses. In 3 hours lots of damage has been caused to houses in the area. All community members in the area lose 1 wealth point.

Figure 1: The main phone client screen At the end of the hour players return to the hosting venue, where a projected display has also been revealing their virtual communities and the communities' interactions to spectators. Each player is given a recreation of a vintage photograph depicting their fledgling community.

2.1 Historical Narrative Exploding Places allows players to engage with 120 years of Woolwich history, with a year in the game passing every 30 seconds. This historical narrative is coarsely location-based, and players are notified of changing events based on the location of their growing virtual communities. The city space in which the game is played is divided into 8 virtual zones, which reflect long-standing and recognizable boundaries or regions within the physical city. As players move through the city, the phone notifies them when they enter a new zone in which they may experience different events. As time passes, players receive messages on their phone describing events that have occurred in the current zone that they occupy, or in the zones in which members of their virtual communities reside. Opening the message reveals a few sentences describing the event in more detail, as shown in figure 2(a), and also the effect it has had on nearby community members (described in the following section). The historical timeline includes a “tag” event for each year, which is used to specify the current game year. Moving these events also gives the ability to compress or stretch game-time while authoring. The main content takes the form of over two-hundred hand-authored descriptions of historical events that the players can experience. These events range from events of national significance that affected the area to events that were local to the area or linked to it, for example (see also figure 2(a)):

Figure 2: (a) Viewing a typical event (b) Adding a new “story” Events that affected the whole area are experienced by all players regardless of their location, whereas events that have specific geographical relevance are tied to one or more virtual zones, and accordingly are only experienced by players who have community members in those areas. In this way, players are exposed to a rich and diverse historical narrative of the local area from 1890 to 2010 – and this narrative also informs the main game mechanic, as described next

2.2 Communities and Members Over the course of the game players create and nurture a virtual community. At the start of the game the players’ devices instruct them to create a “founding” community member – representative of a new immigrant settling in the area for the first time. The device displays an interface that allows the player to set four starting attributes for the founding member by manipulating the limbs of a stylised avatar, as shown in figure 3(a). The player is given a quota of points to spend on the member's attributes (health, wealth, knowledge and participation), introducing a simple resource currency that will affect how the member interacts with unfolding historical events and other community members. Finally the player is invited to “place” the newly created member in the city, by physically moving to an area in which they would like their community to begin. Once placed, the member begins to experience the ongoing historical events, and in particular has their attributes affected by them. Historical events such as the war, that would have changed the health of local residents, have a detrimental effect on the member's health, while changing economic conditions affect the member's wealth, and so on. A basic set of game rules determines how the player's community grows and evolves in response to these events. If a member's health and wealth reach a certain threshold value then a new

community member is “born” nearby, with similar attributes to its parent. If a member's health falls to zero then the member dies (the game prevents the founding member from being killed to ensure that the player can remain in the game for the full hour).

game time as it was created so that it can be experienced by players of subsequent games.

3. IMPLEMENTATION Exploding Places is implemented using a conventional clientserver architecture, and makes use of a custom mobile phone client and server web application. We now give a brief overview of the system and the supporting authoring and orchestration tools used to deploy the game.

3.1 The Mobile Client Players primarily interact with the game using an Android phone. The client is an Android application running full screen in order to partially “kiosk” the application to discourage players from accessing other phone functions.

Figure 3: (a) Creating a community member (b) my community view The attributes of a player's community as a whole also affect the player's abilities within the game. If the combined health, wealth and knowledge of the community exceeds a particular threshold then the player is allowed to create new members in order to expand their community. If the combined participation attributes of the community are great enough then the player is allowed to add their own historical events for future players to experience. The growing community is shown in their “community” screen (figure 3(b)), and is also displayed on the main device screen (figure 1) together with the communities of other players, revealing how each community occupies a particular territory. Even when the player has physically moved to a different area or zone they continue to receive messages about events that affect their members and notifications of births and deaths. A player can also pick up a community member and move it to a new location. In this way, the player can micro-manage the growth of their community in response to unfolding events.

The phone client initially displays a splash screen while the players are being shepherded outside into their starting positions. The main screen (figure 1) embeds a Google Maps component, usually centered on the player’s current location and displaying a live view of community members using custom overlays. Onscreen buttons (when enabled) take the player to further screens allowing them to create new members (figure 3(a)), add their own stories (figure 2(b)) or inspect their community members (figure 3(b)). Notifications of changes in community members are displayed as “toasts” (figure 1, bottom of map) with a sound notification. The most recent historical event is visible as a message overlaid on the map (figure 1, top left of map), which allows the player to tap through to a further screen to view the event in full (figure 2(a)).

3.2 The Game Engine

2.3 Multiplayer Interaction The game employs a number of simple rules that allow communities belonging to different players to interact If a community member is the first to be placed in a particular zone then the member is rewarded with an increase in knowledge and participation attributes. Similarly, if the community grows so that it has more than a certain number of members in a particular zone then they are also rewarded with an increase in attributes. If, however, members from different communities are in close proximity to one another, then the relative values of attributes determine whether the weaker community member is integrated, or “assimilated”, into the stronger community, becoming a member for the other player. While the game does not have an explicit winning condition, some players may choose to engage with the relative size of their community as a primary element of their game play. Finally, as mentioned in the previous section, depending on their attributes, players are able to add their own events or stories to the game. As shown in figure 2(b) the interface for “adding a story” includes a text description and attribute modifiers (like any other event). Once submitted, and after moderation by the game’s operators, this new event is inserted into the timeline at the same

Figure 4. The authoring tool The game engine is implemented as a web application built using the Equip2 middleware platform. This is hosted within an Apache Tomcat installation backed by MySQL on a rented virtual private server. The majority of the game logic – the rules – is implemented within the game engine. The web application exposes a largely generic http interface with which the mobile phone application communicates. Each time the mobile application polls for changes, or reports a change, it is given a list of events that have occurred within the game state since it last connected. This interaction includes application-level retransmission, prioritisation and flow-control to cope with slow

and unreliable networks. Using Equip2’s eventing system, a custom engine controller is notified of any significant changes to players or communities and re-checks the game rules. The game rules are hardcoded in java, although key variables and thresholds are configurable (this includes the ability to turn specific functionality off and on).

possible, for example if a player tries to interact with a community member who has just “died” on the server). Table 1. Total uses of main game mechanisms Mechanism

Total Attempts

% of total

Successful

Success rate

A dedicated HTML5 authoring tool (figure 4) allows the artists to construct and populate the historical timeline of events, while a set of basic orchestration pages allow the artists to create, start and monitor games. Finally, a game overview visualization revealing the complete map and communities, implemented using Flash, is be used to provide a “spectator” view on a large public screen.

Messages viewed

367

36%

367

100%

Members created

534

52%

449

84%

Members placed

208

20%

169

81%

4. STUDYING THE PILOT

Members picked up

31

3%

23

74%

Stories added

18

2%

12

67%

Total

1158

1020

88%

Exploding Places was publicly piloted in July 2011, in Woolwich. Four games were run during the day, starting at about 10:30 and then at ninety minute intervals. In total 25 mobile phones were used, between 5 and 8 in each game. Many phones were used by individuals, but some were also shared by couples and family groups; in total about 35 members of the public played the game. The client application on the mobile phone logged network operations, moving between screens within the application, some GUI interactions (e.g. pressing main UI buttons), game operations (e.g. creating a community member) and game state updates (e.g. changes to community members). The game server independently logged all changes to the server game state, i.e. the server database. The client and server logs were retrieved after the event and archived. Data was analysed and visualised using bespoke log parsing applications and GNU R. Members of the team made a photo record of the locations, activities and artefacts of the day. Four players were accompanied and video-recorded (with their consent) while playing the game. They were informally prompted to describe what they were doing. Three of these recordings were transcribed in full for analysis. In all three cases the player was an adult who played on their own (i.e. did not play as part of a group), and was familiar with the city but not with the district in which the game was played. Two of these players were female (“A” & “C”) and one was male (“B”). Informal exit interviews were also conducted with a number of players, and a simple exit questionnaire was offered to all players. Several iterations of team reflection and discussion were also conducted and documented.

5. QUANTITATIVE OBSERVATIONS 5.1 Overall Use of Game Features As previously described, the game incorporates three main mechanisms for participation: viewing messages, creating and moving community members and contributing stories (usergenerated content). Considering all four games, table 1 summarises the total use of these features for all players. We can see that viewing messages comprises about one third of player game actions, while the vast majority of other actions are creating and then to a lesser extent placing community members. Relatively few stories are added, and it is also relatively uncommon for players to pick up community members that they have already placed. In general, the success rate for all actions is reasonable, and comparable to network reliability observed (game operations can fail either because the client is unable to communicate with the server at the time, or in some cases because concurrent activity on the server means that the action is no longer

5.2 Inter-player Variation We now consider the variation in use of game mechanisms by different players. Figure 5 shows the distribution of number of messages explicitly viewed (clicked on) per player. One player does not view any messages at all (although they receive more than 100 messages). All other players view at least one message, while the median is eleven messages. The largest number viewed is 42. A comparison of the number of messages received and the number viewed reveals no significant correlation, i.e. receiving more or fewer messages does not significantly affect the number viewed. Every player successfully creates at least two community members, while one player creates 92 community members during their one hour game. The median number created is 14 and three quarters of players create between two and twenty community members. From the data in table 1 at most 38% of created community members are actually placed. Again, every player places at least one community member, but no player places more than thirteen; the median is six. From table 1 we have seen that relatively few attempts were made to pick up a community member. In fact, ten players did not attempt to pick up any community members, nine players attempted this exactly once, while the remaining six players did so between two and six times. These attempts were generally successful. Most players did not attempt to add a story, or perhaps did not reach a stage in the game that allowed them to add a story. In total nine players attempted to author a story, and eight of these were successful at least once. Five players added one story, two added two and one added three stories. Most attempts to add stories were successful, but one player made six attempts to add a story (the largest number of attempts of any player) of which only the last one succeeded. Comparing each player’s use of these game mechanisms reveals little correlation or apparent conflict. In other words there is little evidence that – at least in this first playing – players are specializing their engagement with the game. Rather, all players engage to some extent with both messages and community members, while the minority of players also submitting stories are not distinguished by their use of other the game mechanisms.

6

while the maximum was about 800m. We compared these two measures for each player, i.e. total distance walked and maximum distance from the starting point. In general, players who walk further get further from the starting point, but this is not always true. For example, the player walking the greatest distance actually stayed within 250m of the starting point at all times.

5 4 Frequenc y 3

8

2 6

1

Frequenc y 4

0 0

1 2 0 Number 0 of viewed

3 0 messages

4 0

5 0

2

Figure 4: Numbers of messages viewed by each player.

5.3 Details of Message Viewing In this section we look in more detail at players’ viewing of messages. Our measure of “viewing” a message is actually the visibility of the message detail screen; we did not measure user gaze of other indicators of specific attention. The time between a message being presented to a player (on the main map view) and the previous message had no significant effect on whether the player viewed the new message. This suggests that players did not anticipate the arrival of new messages. However, the time until the next message arrived did have a strong effect on the likelihood of the message being viewed. In other words, the more time that a message was visible on the map screen, the more likely it was to be viewed. There was no clear pattern of message viewing over the course of the game as a whole. The majority of times for which a message was viewed (i.e. the message view screen was visible) were between 10s and 30s, median 16s. The minimum time was 1s and the maximum was almost 3 minutes. On average messages were 260 characters long. There was no significant correlation between the length of the messages and the time spent viewing them. There was substantial variation between players in the lengths of time for which they viewed messages. Players who viewed more messages tended to view messages for a greater time in total. Correspondingly, how long a player spent viewing a single message did not vary significantly with the total number of messages that they viewed.

5.4 Overall Patterns of Movement The game client logs the user’s position as reported by the in-built GPS received throughout the game. Figure 6 shows the distribution of the total distances walked by the players. Most players walk between 1.5km and 3km with a median of 2.6km. The shortest distance travelled is about 100m and longest about 5.5km. It is possible that random variations in the reported GPS position may cause these measures to slightly overestimate the actual distance travelled. We also calculated the maximum straight-line distance reached by each player from the starting point, i.e. how far they travelled “as the crow flies”. Most players strayed at most between 200m and 600m from the starting point, with a median of about 400m. One player never travelled more than 50m from the starting point

0

0

1000

2000

3000

4000

5000

6000

Total distance travelled

Figure 5: Total distance walked by each player.

5.5 Temporal Patterns of Activity Binning system and user activity into five minute bands starting at the beginning of each game reveals some simple “shape” to the experience. There are about twice as many authored game messages in the first halves of the games as in the second halves, and this is reflected in the pattern of message viewing, i.e. about twice as many messages are viewed in the first halves of the games as in the second halves. No stories are added in the first half hour of the games, which reflects at least in part the game rules which do not allow players to add stories until their communities meet certain criteria. But after this time stories may be added at any point. The distance that players walk peaks around 10 minutes into the game, and tails off gradually by about one third over the rest of the game. The maximum distance “as the crow flies” that players get increases fairly uniformly in the first half hour of the game and then gradually tails off, dropping more sharply at the end.

6. QUALITATIVE OBSERVATIONS Recordings from the three players mentioned were considered with particular reference to accountable and observable issues of playing and engaging with the game and its various aspects. These observations are presented thematically and grounded in relevant fragments of talk including elements from post-game interviews.

6.1 Goals of the Game In terms of their goals within the game, player B engages with the game, at least to some extent, in terms of competition, specifically to have as many members as possible: “Well I'm trying to create as many as possible”, “We've still got some time left and maybe it'll get me that much closer to colonising the whole of East London”. However he also articulates a degree of self-regulation of this orientation: “I don't want to get too competitive”. Player A, while less explicit, also appeals to the notion of “winning”: “…so I might not win…”. Player C, however, articulates confusion with regard to the (understood) goals and objectives in the game: “A

little confused with the game, I don’t know if I’ve really got the gist of it”, “I have no idea if I’m doing well or not, (laughs) or badly, probably bad”.

6.2 Deciding Where to Go Much of the discussion of where to go revolves around the placement of community members, which is considered in more detail below. However other values and concerns are also expressed in the observed talk. C appeals to her lack of local knowledge as a disadvantage: “It’s quite difficult when you don’t know, you see I don’t know Woolwich at all, so I’m wandering around without knowing where the different areas are, so its sort of (gestures)”. At different points A and B (who also lack local knowledge) also report that they are lost, at least with reference to the map: “I’m not actually too sure where I am now” (A), “totally lost myself on the map though” (B). A refers to “excitement” as one of her guiding principles: “I’m just going to have a look at the map to see where’s exciting that we can go now”. Player A also explicitly refers to the progression of the virtual time, and more specifically its anticipated ending as a cue to end the game: “we’re coming up to 2009 so we should probably finish it”. This is before the phone starts to explicitly notify players that they should return to the starting location. At least on one occasion, C would like to orient her activity in relation to other player(s) of the game: “it’d be nice to see another player….I can’t see any at the moment”.

6.3 Messages All of the observed players comment on the viewed messages on at least a few occasions, and are generally positive about them: “its been interesting finding out about Woolwich” (A), “[I] really like reading about all the things that happened” (C). On at least on some occasions they prompt sharing: “Ha, bet you didn’t know that!” (A, to ethnographer). B reflects more than the others on the link between the messages and framing narrative: “this is quite personal this information so it does, it's not just like playing a game, it's about a real family and real history…so….it's quite a different experience from watching the news…you’re actually paying attention to a person’s biography.” The link between messages and community members is considered in next section.

6.4 Community Members 6.4.1 Reportability and Affect All of the observed players announce changes that they notice in community members. These include changes linked to specific events such as: “Ooh I got 4 health points” (A). The birth or death of community members is particularly newsworthy and also linked to statements of affect: “there’s been a birth in my community, great” (C), “there’s been another birth, that I’m happy about” (C), “[I liked] the huge boom in my population” (A), “Well. Oh dear! There's been a death, of one of the community members” (B), “The highlight was surely when the characters started multiplying themselves, that was really thrilling to see that happen.” (B, interview). The observed players also report positively on the general state of community members: “I'm thriving I don't know how the other players are going but I have a lot of, a lot of characters” (B), “yay all my family members are strong” (A). In the first case the B uses

the first person to talk about his community, while A characterises them as her family.

6.4.2 Monitoring Communities Player A comments specifically on active monitoring of her communities: “just going to check on the health of my community”, “my community over there looks as if it has issues”. Perhaps related to this distinctive concern, A is also the only player to comment on picking up and moving a community member: “my little dude hasn’t started multiplying up there, I wonder whether I should pick him up and move him?”.

6.4.3 Placing Community Members As already noted, the majority of talk about where to go is framed by the decision of where to place community members. This leads to one of the richest themes in the observations which is the range of strategies and resources that players employ (or at least refer to) in making this decision. The map itself (a standard Google map view) is used in an attempt to form decisions: “trying to see, sort of where, where I can go really, where looks good” (C). However, in the absence of other community members (e.g. at the start of the game), the map provides little or no information explicitly linked to the game: “I knew there were 5/6 zones but didn’t know where they were…didn’t know where I was meant to be going to find them” (A, interview). The relative position(s) of other community members on the map is the most often cited resource for placing new community members. In particular, players typically report that they are looking for uninhabited areas, for example: “I think I need to populate another area, so I’m going to head towards I guess, I think the river’s that way…” (C), “I’m trying to find somewhere that’s uninhabited by other people, there’s a big community there” (A). In some cases finer geographical distinctions are used: “I’m going to go down there into that little nick between everybody and I’m going to try and place some people there.” (A). On one occasion reference is made to the possibility tactical placement to try to assimilate another community: “I think I’m going to go just outside of where all their little dudes are, and put my next guy, not that I’m going to try and take over at all”. (A). Player A in particular appeals on several occasions to the character (in particular the apparent affluence) of the local area as a guiding principle: “when it comes to the 2000s they’re going to really do well because they have all of this wealth, affluence around them”, “we’ve got flashy apartments ahead, so as we get into the future it’s going to be a prosperous area”.

6.5 User-generated content Player B notes at one point that they are able to add a story: “I could add a story if I wanted to.” But they do not follow up on this option. Player C makes no reference to adding a story at any point. However player A makes two consecutive attempts to add a story. She begins by reflecting on the opportunity: “I wonder if I have enough points to add a story and what kind of story I would add?”. She begins to author a story about a closed post office: “I’m writing a little story about this area to say that the post office is closed…you can’t post your mail, so your going to lose points because you’ll go into debt..cos you can’t pay your rent”. But she then accidentally presses the “back” button and loses her uncompleted work: “Oh no, where’s my story?” “I pressed back and I had a fat finger that made it disappear from my screen”.

She immediately plans another story: “Ok I’m going to stand out the Weatherspoons and make a story”. She actively plans the effects of the story, apparently before designing the text to fit: “and I’m going to do the attributes before, I’m going to take the health down and the wealth down, and give them lots of knowledge and participation, I’m going to say they drunk and disorderly round here, so they loose money and they lose wealth, but they participate with the local community, they participate with the local police”. She frames her contribution of the story specifically in terms of its anticipated effect on subsequent players: “hopefully this time it might work, excellent, so I might not win, but some poor person is going to have an awful time when they come around here next time”.

7. DISCUSSION In this section we combine and reflect on the observations of each of the game mechanisms in turn and finally their interactions.

7.1 Timeline Messages We have seen from the quantitative results that all players engaged at least to some extent with the messages delivered to them. From the qualitative observations, the observed players reported some positive responses to the messages, in terms of interest, discovery or surprise. Player B also refers explicitly to the narrative link of (a subset of) the messages. The sense of time and history is also reflected by player A (in relation to placing members) and more generally by player B: “inevitably it's going to cover 20th century history which, is always difficult”. However there was also significant variation between players, with some viewing so few messages that they appear to have discounted them in terms of involvement with the game. Messages typically arrive about every 30 seconds, and even the player who views the most messages only looks at about 40% of them. The specific analysis of message viewing suggests that individual messages may be chosen for viewing also almost at random. Anecdotally at least one of the players reported finding the rate of new messages to be almost overwhelming at times. Taken together these observations provides some support for this use of messages and narrative (independent of community members), but also suggests that a subset of players may not engage with this aspect of the experience. It also raises issues about the pacing or consumption of these messages, in that no player has seen even a majority of the authored messages. In part this might be responded by the way(s) in which messages are presented to the user and accessed, for example the current version does not allow players to view old messages. But it also highlights the tension inherent in a game element that players can effectively “take or leave”, both as a whole and also from moment to moment.

7.2 Community Members From the quantitative results we find that all players engage to some extent with the creation and placing of community members, although again there is substantial variation in the apparent extent to which they do so. All of the players in the qualitative observations report positive emotion in relation to positive changes in community members, especially the “birth” of new community members and the growth of communities in general. Player B in particular also reports a strong negative emotional response to the “death” of a community member. The observed players’ talk is also dominated (at least when the game is working) by communities and community members.

This suggests that the community member mechanism may be the most effective element in establishing player involvement. We observe from the both the quantitative and qualitative observations great diversity in the ways in which different players engage with this game element. As with messages, a small number of players make use of the related game action so rarely that one wonders if they evaluate and then dismiss this aspect of the experience. On the other hand, considering players who do engage more with this aspect of the game there appears to be considerable variation in the ways that they engage with it. From the qualitative observation we saw considerable thought and diverse resources being used to decide where to place community members, perhaps to the extent that this dominated players’ understanding of playing the game. Players articulated these decisions in terms of: the map itself; other community members viewed as competition or as a clue to the hospitableness of an area; the visible character of the physical locale (e.g. apparent affluence); the anticipated timeline and events of the game; and the publicly displayed overview of the game. This was in addition to more general consideration of where to go that made reference to: local knowledge; excitement; other players; the ending the game; and challenges of device interaction.

7.3 User-generated Content From the quantitative results we find that only a minority of players contribute (or try to contribute) their own story to the game. Most players who do so contribute only a single story (or two or three at the most). Similarly in the qualitative observation we see only player A contributing a story. The ability to contribute a story was dependent on a player’s communities acquiring certain characteristics over the course of the game. Some players may never have reached this threshold and therefore never been able to contribute. Other players who did satisfy this requirement may not have noticed that the relevant button in the UI was now enabled where previously it had been disabled. Contributing a story was by far the most complex activity in the game, in particular requiring the user to type a relatively extended text using the phone’s on-screen keyboard; this may have discouraged potential contributors. Finally, the potential impact of contributing a story lies outside the current game, as observed by player A, and this may have undermined the immediacy or significance of this part of the game.

7.4 Relationship Between Game Mechanisms In the qualitative observations we saw that players note the effects that events (seen as messages) have on their community members. In this way the inclusion of community members augments (at least some of the time) the subjective significance of the messages and the history that they describe. From the qualitative observations we also saw that players invoke anticipated events (e.g. known future history) as part of the process of placing community members. So in that way the historical backdrop of the game, as expressed through the messages, gives significance or shape to the community mechanism. In terms of the ability to contribute stories we would expect that the prior exposure to messages (and their potential effects on community members) serves as an essential but implicit tutorial on the structure and potential significance of “stories” as authored in the game. It is also clear from player A that their prospective delivery to future players as events (messages and effects) is an essential part of how they are viewed. In this way we can trace the mutually supportive relationships that exist between the three

mechanisms, and we would suggest that messages (events) and community together should be regarded as mutually indispensible elements of the experience.

8. CONCLUSIONS We conclude by reflecting generally on how the combination of the three mechanisms in the game and their presentation can lead to somewhat ambiguous engagement by players, to both positive and negative effect. The concept of design, experiences and games that are ‘evocative rather than didactic’ [6] is considered by the artists to be at the core of their practice. The members of the group recognize that the ‘ambiguity’ of this approach can be in conflict with the development of urban games per se. For example, it is an aspiration for much of the artists’ work that individual “players” can choose how to engage with a “piece”, for example competitively or reflectively.

observe players appealing to quite fine-grained features of the physical environment, such as the market, a particular post office, some expensive-looking apartments or a patch of waste ground. Positively, the game’s ambiguity (i.e. the players uncertainty about exactly what “matters” in the game) allows the players to “flesh out” their subjective experience of the game using the “local colour” of the world around them. In some ways this is an ideal characteristic for a location-based game. Negatively, where player’s working assumptions are in error, the game’s failure to take these manifest features into account may lead to confusion and frustration or imply that the game is simply not working

9. ACKNOWLEDGMENTS We gratefully acknowledge the support of RCUK through grant EP/G065802/1 Horizon Digital Economy Research. Exploding Places was commissioned by Stream and funded by Arts Council and Greenwich Council.

We can trace some of these concerns and tensions in the game and pilot presented here. For example the game does not have a specific winning criterion. From the quantitative observations the great diversity of user behaviours may reflect this, as each player has to decide to some extent for themselves how to orient to the game and its “meaning”. In the qualitative observations we saw two players turning to an ideal of competition, specifically measured by number of community members. In some respects this openness is valuable in giving the players this space of choice and value. On the other hand player C’s manifest confusion about the “gist” of the game and her uncertainty about how to assess how “well” she was doing might easily undermine her total experience.

10. REFERENCES

In reviewing the use of the three main game mechanisms we also saw that in some sense none of them was fully engaged with by any player. For example, only a minority of messages were ever viewed, community members were rarely moved and stories were seldom contributed. The very way in which players are allowed to choose their own approach to engagement to the game might compromise their commitment to every element of it. In an analogous way, the designers and developers of the game must wrestle with potentially contradictory requirements, e.g. to give each element significance and yet allow an individual player to set it aside, or to provide a holistic and integrated experience (within one mechanism, and also across all mechanisms) and yet also have elements of it stand in isolation.

[4] Debord, G. 1995. The Society of the Spectacle. Trans. D. Nicholson-Smith. New York: Zone Books. (Original: 1967. La société du spectacle. Paris: Buchet-Chastel).

Finally, in reviewing the ways in which players engage with the game and especially the placement of community members we see some of the things that players assume to be significant in the game. In some cases these assumptions are correct, but in other cases they are not. A particular point of ambiguity in this game is its spatial granularity. There are six main playing zones in the game, defined as far as possible with distinct characteristics and (local) identities. Community members interact with other members within each zone, but not between zones, irrespective of physical distance. The zones are also relatively large, at least in terms of the distances that players walk. On the other hand we

[1] Benford, S., et al., Can you see me now? ACM TOCHI, 2003. 13(1): p. 100-133 [2] Benford, S., Flintham, M., et al, Uncle Roy All Around You: Implicating the City in a Location-Based Performance. In Proc. Advances in Computer Entertainment (ACE 2004), ACM Press. [3] Cheverst, K., Davies, N., Mitchell, K., Friday, A., Efstratiou, C., Developing a context-aware electronic tourist guide: some issues and experiences, Proceedings of the SIGCHI conference on Human factors in computing systems, p.17-24, April 01-06, 2000

[5] Flintham, M., et al "Day of the Figurines: A Slow NarrativeDriven Game for Mobile Phones Using Text Messaging". 4th Int. Symposium on Pervasive Gaming Applications (PerGames 2007), June 11 - 12, Salzburg, Austria, [6] Gaver, W., Beaver, J., Benford, S., Ambiguity as a resource for design, Proceedings of the SIGCHI conference on Human factors in computing systems, April 05-10, 2003, Ft. Lauderdale, Florida, USA [7] Hull, R., B. Clayton, and T. Melamad. Rapid Authoring of Mediascapes. in UbiComp. 2004 [8] Reid, J., Hull, R., Cater, K., Clayton, B. "Riot! 1831: The design of a location based audio drama." Proc of UK-UbiNet 2004. [9] Shadow Cities, http://www.shadowcities.com [10] Sotamaa O.: All the World's a Botfighters Stage: Notes on Location-based Multi-user Gaming. Proceedings of computer games and digital cultures conference. Tampere University Press, 2002

Suggest Documents