Paper: Art of Defense: A Collaborative Handheld Augmented Reality Board Game

Written by Dan Buckstein


Cooperating with fellow players by representing tangible objects in a virtual world!


Huynh, D., Raveendran, K., Xy, Y., Spreen, K. & MacIntyre, B. (2009). Art of Defense: a collaborative handheld augmented reality board game. Sandbox ’09: Proc. of 2009 ACM SIGGRAPH Symposium on Video Games, 135-142. doi: 10.1145/1581073.1581095


The authors of this paper were interested in exploring the affordances, constraints and player experience of mobile (handheld) augmented reality (AR) games. One of the key motivators is the ability to collaborate and socialize while playing such games; thus they analyze such products as social games. To explore all this, the authors created a game called Art of Defense (AoD), a tangible AR game that requires the players to interact with board game elements in the real world to drive the virtual simulation. The authors first discuss the technical details behind the game, the design considerations and, finally (and most importantly) the player experience. To evaluate player experience, the authors conducted a formal study with twelve participants. Participants played the game in pairs and were recorded on video, and observed, with a feedback questionnaire and interview following the session. From these results, the authors present their lessons learned from the participants’ experiences and discuss considerations for new tabletop AR games.


Structure-wise, the paper is very well organized and the writing really flows from part to part. That’s really all there is to say about that. Content wise, the paper provides an excellent introduction to augmented reality, its current uses in gaming, its caveats, and important considerations for AR game designers. The authors walk us through their design process and motivations. They state from the start that their intentions were to explore the experience side of tangible AR gaming.

Let’s cut to the chase; we’re not here to talk about AR gaming or considerations. If you want to know about what they did, read the paper! We’re here to discuss player experience, and that’s exactly what these authors were after in this paper: determining player experience in the domain of AR games. Although the AR content of this paper is wonderful, I have a few concerns with their evaluation of player experience, and I can also identify a consideration concerning player experience specifically in mobile AR gaming.

In several collaborative AR-related papers (Billinghurst et. al, 2001a; Billinghurst et. al, 2001b; Billinghurst & Kato, 2002; Grasset et. al, 2005), the authors set the stage for multiple methods of collaborating with AR-based products. Mark Billinghurst is truly an influential contributor to the field of AR, and he has practically written a slew of gold-mine papers that frequently focus on the collaborative applications of AR and user experience in specific areas of the game. As such, there are plenty of papers that reflect on players’ claimed experiences while playing AR games. So, if you’re looking for a gold-mine of AR design considerations and stuff, there is certainly an abundance of papers to tell you about all this stuff.

Now, what I mean by ‘claimed’ experience is the reliance on a player to describe, using words, his/her experience playing a game. Questionnaires and interviews, as used in the AoD paper and many others, are good for getting a broad idea of how players feel, but we all know that this is horribly inaccurate because players may not say what they mean or tell the whole truth!

This AoD authors rightfully state: “Designing handheld AR games is challenging… because we do not yet understand how to create effective play experiences.” I agree with this statement completely, as I also have first-hand experience in the difficulty of creating an engaging mobile AR experience. I can also try to explain why the authors’ statement is so accurate. The main reason is that everybody is relying on evaluation questionnaires and interviews! The paper contributes to this list of ‘claimed experience’ papers, because, like many others, they did not attempt to read true affective measurements. Thus, the main concern I have with this paper is that there are no affective measurements. Future studies evaluating this kind of product could make use of EMG and skin conductance data to evaluate what players are really feeling. If not exact, this data will still have a much higher confidence level than simply asking players how they feel, again because most people are either not entirely honest or they simply do not know exactly how they feel. This would also be an opportunity for me to go about my evaluations in the same manner; it really is not easy to develop a good mobile AR game because we don’t know how to get players to enjoy them. Additionally, since we are on the topic of collaborative gaming, this would give us the opportunity to find out how players feel about collaboration and competition in AR games! Also, having only twelve participants does not give a broad enough view of the audience; studies of this nature should have more play sessions to acquire more data.

Another important consideration that I would like to point out is the effect of having a handheld device to display the game world. This is as opposed to some other method of delivery, such as a HMD or shared screen to display the virtual component of the AR. It takes a bit of control and concentration to constantly hold your phone or iPad in place; I bring this up because it can be a problem when dealing with a game like AoD, one that uses a tangible interface. The problem arises because players must focus on interacting with the both tangible parts and the display at the same time. This is really speculative, but I think this would have an effect on biosensor data because players have to constantly monitor that their manipulations in reality are having the desired effects in virtual space; looking back and forth can be strenuous. Thus, having a hands-free solution such as a HMD could help. That’s just my opinion on the matter, but I think this brings up a considerable point for mobile AR designers.


I thoroughly enjoyed this paper, because it involves both a high-level aspect of affective game evaluation and my current research interest, AR. The authors raise some interesting points which I find incredibly valuable since I hope to do some AR game development sometime soon. In the augmented reality domain, they chose not to focus on algorithms and tech, which is okay, but rather the evaluating gameplay elements, which bridges their content to affective gaming and player experience. In the AR domain, my opinion is they did a great job.

In the affective and experiential domain, this paper is… a good start. What really needs to happen in order to better understand the impact of AR games, especially ones with tangible interfaces, is that we need to experiment with biosensors to get real affective data and stop relying on peoples’ opinions. The paper definitely provides a good base for a game to study, but also exposes these considerable points.

Overall, a good work, it just could have used better evaluation methods.

Additional Readings & References

Billinghurst, M., Kato, H. & Poupyrev, I. (2001). Collaboration with tangible augmented reality interfaces. Proc. of HCI International, 234-241. Retrieved from

Billinghurst, M., Kato, H. & Poupyrev, I. (2001). The MagicBook: a transitional AR interface. Computers & Graphics, 25(5), 745-753. doi: 10.1016/S0097-8493(01)00117-0

Billinghurst, M. & Kato, H. (2002). Collaborative augmented reality. Communications of the ACM, 45(7), 64-70. doi: 10.1145/514236.514265

Grasset, R., Lamb, P. & Billinghurst, M. (2005). Evaluation of mixed-space collaboration. ISMAR ’05: Proc. of 4th IEEE/ACM International Symposium on Mixed and Augmented Reality, 90-99. doi: 10.1109/ISMAR.2005.30

3 thoughts on “Paper: Art of Defense: A Collaborative Handheld Augmented Reality Board Game

  1. Brittany Kondo

    I agree with your opinion on how a mobile device may not be the most appropriate choice for such collaborative game. Using the mobile device narrows the player’s viewing area of the game and almost reduces the collaborative aspect, because the player must shift their attention from viewing the game to viewing the other player. However, there are games which have a stronger social interaction component (e.g., Pictionary) and games which involve interaction, but focus more on observing player activity on the game area (e.g. Chess). I’m not sure how prominent the social interaction aspect of this game is, but I think it is still important to easily view the other player. I would also like to see whether the results of this study would change as the number of players is increased. I’m slightly surprised with the shortcoming of affective measures for evaluating AR games. Although I’m not familiar with this field, it seems like integrating affective measures into evaluations would not be a difficult study to design, especially since methodologies currently exist in the affective gaming literature.

  2. Erik Paluka

    I would argue that the requirement of using the mobile device to play the game would actually increase collaboration for two reasons. First, occlusion issues occur when more than one player is attempting to view the same game area from the same or similar vantage point. This forces interaction among the players to mitigate this problem when it occurs. Second, the handheld AR interface causes a fog-of-war effect, which increases the chance of communication between players. If one player is able to view a certain portion of the game space that is of interest to the other player, it is easier to communicate the state of that space than to physically move. This is especially true when the game is situated on a table, and players are seated.

    With the popularity, size, cost, and mobility of smart phones, handheld AR interfaces have become an appropriate AR solution. Nevertheless, I would like to see more projects that make use of steerable displays [1]. They consist of a projector mounted on a motorized platform, and do not require the user to wear or hold any device to see the augmented reality graphics.

    [1] Andrew Wilson, Hrvoje Benko, Shahram Izadi, and Otmar Hilliges. 2012. Steerable augmented reality with the beamatron.

  3. Dan Buckstein Post author

    Good insight, Erik. The complaint here is that the use of such a mobile device with such a board game might hinder biosensor data, if it is used, not necessarily any collaborative (or competitive) advantages.
    Throughout the course we have frequently seen the whole collaboration vs. competition theme, so here’s some food for thought: do you think that handheld devices would enhance only collaboration, only competition, or either? I think that with the right selection of a mobile platform for tabletop AR games, the play experience can be enhanced for both collaborative and competitive games. I think a valid concern here is the choice of the device used; in this paper the device is a small phone with a small screen. The potential for issues is likely reduced when using a device with a (slightly) larger display, such as Android- or iOS-based phones and tablets.
    In my experience, having developed a mobile AR prototype, the nature I have dealt with is competition, and in such a case players are not going to be entirely focused on communication… aside from the occasional taunt. On the other hand, we have games like AoD which focuses on the collaborative side of things. With the right setup, mobile devices can be beneficial for both types of games. We should remember, however, that the interaction that we share with both real and virtual parts can be quite versatile, thus having an effect on any biofeedback measurements that we may run for experiments in this kind of scenario.

Comments are closed.