Tangible interaction in tabletop games

Written by João Costa


Reference Information

Bakker, S., Vorstenbosch, D., van den Hoven, E., Hollemans, G., & Bergman, T. (2007, June). Tangible interaction in tabletop games: studying iconic and symbolic play pieces. In Proceedings of the international conference on Advances in computer entertainment technology (pp. 163-170). ACM.


This paper describes a study on the role of game tokens (tangible objects) in the process of learning the game mechanics of a tabletop game.

The research group focused on digital tabletop games while using two sets of objects for this study: i) a set of iconic tokens, which are representative of the meaning of the token in question; and ii) symbolic tokens, which are an abstract representation of the component of interest. They conducted a thorough study with 30 participants, in which they looked after which of these two types of tabletop tokens conveys a better understanding of their role in the game, the game itself, as well as which creates a better game experience in terms of fun.

In the end, their statistical results were deemed as not being relevant, although, from the data that was gathered from the questionnaires, the participants demonstrated a clear preference for iconic tokens, reporting that they had felt more fun while using such game pieces.


The authors begin by describing the advantages of having digital tabletop games in opposition to playing them on the traditional (and well-succeeded)  cardboard medium. The progress of computers and the advances in Human-Computer Interaction have allowed, and will further allow, to explore new ways of conveying information to us, humans, in more natural ways. One such example is the advent of digital tabletops, that combine the digital world with physical interaction. These tables can detect physical objects users interact with, making these tables an ideal candidate for playing digital board games. One other advantage of such computing devices is that it allows for multiple people to gather around it and interact with it at the same time, just like what is possible in the traditional medium.

Following this brief explanation of what digital tabletops are capable, the authors take a glance at how the players of digital tabletop games can interact with them. It was mentioned that some of these interactive tables allow the use of objects which are detectable upon interaction. Physical objects that are tangible and graspable by the players is very important because it bridges the gap of interaction in a way that is more natural when playing tabletop games, rather than dragging sprites across a screen.

These tangible objects, just like in regular board games, can either have a symbolic or iconic design. An example of an iconic tangible object is a cannon or soldier token when playing a game of Risk. This sort of object is immediately recognized by the player, despite his level of experience. By opposition, a symbolic token, and considering the same example from above, would be replacing the cannon by a wooden cube and the soldier by a plastic disk. Upon looking at such tokens, it is not immediately clear to the player which is which and what is its intended purpose.

In order to study the differences between the use of iconic and symbolic tangible play pieces, the authors developed and implemented a game called Weathergods, a game set in the Arabic savannah, for two to four players.

A draught is threatening the harvests, and each player plays as a camel rider from his own village in a quest for gold that will be spent as tribute to the gods. The first player to buy all the offerings, in the correct order, is declared the winner of the game. The game board  displays a map that is divided into 128 square tiles where the camel riders can move across. For this game, three different classes of tangible play pieces for interaction were provided, in both iconic and symbolic versions: camel rider, bandit and detector. 

The camel rider represents the first piece and shows the location of the player on the game board. The second play piece, the bandit, steals money from players. There is only one bandit on the board which can be used by every players, a concept similar to the thief available in the game Settlers of Catan. The third play piece is the detector. The function of this token is to reveal hidden gold. Each player has a detector and can place this piece on one of the tiles near his camel rider. If gold is hidden underneath that tile, the crystal ball in the detector will show a yellow light, only visible to the player handling the detector at that moment.

Tangible play pieces used in the game Weathergods.

Tangible play pieces used in the game Weathergods.

The authors sought after identifying the impact in the gameplay that having play pieces with different representations might cause. It focused on the actual difference in game experience, divided into three categories; the understanding of the game, the understanding of the play pieces used in the game and the fun experienced during the game. For this matter, the participants played the game together with two experimenters. After 10 minutes of play (test 1), the play pieces were replaced by the other available version (test 2). Half of the study population started with the iconic pieces, switching to symbolic pieces later on, while the other half began playing with symbolic pieces first. The participants were not given an entire explanation of the game in order to get better results. After the first play session (10 minutes), the participants filled a questionnaire with questions regarding the topics already mentioned. They then played for another 10 minutes playing with the other play pieces, being interviewed after this second play session about the same topics.

In order to interpret the results for understanding of the play pieces, scores were calculated for each participant. The participants were evaluated  during each condition by being asked to name the functions of the play pieces. Each correct function was worth 1 point, or 0.5 if only the representation was known. To understand whether the participants had learned between test 1 and test 2, the scores on each test were calculated. If a participant had an increase of score from test 1 to test 2, the authors assumed that the increase of score (naturally) reflected increased learning of the role of the play pieces. Despite the fact that, participants who were tested by playing the game with iconic pieces first scored higher learning scores than those who played in reverse, a t-test for statistical correlation revealed that this result was not significant ( t = 5.06, df = 28, p < 0.001).

The questionnaire filled in between play sessions was used to assess the level of understanding of the game from the players. If certain key words or sentences were mentioned in the answers of the players, they were awarded greater of fewer points, thus adopting a similar scoring mechanism than the one previously mentioned. Once more, the participants from the first condition scored higher, but no significant correlation was found (t = 0.91, df = 28, n.s.).

The final result described by the authors is the level of fun during the game, which was assessed with the aforementioned questionnaire as well, by adopting again the same sort of scoring mechanism. The means of both groups did not differ much in regards to scores, and,  after conducting a t-test for independent groups, the result was revealed to be not significant (t = 0.61, df = 27 , n.s.), which indicates that both groups have experienced the game as equally fun, regardless of the understanding of the play pieces and the game itself.

Nevertheless, despite most of the results being indicated as not statistically significant, the authors gathered the personal opinions of the players from the interviews conducted after both play sessions. These interviews revealed that participants, as a whole, found the most fun when playing the game with iconic play pieces, even thought the statistical results showed that both groups had the same level of fun. When asked about preference, the majority of the participants also stated that they preferred to play the game with the iconic tokens rather than their symbolic counterpart, particularly because these play pieces were compatible with the game’s theme.

The discussed reasons for change in this study are mostly the fact that the sessions were timed, hence the slower players did not experience the same events during the play sessions as faster players, and the fact that frequent players were not equally divided between the two groups. As it is predictable, upon conducting a t-test over the average scores of understanding of the game for frequent players and non-frequent players, the more seasoned players scored higher in understanding, a fact that was concluded as statistically significant.


The authors conclude that both study groups understood the play pieces and the game in an equal manner, despite the play pieces with which they played first. The experience, understanding and fun of the game is not influenced by the appearance of the play pieces. What the authors found remarkable was that, participants from group 1 (iconic play pieces first) scored higher both in the talk-aloud questions about the role of pieces, as well as in the questionnaire, when compared to group 2 participants, revealing that there is a higher understanding of the game if you play with more representative tokens first. To me, this is no real surprise, because this is what generally happens when you are first trying to understand how the game works. All the pictures, tokens and game mechanics details are new to you, so once you get used to them (after a play session), you are already accustomed or attuned to them, not really needing an accurate visual representation of what the token is supposed to be or do.

In my opinion, it is a bit sad that all the results revealed to be statistically insignificant, other than the fact that iconic play pieces were perceived as being more fun to play with and were the choice of preference. Nevertheless, what this tells us, in my view with the game designer hat is that, when you design a good game mechanic, the visuals are only but a mere accessory. This is visible to some extent in other types of games like the trend of indie games, where visuals can be deliberately crude, while the experience can still overthrow some of the best AAA titles.