figure 3: in this inversion-problem game, given an input, Player 1 produces an output, and Player 2 guesses the input.

Player 1

INPUT

Player 2

(t ) output 1, 1 1, 1

 

(t ) output 1,n 1,n

(t ) output 1, 1 1, 1

(t ) output 2, 1 2, 1

(t ) output 1,n 1,n

(t ) output 2,m 2,m

Players win if/when output = INPUT

2,i

figure 4: in this input-agreement game, players must determine whether they have been given the same input.

Player 1

 

INPUT 1

Player 2

 

INPUT 2

(t ) output 1, 1 1, 1

(t ) output 2, 1 2, 1

(t ) output 1,n 1,n

=≠

(t ) output

2,m 2,m =≠

Win if players guess whether INPUT = INPUT

12

 

input; all they know is that they must “think like each other” and enter the same output.

This game structure accomplishes several goals at once: a good “winning” strategy for the players is to produce outputs related to the only thing they have in common—the input; when the two players provide the same output, this partially verifies that the output is correct, since it comes from two largely independent sources; and trying to agree on the same output with a partner is an enjoyable social experience.

Inversion-problem games. Resulting from any of three seemingly different games—Peekaboom, ²⁵ Phetch, ²³

and Verbosity²⁴—they, in their most general form (see Figure 3), can be described through the following rules:

Initial setup. Two strangers are randomly chosen by the game itself from among all potential players;

Rules. In each round, one player is assigned to be the “describer,” and the other player is assigned to be the “guesser.” The describer is given an input. Based on this input, the describer produces outputs that are sent to the guesser. The outputs from the describer should help the guesser produce the original input; and

Winning condition. The guesser produces the input that was originally given to the describer.

Verbosity (see the sidebar) is an inversion-problem game where the input is a word and the outputs are commonsense facts related to that word. Following the input word “milk,” the game might output such facts as “it is white” and “people usually eat cereal with it.”

Although the design of Verbosity involves other game elements not described here, the basic idea is that players need not be asked directly for facts about “milk.” The game is designed such that facts are collected as a side effect of playing. Players told to “please enter facts about milk” might not be motivated to do so or enter incorrect information.

In inversion-problem games, partners are successful only when the describer provides enough outputs for the guesser to guess the original input. If the outputs are incorrect or incomplete, the guesser will not be able to produce the original input. Therefore, the game structure encourages players to enter correct information. At the same time, having one player guess the input while the other describes it is an enjoyable social interaction, similar to the popular children’s game “ 20 Questions.”

Additional elements can be added to inversion-problem games to increase player enjoyment, including transparency and alternation:

Transparency. In post-game questionnaires, players of inversion-problem games have expressed a strong desire to see their partner’s guesses. We therefore experimented with adding a level of transparency between players

so the actions of one would be visible to the other. In games like Verbosity and Peekaboom this transparency is achieved by displaying partner guesses to the describers and allowing them to indicate whether each guess is “hot” or “cold.” This design feature increases the social connection between the players without compromising output correctness.

Alternation. Unlike output-agreement games (where both players continually perform the same task), inversion-problem games are asymmetric in that each player in the pair performs a different task. In some games of this type, one of the two roles involves more interaction or is faster-paced and thus more enjoyable than the other role. In such cases, to balance the game and maintain an equal level of player engagement, player roles can switch after each round; the guesser becomes the describer, and the describer becomes the guesser.

Input-agreement games. Representing a generalization of games like Edith Law’s TagATune⁹ (see Figure 4), they can be described through the following rules:

Initial setup. Two strangers are randomly chosen by the game itself from among all potential players;

Rules. In each round, both players are given inputs that are known by the game (but not by the players) to be the same or different. The players are instructed to produce outputs describing their input, so their partners are able to assess whether their inputs are the same or different. Players see only each other’s outputs; and

Winning condition. Both players correctly determine whether they have been given the same or different inputs.

In TagATune, the input is a sound clip, and the output is a series of labels or tags for the clip. The two players achieve the winning condition (and obtain points) only if they both correctly determine whether they have the same input song. Because players want to achieve the winning condition, they each want their partner to be able to determine if their inputs are the same. This means it is in their own best interest to enter accurate outputs that appropriately describe their individual inputs.