Charles Suscheck presented how he uses a variation of the card game Rummy to teach the importance of communication, planning, and collaboration on projects at Agile2008. The game explores the effects of various levels of distribution on a team, as well as the impact of adding or removing experts on the team during a project.
Charles' session began at 2:00 PM on Wednesday at the Agile2008 conference in Toronto. The participants were broken into 3 groups, each consisting of 5 or 6 people.
- The first group was told that they could only communicate via 'email', which would be simulated by written messages passed around the group by a designated 'mail server'.
- The second group was told that they would only be able to communicate via 'telephone'.
- The third group was given no limits on how they could communicate.
Each team had one member designated as the 'user'. Once the game started, the users would be given complete game rules, as well as variations to the rules that would be introduced in each succeeding round of the game. Another set of rules would be distributed among the rest of the team such that each player had only a partial set of rules.
Charles' design for the game scales up to having 6 different types of groups in the game at once:
Group 1: Via Email Only, No additional user
Represents Team geographically distributed across several time zones.Group 2: Via Email Only, Additional user involved
Represents Team geographically distributed across several time zones.Group 3: Via Telephone Only, No additional user
Represents Team with no documentation.Group 4: Via Telephone Only, Additional user involved
Represents Team with no documentation.Group 5: Open Communication, User inserted at last 5 minutes
Represents Team falling behind that has an expert inserted to help.Group 6: Open Communication, User involved until last 5 minutes
Represents Team that is doing well, so has an expert removed.
The teams were given 3 minutes, without communication restrictions, to organize themselves, plan how they were going to communicate, what the roles on the team would be and generally what their strategy would be.
Next, the rules were distributed. Participants were told not to show or read the rules to each other, though they could describe their own interpretation of the rules using the communication channels available to their group. The stated goal for each team was to complete as many rounds of modified Rummy as possible, with the team completing the most rounds being the winner. At least one player was given a rule that could be interpreted as indicating that the individual player with the best score would be the 'winner' on their team.
The game ran for 20 minutes, with teams playing a continuously changing version of Rummy. As the game ran, gaps in each team's processes, understanding of the rules, and strategy where uncovered and the teams had to adjust.
When the 20 minutes had expired, the team with no communication restrictions had completed nearly 11 rounds of Rummy and had simultaneously created performance documentation that had been requested as an additional deliverable. That team had quickly decided that since the object of the game was to complete as many rounds as possible, that they would play all hands face up and collaborate to help the first possible person 'go out', without regard for individual 'scores'.
The 'telephone team' had decided to use a conference call model to allow as much communication as possible. Even so, they didn't hit upon the notion of collaborating to help one person 'go out' until after they had played one round. Ultimately, the team completed 2 rounds of Rummy.
The email team never got to the collaboration stage, with one or two members playing aggressively to be the 'winner' on the team. This team managed to complete one round of Rummy in 20 minutes.
After the game was played, Charles facilitated a debriefing session where the participants shared experience, lessons learned, and drew parallels to their real project experiences. Many participants shared that the dynamic of internal competition was a significant block for true collaboration on teams that they had worked with. Even though all teams were given up-front planning time, the value of high-bandwidth communication for adjusting the plan on the fly was striking. This aspect of the communication restrictions may have been more significant to the results than the restrictions' direct effect on the speed of game play.
A full set of guidelines for playing the game can be found on the Agile2008 website.