Agile, Lean & Kanban Software Development

Expanding the knowledge base of Software Development Excellence

The Ball Game

I first read about The Ball Game about 7 years ago.  Maybe it wasn’t 7 years ago but it was long enough ago that I don’t even remember my first encounter.  It was created by Boris Gloger (Kudo’s Boris) and frankly it’s brilliant.  It has all the hallmarks of a great learning game 1) it’s a ton of fun to play, 2) it’s rules are essentially simple and 3) teams actually understands the point of the game (that they can improve by self-organization).  I’ve run this many many time so I know it works as advertised.

What is the primary take away we are looking for?

To teach teams that they can improve by using self-organization.  By stopping at the end of each Iteration to inspect their performance and adapt to what they have just experienced.  They also will understand that if they want to experience an improvement they have to really CHANGE the system/process/flow (the old adage about the “definition of insanity” is true).   And of course that you can have fun AND improve.

What do you need?

  • A lot of balls. I have a container of tennis balls (about 25 or so) I use but ping pong  balls or racquetballs would work as well.  I used marbles once and that worked too.
  • A whiteboard or large paper taped to the wall to graph the progress.
  • The team

What is the objective?

To see how many balls the team can through their process within a certain time box.  Allow them to learn and change their process and try again.  Track how their changes affect the throughput.

The Rules of the Game

  • You are One Big Team
  • All balls must travel through the air from person to person
  • You cannot pass a ball to a direct neighbor
  • For each ball the Starting Point = the Ending Point
  • Each Iteration will take 2 minutes
  • Between Iteration the Team will have 1 minute to Inspect and Adapt the system
  • There will be 5 Iterations

Twl2mod2sm1024

The Playbook

  • 2 minutes Introduction to the Game
  • An initial 2 minute preparation time
  • Before an Iteration begins
  • The Team give an estimate of how many balls will go through the system in 2 minutes
  • Graph the estimate for the Iteration
  • Run the Iteration for 2 timed minutes
  • Graph the Actual number that made it through the system on the same vertical as the estimate
  • Give 1 minute Inspect and Adapt time between Iterations
  • Run 5 Iterations
  • Follow the 5th Iteration with a Debrief (5-10 mins).

Chart1_cr1024

The Debrief

  • What happened?
  • What was the Iteration that felt best?
  • Did you experience a rhythm?
  • How did you improve with no “leader”?
  • What did you have to do in order to increase your flow-through rate?

A system has a natural velocity; each team has to find theirs

Once you have the basics down you can try these options:

Experience how it feels to disrupt an Iteration’s Flow
Run 6 Iterations and in Iteration 4 disrupt the flow by pulls teams away from the team.
Debrief, how did you handle it when your plan was disrupted?  Protect your Iterations.

Use dropped balls to represent a commitment to Quality
In Iteration 1 and 2 allow the team to pick up dropped balls but in 3, 4 and 5 they must stay on the ground (improvement by paying purposeful attention to quality).  Note: If the velocity of Iteration 3 goes down keep at it and 4 and 5 will increase because they will “learn”.

How does “getting closer” to each otherhelp?
Team will invariably “toss” the balls in the air to each other which increases the chances of dropped balls (“defects”).  Have the people across from each other hold their hands above and below each other 6 inches apart or so and simply simply “drop” the ball.  Debrief: Getting “closer” to each other decreases the probability of the occurrence of defects which usually results in an increase in flow-through.

Smaller Teams can go faster AND better
Advanced: If you have a very large team of say 20 or so you can still do this.  But as an added exercise split the team into two teams of 10 each and run one more Iteration with no time to Inspect and Adapt.  Simply take each team of 10’s estimate, add them together and plot the combined estimate then run the Iteration and combine and plot the actuals.   The results will be fairly dramatic.  Why would the “improvement” be so great.  Hint: It has to do with smaller teams operate more efficiently than larger teams.

Further Innovations: I’ve been toying with using the large doctor’s office-type tongue depressors to see how they would work for this game.  They are oddly shaped and could be good to see how they handle the “quality” and also they would be REALLY easy for transporting to and from training!

 

About

View all posts by