An interactive learning event incorporating the game takes less than a couple of hours; however it is concentrated learning. The game is easy to set up.
You will needs a good understanding of variation: the causes and effects.
PREPARATION
• Participants: teams of five players
• Resources: flip charts and pens
• One die and 50 ‘widgets’, e.g. buttons, pennies, etc. per team,
• Score sheet for each player
INSTRUCTIONS TO PARTICIPANTS
• Ask each team of five players to sit in a row and give each player a score sheet
• Number the players 1 to 5; each player is a step in the end to end process
• Put the pile of widgets on the left of player 1
• Explain that the widgets (or whatever commodity the team normally work) should be passed to the next player to their right; that the players represent a process; that each turn represents a week; and they have been contracted to deliver 35 widgets to the end of the process within 10 weeks. Note: this is based on the fact that each person can throw anything from a 1 to a 6, and on average, will throw 3.5, which is equivalent to 35 at the end of 10 weeks.
• Explain that a key part of the process governance is completing the score sheet (see the relevant section)
PLAYING THE GAME
• Give the die to player 1
• Player 1 rolls the die and takes the relevant number of widgets from the box and passes them and the die to player 2. Player 2 rolls the die and passes what they can to the third person, and so on
• When player 5 has rolled the die and delivered as many as he can to the far end of the row (last widgets box), you can start round 2
• The game lasts for 10 rounds – each person will have the opportunity of throwing the die 10 times
COMPLETING THE SCORE SHEET
• Each person records their position in the process (1 - 5)
• The round they are on
• The number they threw on the die
• The number of widgets they moved to the next player
• They are contracted to move 3.5; so you might pre-print this
• The difference between what they moved and what they were contracted to do
• The cumulative performance for each round (1+2, etc)
| Round Number | |||||||||||
| Example | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | |
| Capacity (The number that you threw) | 6 | 4 | 6 | 1 | 2 | 4 | 4 | 1 | 1 | 4 | 3 |
| Activity (The number that you could move) | 3 | 4 | 6 | 1 | 2 | 4 | 4 | 1 | 1 | 4 | 3 |
| Contract (The number you should have moved to meet the contract) | 3.5 | 3.5 | 3.5 | 3.5 | 3.5 | 3.5 | 3.5 | 3.5 | 3.5 | 3.5 | 3.5 |
| Activity minus Contract | -0.5 | 0.5 | 2.5 | -2.5 | -1.5 | 0.5 | 0.5 | -2.5 | -2.5 | 0.5 | -0.5 |
| Cummulative performance | -0.5 | 0.5 | 3 | 0.5 | -1 | -0.5 | 0 | -2.5 | -5 | -4.5 | -5.5 |
AFTER 10 ROUNDS
• Ask all players to hold up their score sheets so that the whole table can see the cumulative performance of the system. In general the cumulative performance should get worse (more negative) for each round and the further downstream in the process you get
• On the flip chart draw up a table with each team’s name (e.g. table A, table B, etc) and ask the teams to call out the number of widgets they actually processed at the end of 10 rounds
• Congratulate the team that scores the most and ask them what was different about their system when compared to the other teams. The answer should be nothing – their score due to luck.
LEARNING – DISCUSSION
Q: What is going on?
A: If player 1 rolled a 5, they would pass 5 widgets on and record that they passed 1.5 more than the average activity required of them. If player 2 rolled a 4, they would pass on 4 and record that they passed 0.5 more than the average activity required of them. They would have a queue (backlog of 1) If player 3 rolled a 2, they would only pass 2 along. However they delivered 1.5 less than the average activity required of them. They also had a queue of 2 (backlog). As the game progresses the players downstream of position 1 notice that there are rounds when either:
• They have more widgets waiting than they have capacity (throw on the die) to pass on and have a backlog
• Or they have fewer widgets to pass on than they threw on the die; they are starved of work
Q: How could the teams improve their systems performance so that it always meets their target of 35 widgets at the end of 10 weeks?
A: The most common strategy is that each player would always have a pile (queue)of widgets waiting so that if they throw a big number then they will always have widgets to pass along. But it is not good to always plan to have a queue; work in progress can cost a lot of money
Q: What is the real problem for the team?
A: The real problem is the assumption that the capacity of the system is the average for all the departments in the process. Since the effective capacity of each station in the chain varies between 1 and 6, the average effective capacity is 3.5. However this isn’t the case.
Q: What is the chance that the team will process at least one widget?
A: 100%. Each position is bound to throw 1 to 6 and will therefore be able to pass one widget along from position 1 to position 6.
Q: What is the chance that the team will process six widgets in each round?
A: The chance of producing at least 6 widgets = 0.013 %
Q: So what is the chance of the team producing 3.5 widgets in each round?
A: About 9%. This system has a 9% chance of meeting its target! Chance of at least 3 widgets = 13 % Chance of at least 4 widgets = 45 = 3 %
Q: Is this why the delivery process is difficult to manage?
A: Often no one understands the real flow through each process step.
Q: What can we do?
A: You have a choice of three strategies:
Strategy 1: Measure the variability of the capacity in all the steps continuously and build a computer model for predicting the amount of backlog that each department needs to hold in order to optimise the performance of their process? Comment: Don’t even think about it!
Strategy 2: Increase capacity – more dice. Comment: Doubling the capacity with the same amount of variation will not help.
Strategy 3: Reduce the variation – load the dice so they always throw a 4. Comment: This is the best solution. Find out why the capacity varies so much and stop it. 80 - 90% of the variability is within the system’s control.
Further questions to stimulate the debate
• What causes the capacity to vary?
• How would the team set about stopping it?
• How should the contracting process work?
• What would be the appropriate measures of performance?
• Which do we currently do?
No comments:
Post a Comment