My current work made me refresh my documentation a bit on this little agile game, the coin game, or the “coin toss”. Simple but very effective, take a good fifteen minutes.
And as this old Chinese proverb says: Tell me and I will forget. Show me and I will remember. Involve me and I will understand. – Chinese proverb
The objective of the game will be to demonstrate some fundamental benefits of agile: a focus on value by maximizing it (maximize value, minimize effort), and delivering it as early as possible (“time to market”) while integrating a capacity for change.
Rules
4 participants, each represents the phases of a classic project (that is to say the old-fashioned way…): let’s say for example: definition, specification, implementation, acceptance. One person times: the time of the first phase, and the total time. A pile of coins (a little more than one euro in small coins from 1 cent to 20 cents). “Flipping a coin” is the metaphor for carrying out an activity around a feature (its value is that of the coin). The coins, after being flipped, pass through all the phases of the project until its end.
First round: the classic project
On the classic project the team will take around 1min, 1min30 to pass all the coins, and will reach let’s say 1 euro and 15 cents (randomly). On a classic project the coins will first all be flipped by phase before being passed to the next phase (we finish all of the definition, before moving on to specifications). This is the rule of the classic project (even if some think it’s a caricature, we still encounter it quite often).
Display on the board: the time of the first person, the overall time, the sum earned.
Observations
Some people don’t work while others work. In real life you’ll be told they’re working on “something else”. Problem: in real life there can quickly be a bottleneck effect Problem: we can’t switch beyond two activities without losing productivity, so “working on something else” very quickly has limits.
Second round: the agile project
Agile switch: everyone works at the same time (as soon as we have flipped a coin we can pass it to the next phase) to have finished things (at the end of the chain as quickly as possible) because we want value, and value (and learning) only comes with finished things. Ask people in advance: “Will this be shorter?” If some say yes, emphasize “Really shorter? … Well we’ll see!”. In fact it will be 4x shorter (to put in perspective with the loss of productivity beyond 2 activities) with the same number of coins, and the same sum.
So we will manage to pass all the coins in four times less time (and for the same value), let’s say 1.15 euros in 34 seconds instead of 1min30.
Observations
We’ll note the feeling of fatigue linked to this genuine interaction. We must be careful about this in agile (“sustainable pace”).
We’re always asked to “go faster”. Yet it’s often futile to go faster, to accelerate the pace or to increase the amount of work. To go faster: we must change the system, think differently. This is what we’ve just done.
We go from 1min30 to 34 seconds, but especially the first delivery with value arrives around 14 seconds (in any case much earlier, from the beginning, if in the end it’s not the value we wanted, or not the expected value, we can stop at lower cost or change our approach).
Third round: brief visit to Kanban
If we note a bottleneck linked to the new rhythm (a more tired participant ? more clumsy? less skilled? not in shape today? no matter it’s the symbol that phases aren’t alike and require different attention), in short if we note bottlenecks and pile-ups in places we can add an element of Kanban (one of the agile approaches): we place a post-it or a piece of paper between each phase: it’s the passage point, the exchange ticket, we can place one or two flipped coins on it only if the paper is empty, otherwise we must wait for the next phase to be in rhythm: this regulates the flow to “just in time”.
I’ll repeat: each participant can only pass the flipped coins to their neighbor, the next phase, only if they have freed up space on the paper by taking care of the previous coins.
Observations
With this “just in time” regulation, no stock, fewer coins in the flow, if needed we can easily change our approach. And we don’t go any slower.
You can explain the variations between the different stages by the fact that the forces at play in real life aren’t the same.
Fourth round: the true agile project
- In fact we never have time in real life to do everything. We must
- focus on value. We must prioritize it. We’ll set ourselves
- 20 seconds max, and try to have the maximum value in 20 seconds.
- The person who initiates the flow can therefore choose the coins. They don’t
- have the right to sort them, to anticipate, but it’s useless, as in
- real life they know where the big pockets of value are. Finally
- at ten seconds, we remove all the coins and replace them with others
- “the market just changed!”. The new coins introduce new large coins, as in real life: new pockets of value have appeared.
Observations
In 20 seconds we manage to have 97 cents (for example), compared to the starting point of 1min30 and 1.15 cents. From the 14th second we obtain value, while having absorbed a market change.
You can announce 25 seconds, and finally reduce to 20 seconds to feel the reality even more…
Is this Fordism???
The rigorous and repetitive aspect can quickly be equated to Fordism, to Taylorism. When a Scrum dynamic is in place the impression can be the same! At least in appearance, because deep down the people involved decide their way of working, they are autonomous and empowered, this is fundamental.
Addition: The fourth column riddle
[Edition January 26, 2017]
For several months I’ve been adding information: the fourth column. I save space for a fourth column that I add at some point in the form of a riddle: what if there was a fourth column? What would be interesting to measure? Do it when you feel it but generally after the second round (the coins two by two) it makes sense.
The answer is the time between the start and the first production release, the first delivery of value in production as Alexandre says. Often between the first run and the second the difference is striking. This allows us to discuss Time to market, cost of delay, etc.
A quick little sketch (click on it).