I recently played game named “Lost in the desert” to train on the backlog prioritization. I really like it to bring some tools to the audience, but this game can be so much more powerful when it’s used as well to share the importance of the product vision for a product Owner.

What is Lost in the desert ?

For those who don’t know this game I advise you to have a look on google about it, there is a lot of reference and even the US Navy have proposed a solution. Here is a quick summarize : You are playing the role of a plane crash survivor, all the crew is dead and all the passengers are perfectly safe. You are crashed in the middle of Arizona desert, 80km away from the standard flights corridors and 110 km far from the closest town.

I know… There are some better days… but don’t worry you have saved 15 items from the burned plane and you need now to prioritize them from the most important to the less important to survive.

Most of the time, the game is organised in two session, the first one 10 min of individual prioritization, (it’s important to ask everybody to not communicate). And then the second one, 15 minutes of group prioritization, the team should find a consensus and propose one unique list.

What is happening ?

First phase

All the elements might be really important for you depends of the strategy you choose to survive, so during the first phase each of the participant is defining is own approach and is trying to prioritize the items.

You can let the audience do the prioritization without assistance but it’s from my perspective the perfect moment to pause the timer after 5 min and explain what is the MoSCoW method.

The players will then use it according to the strategy they have chosen. For those who don’t know what is MoSCoW don’t worry, nothing « rocket science » here, you can refer to the holy wikipedia page and contact me if you would like some best practice around the tool : https://en.wikipedia.org/wiki/MoSCoW_method.

Second phase

In this phase we ask the whole team to propose a common prioritisation. Generally by groups not exceeding 5 persons. I prefer creating a challenging environment for the team and create larger groups.

My objective here is clearly to underline the difficulties the team will have to get aligned as there will be more different interactions. So I like to let them struggle ; as there are not align on the strategy, there are discussing for each item if it’s really the most important and they are losing a lot of time. If needed, I disturb them from the possibility to agree on a common strategy and share a vision.

After 7 minutes generally the prioritization is not advanced sufficiently to complete it by the end of the allowed time. So I stop the stopwatch to do a 5 minutes debrief (a retro :)).

By simply asking to 2 or 3 person which strategy to survive they prefer, everybody realize quickly that they need to be align on the vision.

This being done I let the team complete the prioritisation which goes generally super quickly having a shared vision and using MoSCoW.

Debrief

After that comes the debrief and I can deeply insist on the importance for a product owner to have a clear product vision and to share it to the rest of the developpement team. This should be done from the early beginning in the backlog grooming and the sprint planning for exemple, and all along the developpement.

I would like to remind here the scrum guide on the definition of the increment :”The increment is a step toward a vision or goal”*. That’s why the product vision should not only be shared through the items of product backlog, because without a clear shared vision we cannot collectively reach the following step.

Different ways to drive the game :

In English : http://bitsofknowledge.waterloohills.com/communication/agile-games-night-lost-in-the-desert/

In French : http://coach-agile.com/serious-game/perdus-dans-le-desert/

Reference :

*“The increment is a step toward a vision or goal” : Scrum guide, section Sprint backlog / Increment. https://www.scrumguides.org/scrum-guide.html#artifacts-sprintbacklog