The slides used for this session are available to download from here.
The estimation of software projects has been an ongoing topic for many decades. It is notoriously hard to do, imprecise and causes toxic behaviours in many situations. It has been addressed in many ways, all of which come with their own sets of benefits and challenges. We suggest that the issue with estimates is that humans are not wired to do them well: we are much better negotiators and traders than estimators.
Prices in most parts of our economy are driven by liquid markets, supply and demand, bidding and asking, and an element of gambling. The price of projects, however, is most commonly determined by individuals assigning sizes. This frequently leads to a lack of shared understanding amongst stakeholders of when a project will be delivered, what functionality will be in the delivery, and how much it will cost. Disagreements on sizes and content cause uncertainties and frustrations and cannot be resolved.
Our aim in this session is to introduce a gamification of project estimates in order to create this shared understanding and to resolve disagreements in a playful way. The game introduces a liquid economy, where sizes (prices) are set by the players who will buy items (tasks, stories, epics etc) they think are undervalued and sell the ones they think are overpriced until an equilibrium is reached. When a new piece of information impacts sizes, players will adapt prices and there will be trading activity until a new equilibrium reflects the impact.
Producing reliable forecasts is the main goal of the game, but there are additional benefits to playing it. It drives practices like removing blockers quickly, collaborating, delivering fast, establishing transparency and focusing on quality.
The game has simple rules and can be played by anyone involved in the project, not just the delivery team. We'll use half the session explaining the game and discussing benefits and shortcomings, and the other half playing it. We will run a project simulation with random events that will produce unforeseen surprises that impact item sizes. Neither presenters nor participants will know how large the simulated project will end up being, so you will be able to assess the forecasting capabilities of the game in the session itself.
The game is designed to be played by the whole team, so the session is open to all and no initial skill set or background is required. We may have to limit the number of players but will make sure everyone attending will receive appropriate exposure to the game.
Ernst is a former consultant who is now working as COO of Building Intellect. Previously, he managed many large and small software projects in various industries. He has helped organisations transition from traditional to agile development approaches and has used many estimation and no-estimation techniques with his teams. Ernst does technical as well as managerial work. He has looked for ways to answer questions about when things will be delivered, what is going to be in the delivery and how much it will cost from both sides.
Paula de Matos is an independent UX consultant specialising in complex data fields, specifically in the life sciences and biotech sectors. Her approach to designing in complex environments is based upon the notion that complex problems often require solutions from multiple domains and fields. She uses expertise from other domains such as visualisation to solve complex problems - hence her interest in bridging the gap between these disciplines.
Paula originally trained as a software developer, then moved to bridging the gap between users and developers in the role of Group Coordinator and User Experience Analyst at EMBL-EBI (the European Bioinformatics Institute). She completed a masters in human computer interactions at University College London. She blogs at http://www.pauladematos.co.uk.
To ensure you're the first to hear about Agile Cambridge 2017, join our mailing list.