The Weather Roulette: a gambling game to communicate probabilistic climate predictions

16:15 Tuesday 28 May


Room S10


Marta Terrado (Spain) 1; Isadora Christel (Spain) 1; Dragana Bojovic (Spain) 1; Llorenç Lledó (Spain) 1; Rodrigo Manzanas (Spain) 2; Albert Soret (Spain) 1; Francisco J. Doblas-Reyes (Spain) 1

1 - Barcelona Supercomputing Center (BSC); 2 - University of Cantabria

Climate information and knowledge responding to user needs are essential to assist decision-making in climate-sensitive problems. Although probabilistic climate predictions (from three weeks to a decade into the future) can provide valuable information for different socio-economic sectors, they are not easily adopted by users. Barriers hindering user uptake include the reduced skill of such predictions compared to the skill of weather forecasts, the difficulty to deal with probabilistic outcomes and the lack of a common terminology between the communities of climate scientists and users. The limited understanding of the concept of climate prediction skill prevents users from building trust and changing their current practice (often based on the use of past observations to estimate future conditions).

We present gamification as an engagement tool to break communication and understanding barriers. Using the Weather Roulette gambling game (Hagedorn and Smith 2009), previously applied namely to weather forecasts, we illustrate the added value of applying probabilistic climate predictions. We translate the concept of skill in economic terms using interest rates and return-on-investment concepts, which are more familiar for non-climate scientists. In particular, we apply the Weather Roulette to seasonal climate predictions of wind speed provided by the Resilience prototype ( for a period of 33 past years at locations with installed wind power capacity. We find that for skillful regions the economic benefits of using such predictions in the long term increase. By contrast, either gains or losses can be obtained in regions of limited skill. As a practical example, we present the Weather Roulette app directed to the wind energy sector.

We show how a tailored communication can help users understand the usefulness of probabilistic predictions and build enough trust and capacity within the user community to ensure the uptake of state-of-the-art scientific products. Beyond helping to transfer the knowledge arising from climate research to potential users, the obtained results can have important practical implications for the development of climate services.