While a large literature has studied how people make forecasts, less is known about how lay people process and interpret forecasts presented to them. We contrast two common ways of communicating an uncertain fore- cast, as either a chance (e.g., the probability of winning) or as an expected margin (e.g., the point spread). Across five studies, we find a robust chance-margin discrepan- cy: people tend to treat a chance forecast as conveying greater probability of the higher-likelihood outcome than the statistically equivalent margin forecast