It is often useful to predict the decisions of others, especially in conflicts. Everyday experience suggests that expert judgements provide the most accurate forecasts. For example, the media routinely turn to experts for their opinions on what will happen in conflicts. A survey of 62 experts found they believed experts' forecasts would be more accurate than novices', irrespective of the method used. As part of ongoing research, experts predicted the decisions that would be made in eight real conflicts, such as whether a supermarket chain would agree to a plan to sell major appliances. In addition to experts' forecasts, novices and game-theory experts provided forecasts, other experts used structured analogies, and novices played the roles of conflict protagonists in simulated interactions. Experts using their unaided judgement (32% correct , n=106) and game theorists (31%, 101) were not significantly more accurate than novices (29%, 169). Using structured analogies does not involve using expert judgement to forecast and 46% of forecasts were accurate (97). Sixty-two percent of simulated interaction forecasts were correct (105). More experience was not associated with greater accuracy. In summary, by itself, expertise was of little value. Nevertheless, expert knowledge of the target situation and of analogous situations was useful.