Results Experiment 1: fast and slow pointing in the direct and indirect pointing conditions.

A) End-point error versus target location for an example participant in the fast pointing condition. The different symbols, indicating the response (left or right of target) for each trial, are clearly segregated with respect to the random error and not with respect to systematic errors. B) Psychometric curve (cumulative Gaussian) for the data shown in A. Blue and red symbols indicate the response versus random error for individual trials. Grey dots indicate the proportion of “right” responses for binned trials (bin size = 10mm). The size of the dots is indicative of the number of trials taken into account for each bin. The solid curve indicates the obtained psychometric fit from which we obtained the perceptual noise . C) the perceptual noise for the left/right discrimination of the errors (red data point corresponds to the example from B). For comparison, the dashed line shows the average across participants for , i.e. the standard deviation in random pointing error. Bars and error bars indicate the median and 25% and 75% percentiles across participants respectively. Individual participant results are indicated by the separate points. Closed symbols indicate participants with significant knowledge of their random error; open symbols participants whose performance did not differ from chance. D) The standard deviation in random error . Bars and error bars indicate the median and 25% and 75% percentiles across participants respectively. Individual participant results are indicated by the separate points.