Unfortunately there is an error in the percentages in paragraph 5 of the discussion section. The corrected paragraph is given below. In the sound-off condition, players on average estimated that they won 33 times when in reality they were only exposed to 28 wins (thus, on average they overestimate by 5 (i.e., 18 %) the number of times they won). Crucially, this propensity to overestimate these wins is exacerbated when sounds accompany the losses disguised as wins. In this sound-on condition, players estimated that they won on average 36 times (an overestimation of 8 (i.e., 29 %)).