But to complicate matters further, Uber expects this to happen, and adjusts the wait times accordingly based on the "statistically expected time." So they are almost always going to be wrong, Kalanick admits, because they make them wrong on purpose. You’re not getting an ETA based on how close the nearest car is at that moment, but instead, it factors in the probability that you might have to wait for the second-closest car. That method, Kalanick argues, means that your estimate "will be less different/wrong on average." So, less wrong, but not necessarily more right.