It's definitely faster to just assume that there is evidence to reinforce our existing biases than it is to do the same after succeeding or failing to locate that evidence, so considering that all three methods lead to the same result, the one with the lowest time cost is optimal.