A state of mind that causes a forecaster to think that the probability that a forecast is correct is greater than the actual probability. This leads <a href="search.pl?p=prediction intervals">prediction intervals</a> to be too narrow. Experts are overconfident because of various biases, such as an unwarranted feeling of control or a desire to see things turn out well. Overconfidence is widespread. For example, when subjects are asked how many times the letter F appears in: “Finished files are the result of years of scientific study combined with the experience of years,” about half answer incorrectly. Most are sure that their answer is correct for this problem, and those who are more confident are no more accurate than those who are less confident. (The correct answer is six.) See Arkes (2001). <o:p></o:p></P>
<ul>
  <li>Arkes, H. R. (2001), Overconfidence in judgmental 
forecasting,” in J. S. Armstrong (ed.), <I>Principles of Forecasting.</I> 
Norwell, MA: Kluwer Academic Press.</li>
</ul>