I wouldn't have thought you needed to prove that overconfidence in war could be dangerous, but ever since I lived through the Bush Jr. Republicans' apparent efforts to repeat the failures and crimes of British colonialism in Iraq, I guess you do need to prove scientifically that arrogance can have a bad effect in dangerous situations.
Overconfidence is a disadvantage in war, finds study
Roxanne Khamsi, NewScientist.com news service
Overconfident people are more likely to wage war but fare worse in the ensuing battles, a new study suggests. The research on how people approach a computer war game backs up a theory that "positive illusions" may contribute to costly conflicts.
"It supplies critically needed experimental support for the idea that positive attitude - which is generally a [beneficial] feature of human behaviour - may lead to overconfidence and [damaging] behaviour in the case of war," comments Peter Turchin of the University of Connecticut, US...
http://www.newscientist.com/...
Flip down for further excerpts, it's pretty interesting.
...Dominic Johnson of Princeton University in New Jersey and his colleagues recruited 200 volunteers to play the role of the leader of a fictitious country that is in conflict with another over newly discovered diamond resources that lay along a disputed border.
Before the game, volunteers were asked to predict how their performance would rank compared with the other 199 people in the experiment. They then played anonymously against other volunteers and received $10 if they won the game, that is, if they amassed the most wealth or defeated their opponent in war...
...Careful negotiations with opponents could win players additional resources in exchange for the diamonds. But they also had the option of waging war. Their victory in battle was determined by how much they had invested in their military, along with an element of chance.
Players who made higher-than-average predictions of their performance - those who had higher confidence - were more likely to carry out unprovoked attacks... [also] people with higher self-rankings ended up worse off at the end of the game. "Those who expected to do best tended to do worst," the researchers say. "This suggests that positive illusions were not only misguided but actually may have been detrimental to performance in this scenario." [Tests yielded gender as a slight but non-determinate factor, with males being more aggressive but having little to do with tested levels of actual testosterone.]
...Those who launched unprovoked attacks also exhibited more narcissism, scoring 13 out of 15 on a standard psychological test. More peaceful types scored 11 on average on the same test. The trend applied to both men and women. "So it's not maleness per se but narcissism that makes some people overly optimistic and aggressive," suggests Bertram Malle at the University of Oregon in Eugene, US...
...Malle agrees that the study raises worrying questions about real-world political leaders. "Perhaps most disconcerting is that today's leaders are above-average in narcissism," he notes, referring to an analysis of 377 leaders published in King of the Mountain: The nature of political leadership by Arnold Ludwig.