In early 2010, Allessandro Pluchino and his team from University of Catania in Italy shot to scientific fame when they published the first computational study of the Peter Principle. The principle, devised by Canadian researchers in 1969, states that "Every new member in a hierarchical organization climbs the hierarchy until he/she reaches his/her level of maximum incompetence," meaning that an employee is promoted until he or she can no longer work effectively.
The team, consisting of two physicists and a social scientist, demonstrated the validity of the principle, "but above all, we have found a possible, apparently paradoxical, solution to avoid its negative effects: adopt random promotions," the trio said during their acceptance speech at the 2010 Ig Nobel Prizes.
(To run a simulation on your own organizational incompetence, click here.)
Since then, the Italian researchers have turned their minds - and their computer models - to other areas, including the operation of government (see their paper published in March 2011, "Accidental Politicians: How Randomly Selected Legislators Can Improve Parliament Efficiency") and now computing grids.
"Inspired by our findings on the Peter Principle, we are trying to test the power of random strategies on the process of submission and queue management of many jobs on the Cometa grid, using the grid itself to run simulations," Pluchino said.
"We are comparing the actual mechanisms used by grid administrators for optimizing the choice of the queue to assign a given job (based for example on the estimated traversal time, or on the number and the kind of jobs present in a given queue at a given moment) with a completely random choice of the queue.
"We strongly suspect that this latter strategy could be more effective of the others, which are in any case affected by several errors mainly due to an unavoidable delay in recovering the information about the state of the system," Pluchino said.