The holy grail of personnel selection

Many human activities have a one thing in common: Very few per cent of the men involved contribute a very disproportionate (like tenfold) share of the output. Few per cent of fighter pilots contributed more than half the air combat kills in the World Wars, for example. Similar results are known of submarine commanders, snipers, infantrymen and many more. Even an analysis of epidemics and friendship networks yields similar results; a few individuals are way more active/successful than all else. This is interesting even if the result is in one instance not that extreme; let's take an excerpt of a lessons learned report by Lt.Col. L.Wigram (Sicily Campaign 1943), for example.
The Pl in action is almost invariably twenty-two strong and of whatever Regt good or bad, every Pl can be analysed as follows: Six gutful men who will go anywhere and do anything, 12 `sheep' who will follow a short distance behind if they are well led, 9-6 who will run away. I have discussed these figures with many people and they all agree, although there is some slight disagreement on figures. These figures are roughly accurate as shown by the number of Court-Martials for running away that follow every Campaign. Every Bn has between forty to sixty and there are, of course, many others who aren't caught.
He also points out that the brave ones have a higher attrition rate, so these proportions are not representative. Nevertheless, it's obviously advantageous to know the brave ones and to assign the cowards to units where their behaviour doesn't harm. The quote does by the way also explain why veteran armies are usually stubborn and competent in the tactical defence but almost useless in the tactical offence; their ranks are depleted of especially brave, aggressive men - leaving behind only experienced average Joes. Many accounts of disappointing veteran forces are known in military history - usually associated with tactical offence. Back 2topic; the holy grail of personnel selection is to know the key traits in advance and to assign the right people to the right jobs. The performance gain by doing so is huge and for free. You don't want to employ a potential fighter ace as logistics soldier or an excellent organizer as fighter pilot. The 'Western' and most accepted method to assign the right people to the right positions is meritocracy:
Meritocracy is a system of a government or other organization wherein appointments are made and responsibilities assigned to individuals based upon demonstrated talent and ability (merit).
The result is the Peter Principle.
The Peter Principle is the principle that "In a Hierarchy Every Employee Tends to Rise to His Level of Incompetence."
An excellent Captain can become a poor Colonel. An excellent Colonel can become a terrible General. The requirements change a lot during most career paths, so earlier performance is a poor indicator for future success. The other side of the problem is that we likely ignore dozens or hundreds of phenomenal potential Generals every year because these people fail as Captains. The answer is known long since; neither standing nor previous performance are of much importance. The most important attribute is potential for a job. An old invention to identify potential is the assessment centre. It's nowadays known in Germany as a supposedly Anglo-American human resources tool, but it's in fact an invention of the Reichswehr (1920's). The Reichswehr had to make use of non-aristocrats for officer positions and therefore wanted to differentiate between promising and non-promising candidates. It had a great interest in maximized officer corps quality because it was restricted to 100,000 personnel and trained all its soldiers to be able to replace their superior in order to enable a quick future expansion. The result was that process of testing their potential. 
The importance of potential over performance is officially recognized in many armies, but it's not influential enough. There are still many reserves, many opportunities for improvement. That's interesting and important because an unpleasant reform in this topic could yield much better results than most if not all huge procurement projects - for free! We should also spend more on specific social research, for it would be of extreme value if we learned more about how to identify 'aces' and 'duds' in advance. Such focused research could also help the civilian world very much.


  1. You should read Horner, Thomas A. “Killers, Fillers, and Fodder.” Parameters:, Journal of the U.S. Army War College, September 1982.

    Horner goes into exactly these issues: a small number of men do the real work of killing, a larger number ("fillers") follow along, and a third group ("fodder") are either cowards or so inept that they are eliminated almost immediately.

  2. 'The answer is known long since; neither standing nor previous performance are of much importance. The most important attribute is potential for a job.'

    I'm not sure about that. Even with your cherished 'potential' you’re still guessing. I mean how can you evaluate the candidate's potential? Only by assuming based on past or current performance imho. You still don't know how he/she will perform in the future, maybe with a crystal ball...

  3. You can analyse the new job to learn about the requirements and the share of different tasks.

    Then you can look at the candidate and his past performance, but weigh the performance in certain tasks according to the new job.

    Old Job 90% Task A, 10% Task B
    New Job 10% Task A 90% Task B

    A candidate who's excellent in his old job can suck in the new one while a dud in the old one can be excellent in the new one.

    Now consider
    Other new Job 90% Task C and 10% Task B

    It's crucial to look for traits in this case - or to do tests that resemble Task C.
    - does he learn quickly?
    - ability to improvise
    - personal style
    - basic traits (like arrogance, humility)
    - resistance to stress
    - ability to organize his time well

    The potential-based selection of people for tasks is an area of research for hundreds if not thousands of personnel economics professors around the world, and it's indeed the next big thing after mere meritocracy.