Ed Miliband says the BBC needs a "strong" DG. It's sad that the leader of a so-called leftist party should still subscribe to a crypto-fascist ideology of strong leadership.

To see my point, start from Lucy Kellaway's entirely correct analogy - that jobs are just like marriages. When we see a marriage we think likely to last, we don't usually say 'she's like Victoria Coren" but rather 'they are right for each other.' The same thinking should be true for jobs. What matters is not getting the 'best" person but rather the best fit between the organization's requirements and the manager's particular skills.

Research by Boris Groysberg at Harvard Business School has shown just this.He looked at the performance of former executives at General Electric when went onto become CEOs at other firms. He found that where the skills of the executive matched those needed by firms - for example, whether the executive was a cost-cutter, sales grower or manager of a cyclical business - the firm did well. But where there was a mismatch, it did poorly. Managers with similarly impressive CVs, then, can have very different performance.

Football fans, of course, should know this. There are countless examples of coaches doing well with one team and poorly at another. Think of Andre Villas-Boas at Porto and Chelsea, or Fabio Capello at Milan and England, Brian Clough at Leeds and Nottingham Forest. And so on.

What matters, then, is not having a "strong" leader, but rather the right leader. It's the match that matters, not (just) the individual.

In seeking a new DG, what Chris Patten should do is not ask: "who's the best candidate?" but rather "what exactly do we want the DG to do?" and "who has the requisite skills to do this precise job?"

This, though, raises a danger. In defining precisely what the new DG needs to do, there's a risk of over-reacting to current circumstances, and ending up hiring a general who's good at fighting the last war. For example, a DG who's good at cleaning up the standard of journalism might not be good at the many other things a DG has to do. Worse still, the qualities that equip a boss to do well at one task might well mean he does badly in others. For example, the man who is good at improving line-management process is also a stifling micro-manager; the cost-cutter fails to inspire his staff; and so on.

Even if one pays proper attention to matching, therefore, there's no assurance of making the right choice. But at least there's a better chance than there is if you stick to a silly ideology of egomanaical managerialism.