Watching some of Arsenal's recent performances has been like seeing the debate about Syria - and not just in the sense that both are utterly dispiriting.

What I mean is that in both recall to me the wise words of Charlie Munger - that "knowing the edge of your own competency" is a great skill. And it's a skill that has deserted Shkodran Mustafi lately. Against Palace, he attempted six tackles and missed all of them, often giving away free kicks in dangerous areas. And he and his team-mates have sometimes conceded goals by challenging for headers and missing them. In these respects, there's a tragic contrast with Tony Adams. One of his many great abilities was knowing which balls he could win and which not: if he couldn't win a header, he'd hang back and win the second ball rather than be caught out of position. Adams knew the edge of his competency: Mustafi does not*.

Which brings me to the Syria debate. This is a horribly complex problem: murdering children is of course a simple issue, but what we can effectively do about it is anything but. Any sensible discussion must have as its central tenet a recognition of the edge of our competency - the fact that we cannot predict the effect of interventions and that ground truth (which is what matters) is largely elusive. Unless you've devoted a lifetime to Syrian politics or military affairs, you're very likely over-confident about your opinion of what to do.

Such recognition, though, is often lacking. Some debates tell me plenty about tribal divisions within the Labour party and zippety-diddly abut Syria.

What I'm complaining about here is a very common phenomenon. To take another example, there was a piece in Monday's Times by Libby Purves telling us why Cressida Dick was the right woman to run the Met**. Now, Ms Purves has many virtues, but an in-depth knowledge of the Met's personnel has not hitherto been the most celebrated of them.

We might call it the columnists' fallacy (or the PPEists' error!) - a tendency to exaggerate one's knowledge and to ignore the edge of your competency. Chris Shaw is right to complain that, very often, "gaps in knowledge and complexity are tacitly ignored." I sometimes try to avoid this error and try to recognise ignorance in my day job - for example by claiming that we cannot know how great a problem is household debt or how worried we should be by the overvaluation of US equities. But nobody thanks me for it.

And herein, perhaps, lies the problem. Not only do pundits (and politicians) not recognise the edge of their competence, but there's little demand that they do. As Margarita Mayo says, humble people make the best leaders but it is arrogant narcissists who get hired or elected.

Acknowledging complexity and ignorance is not something many of us want to do.

* I'm setting an impossibly high bar here. Expecting defenders to be like Adams is like expecting novelists to be Fyodor Dostoyevsky or singers to be Hank Williams.

** I'm picking this example simply because it's the first that comes to mind. You can all no doubt think of other examples, by perhaps more serial offenders than Ms Purves.