Do you read this newspaper every day? Enough to keep yourself moderately well-informed? Fancy a ministerial car? Good. Stand by your phone. I am about to recommend you to our new Prime Minister.

Gordon Brown has his own ideas, of course. He tells us that he intends to form an administration of “all the talents”, a category in which he eccentrically places Paddy Ashdown. His audacious attempts to recruit such people have captured headlines if not, yet, any talented people (Harriet Harman, I think we can agree, doesn’t count).

I believe that forming governments of “all the talents” is a silly thing to do and I am convinced that if anyone actually agrees to Mr Brown’s offer, the whole thing will end in a total mess. But I have been cheering him on, and not for mischievous reasons, either. I think, don’t you, that I’d better explain myself.

What does Gordon Brown mean by a “government of all the talents”? He can’t mean that everyone talented will be in it, and I think it is unlikely that everyone in it will be talented. Instead, it appears that he wants to augment the Labour administration with experts, even if they don’t support his party.

So the Single Transferable Policeman, Lord Stevens, was invited to tackle the crime portfolio, Julia Neuberger was sought so that she could bring into the Health Department the health expertise she has accumulated at the King’s Fund, various businessmen are being empanelled (and some may be ennobled) to report sagely on this and that. And so on.

What Brown is doing is taking the in-vogue political idea of the age and flipping it around. Instead of taking issues out of politics and letting them be settled by independent experts, an approach canvassed almost every day by somebody about something or other, he intends to bring the independent experts inside the government. The central idea – the superiority of the political judgment of experts – is maintained. And that central idea is wrong.

It’s easy to understand why we respect expertise. Nassim Nicholas Taleb fills many pages of his new book, Black Swans, with caustic comments about experts, but he still finds room for this question: “Would you rather have your upcoming brain surgery performed by a newspaper’s science reporter or by a certified brain surgeon?”

Taleb then identifies fields in which experts are useful – livestock judges, test pilots, brain surgeons, accountants – and those where, as he puts it “experts tend not to be expert” – stockbrokers, personnel selectors, intelligence analysts. Simply, he argues that “things that move”, requiring anticipation and prediction, do not usually have experts, while “things that don’t move” seem to have some experts. Politics is very definitely in the former category.

Taleb’s friend, Professor Philip Tetlock, has subjected this idea to empirical tests and the results are contained in his excellent book Expert Political Judgment. He has surveyed thousands of political predictions made over decades by observers with varying degrees of expertise, using the accuracy of their predictions as a way of gauging how well they understand the world.

The results show that expert political predictions are only very slightly better than those you can expect from a chimp. Tetlock then examines what he calls, er, “the diminishing marginal returns from expertise hypothesis”. According to this theory, people who read a newspaper and have some general acquaintance with the issues make mildly better judgments than would be made by a chimp, but your judgments don’t improve much with further knowledge.

This is counterintuitive, so let me provide an example of the problem with expert judgment. The FA Cup Final. Chelsea faced this year’s Wembley decider without their doughty Portuguese defender Ricardo Carvalho, who was nursing an injury. It takes a little knowledge to appreciate that this matters at all to the result. But experts, realising quite how good Carvalho is, proceeded to overstate vastly the importance of one player being absent. Why? Because people with knowledge have a tendency to regard the things they know as more important than they actually are.

Using a political example, look at the reaction to the defection to Labour of the Conservative MP Quentin Davies. You have to be a political expert to know exactly who he is and acquiring such knowledge is painful (trust me). Having done so it is natural to assume that what he does matters. If you have no idea who Quentin Davies is, or perhaps just a vague idea culled from the papers, then you are more likely to correctly understand that what he does doesn’t matter awfully.

So that’s why I see little advantage in trying to assemble a “government of all the talents”. Except in one, slightly surprising way. Introducing members of other parties and none would be a messy business and that’s a good thing.

Media and politicians are obssessed with tidiness, with unity, with everyone saying the same thing, “singing from the same hymn sheet”. But much that we have learnt recently about taking decisions suggests that such tidiness is not a good thing at all.

The best way of making good collective judgments is to aggregate many independent points of view. But tidy politics works on the opposite principle. It is organised as a conspiracy in which everyone defends everyone else’s mistakes.

In a “government of all the talents” one minister might say something different from another. How refreshing that would be. And how helpful in restoring trust in politics.

Take Harriet Harman. During the deputy leadership election she said that Labour should apologise for the Iraq war (she unquestionably did say this, by the way). Now she denies ever having done so. Why couldn’t she just say that she had one view and other ministers had another? What’s wrong with that? It isn’t tidy. We’d call it a mess.

That’s why I do so wish that the Lib Dems had said yes to Gordon Brown. Oh for the brave new era of messy politics.

Times Recommends

• Treasury needs a new golden rule fast
• Testing doctors is no magic cure
• We must cry out for the poor