In the day job I call Arsene Wenger a great economist. I'm making a serious point.

An economist is not someone who sits in an armchair pontificating about what's wrong with economics or - worse still - about the future. He is someone who tries to solve or ameliorate particular problems. As Simon says, economists are like doctors: much of my day job consists of giving advice on financial health.

From this perspective, football managers are economists. The task of picking a team is a constrained optimisation problem. And the job of a coach is largely to find the best solution to trade-offs. For example, do you press the opposition and risk being caught out of position, or do you sit back, keep your shape but concede possession? Do you play it long and have a high chance of giving the ball away in safe areas, or play it short and run the smaller but more costly risk of giving it away in your own third? Do you get men behind the ball and not concede goals but then lack options when you attack? And so on.

But here's the thing. When we judge whether coaches are making these trade-offs correctly, we often fall into error because of the interaction of two biases.

One is the outcome bias - our tendency to judge decisions by results rather than by circumstances at the time the decision was made: as Dostoyevsky wrote, "everything seems stupid when it fails". The other is a version of the error Frederic Bastiat warned us against - a tendency to look at the seen rather than the unseen.

Take, for example, something Arsenal are often criticized for - pushing fullbacks up too far and so being vulnerable to counter-attacks. It's easy to produce examples of this being costly - though I'll not bother. That's the seen. The unseen is what would happen if the full-backs stayed back. Somewhere in a parallel universe, Arsenal's fullbacks played deeper last Sunday, Nacho Monreal didn't win that penalty and Robbie Savage is criticising Arsenal for lacking an attacking edge. (One implication of the multiverse hypothesis is that there are an infinite number of Robbie Savages - and they call economics the dismal science.)

I say all this for two reasons.

One is that this point generalizes. In a complex world, even the best solutions to trade-off problems will entail some costs. Those solutions should not be criticized for this. For example, holding tracker funds will incur the costs of sometimes missing out on some profitable investments - for example when mega-caps underperform and so most stocks beat the market or when momentum stocks do well. But these costs must be weighed against costs avoided, such as high dealing charges. Decisions should be judged by their totality of costs and benefits, not by single costs. The same, of course, applies to macroeconomic policy-making: the cost of higher inflation must be weighed against the cost of higher unemployment. And it applies to politics generally.

Secondly, there is a close affinity between economics and sport; each can illuminate the other. I suspect you could learn more about economics from football than you could from the empty suits at Davos this week.