Diego Costa has something in common with George Osborne. I don't just mean the obvious. I mean that both could be beneficiaries of a common cognitive error - the outcome bias.

Some observers are praising Costa for engineering Gabriel's sending off. "There is barely a team on the planet who would not benefit from Costa's streetfighting approach" says Oliver Kay in the Times. "He deliberately and skilfully got an opponent sent off" says Barney Ronay in the Guardian. "This was wasn't a mugging. It was a heist, and an expert one." And Arry Redknapp adds that "you would like to have him in your team."

However, if the game had been reffed properly, Costa would have been red-carded and so Gabriel, having nobody to kick, would have stayed on the pitch. That might have cost Chelsea the game. We would not then be hearing praise for Costa, nor talk from Mourinho about emotional control.

In this sense, what we're seeing is the outcome bias. Costa's behaviour looks good not because it was skilful but because he had the good luck of having Mike Dean as ref. As Daniel Kahneman has written:

Hindsight bias has pernicious effects on the evaluations of decision-makers. It leads observers to assess the quality of a decision not by whether the process was sound but by whether its outcome was good or bad. (Thinking Fast and Slow, p203).

The outcome bias, he says, can "bring undeserved rewards to irresponsible risk seekers." Maybe Costa was one of these: he recklessly risked getting sent off himself.

This bias is a common one. It's common to praise bosses of successful companies, without asking whether that success was because of the CEO's decisions or just luck:Alex Coad's finding that corporate growth is largely random, and Ormerod and Rosewell's claim that bosses can't predict the effects of corporate strategy both suggest that we understate the latter. Here's Kahneman again:

Because luck plays a large role, the quality of leadership and management practices cannot be inferred reliably from observations of success.

A similar thing might happen in medicine: doctors get excessive praise for lucky but correct diagnoses and too much blame for reasonable but wrong ones.

Which brings me to Osborne. Some of his supporters regard the decent growth we've had had since 2013 as evidence that austerity worked. This too, is an example of the outcome bias - interpreting a bad decision as a good one simply because it eventually led to a happier outcome. But as Simon says, this is absurd:

Imagine that a government on a whim decided to close down half the economy for a year. That would be a crazy thing to do, and with only half as much produced, everyone would be much poorer. However, a year later when that half of the economy started up again, economic growth would be around 100 per cent. The government could claim that this miraculous recovery vindicated its decision to close half the economy down the previous year. That would be absurd, but it is a pretty good analogy to claiming that the recovery of 2013 vindicated the austerity of 2010.

There is, though, a problem here. The outcome bias isn't wholly stupid; as I've said, cognitive biases persist because they have a grain of usefulness. In an uncertain world, it's impossible to foresee the effects of choices and so it's hard to judge the quality of a decision at the time - especially if we don't know the decision-maker's information set. A good outcome might then alert us to the possibility that there was more wisdom in the decision than we thought at the time. I don't think this applies to Osborne's austerity - mainstream economics told us at the time that it was a bad choice - but it might apply to Costa. Maybe he judged that Mike Dean would not take a big decision to favour Arsenal and so thought he could get away with his behaviour. If so, praise for him is warranted. Sometimes, it's hard to distinguish between rational and lucky behaviour.