During WWII the US Airforce had a problem. Its bomber fleet was sustaining unacceptably high losses on each raid over Europe. On some missions the chances of any one plane making it home were poorer than 50%; for once the word decimation was not hyperbole, but understatement.

Airforce command reasoned that more armour was the solution, but any increase in weight affected the distance, height and the duration of a mission. So they examined the bombers that had returned. They recorded every bullet hole, and overlaid their results.

They detected a pattern. Bullets were clustered around the wings, down the centre of the fuselage, and about the tail gunner. So they reinforced these areas, leaving the rest unclad – a trade off between strength and agility.

The result? Nothing. Statistically, losses remained the same. In bewilderment the airforce took their problem to a think tank called the Applied Mathematics Panel attached to Columbia University. There it came to the attention of a Hungarian emigré and mathematical genius called Abraham Wald.

Wald listened in silence before delivering a wholly unexpected solution, one which presaged a new approach to statistical analysis. “You’ve got it the wrong way round”, he told them.

What did he mean? Just this: the airforce had analysed only survivors, so had highlighted nothing more than the amount of damage a plane can take and still survive. A wise use of resources, Wald correctly hypothesised, would be to reinforce the areas which hadn’t been hit, since planes that sustained damage to those areas must now be lying at the bottom of the English Channel.

And how, you may be asking, oh patient reader, does this relate to the setting up of a creative studio? It’s like this: you have as much, if not more to learn from failure, as from success.